Biotechnology Valuation Using Real Options

The Acquisition of Alder BioPharmaceuticals Inc. by H. Lundbeck A/S

Master Thesis

M.Sc. Economics & Business Administration – Finance & Investments

Gerrit Martin Jungen

S-123293
[email protected]

No. of Pages: 79
No. of Characters: 163,538
Submitted: 29th of May 2020

Supervisor: Rune Dalgaard

Spring Semester 2020

Abstract

Traditional discounted cash flow methods face inherent weaknesses when confronted with multi-staged
uncertainty such as drug development. Real options valuation holds promise but is in practice rarely adopted.
Meanwhile, big pharma continues to spend large sums on acquiring small biotechnology firms in the quest to
create shareholder value. One recent example is H. Lundbeck A/S buying out Alder BioPharmaceuticals Inc.
This thesis investigates if Lundbeck has created value and how a real options approach to valuation could help
understand if Lundbeck paid ‘fair value’. The migraine prevention market is discussed and modelled as an
input for a risk-adjusted discounted cash flow analysis of Alder’s lead asset Eptinezumab while an
epidemiology model is paired with a real options approach to value Alder’s early-stage asset ALD1910. A
verdict is reached after an event study is performed on Lundbeck’s share price reaction on the announcement
day.
Contents

1. Introduction ............................................................................................................................. 1
1.1 Problem Definition ................................................................................................................................... 2
1.2 Research Question & Hypotheses ............................................................................................................ 2
1.3 Relevance ................................................................................................................................................. 3
1.4 Background to the Biotechnology Sector ................................................................................................. 3
1.4.1 Introduction to Biotechnology .................................................................................................... 3
1.4.2 Drug Development in the U.S..................................................................................................... 4
1.4.3 Drug Approval in the European Union ....................................................................................... 6
2. Methodology ............................................................................................................................ 7
2.1 Philosophy of Science Statement ............................................................................................................. 7
2.2 Outline of Analysis Approach .................................................................................................................. 8
2.2.1 Company Overview, Market & Strategic Environment .............................................................. 9
2.2.2 Eptinezumab ............................................................................................................................... 9
2.2.3 ALD1910 .................................................................................................................................... 9
2.2.4 Price vs. Value Analysis ........................................................................................................... 10
2.2.5 Lundbeck Event Study .............................................................................................................. 10
2.3 Data and Sources .................................................................................................................................... 11
3. Theoretical Valuation Background ..................................................................................... 12
3.1 On the Critical Importance of Valuation ................................................................................................ 12
3.2 Overview of Valuation Methodologies .................................................................................................. 12
3.3 Discounted Cash Flow Valuation ........................................................................................................... 13
3.4 Real Options Valuation .......................................................................................................................... 15
3.4.1 Introduction to Real Options .................................................................................................... 15
3.4.2 Overview of Different Approaches........................................................................................... 17
3.4.3 MAD assumption ...................................................................................................................... 21
3.5 Advantages & Disadvantages of DCF vs. Real Options ........................................................................ 21
4. Empirical Industry Background.......................................................................................... 23
4.1 Drug Development Costs ....................................................................................................................... 23
4.2 Empirical Data on Clinical Trial Success Rates ..................................................................................... 24
4.3 Sales Forecasting & The Patent Cliff ..................................................................................................... 25
5. Analysis & Discussion ........................................................................................................... 26
5.1 Company Overview, Market & Strategic Environment ......................................................................... 26
5.1.1 Company Overview .................................................................................................................. 26
5.1.2 The Migraine Prevention Market .............................................................................................. 28
5.1.3 Strategic Environment .............................................................................................................. 29
 5.2 Valuation of Alder BioPharmaceuticals ................................................................................................. 34
5.2.1 Eptinezumab – DCF Assumptions & Model Construction ....................................................... 34
5.2.2 Eptinezumab – Valuation Results ............................................................................................. 42
5.2.3 ALD1910 – Static DCF Assumptions & Model Construction.................................................. 47
5.2.4 ALD1910 – Static Valuation Results ........................................................................................ 52
5.2.5 ALD1910 – Real Options Assumptions & Model Construction .............................................. 54
5.2.6 ALD1910 – Real Options Valuation Results ............................................................................ 57
5.2.7 SOTP Valuation (Standalone Case) .......................................................................................... 58
5.3 Price vs. Value Discussion ..................................................................................................................... 59
5.3.1 Transaction Overview at Announcement.................................................................................. 60
5.3.2 Purchase Consideration & Structuring ..................................................................................... 60
5.3.3 SOTP Valuation (Acquisition Case) ......................................................................................... 61
5.3.4 Price vs. Value Comparison ..................................................................................................... 62
5.3.5 Lundbeck’s strategic rationale .................................................................................................. 64
5.4 Event Study of Lundbeck’s Share Price Reaction .................................................................................. 65
5.4.1 Event study construction........................................................................................................... 65
5.4.2 Event study results .................................................................................................................... 67
7. Conclusion.............................................................................................................................. 69
7.1 Summary of Results ............................................................................................................................... 69
7.2 Limitations ............................................................................................................................................. 70
7.3 Avenues for Future Research ................................................................................................................. 72
II. References .............................................................................................................................. 73
Appendix A-1: Unlevered Free Cash Flow Derivation .................................................................... 88
Appendix A-2: Exemplary Decision Tree of R&D Projects............................................................. 89
Appendix B-1: Porter’s Five Forces Framework .............................................................................. 90
Appendix B-2: CGRP & PACAP Mechanism of Action ................................................................. 91
Appendix B-3: Standard Drug Sales Curve ...................................................................................... 92
Appendix B-4: Quarterly Sales and Market Share Development of CGRPs since Launch ............. 93
Appendix B-5: Overview of Preventive CGRP Forecast (I/II) ......................................................... 94
Appendix B-6: Overview of Preventive CGRP Forecast (II/II) ....................................................... 95
Appendix B-7: Benchmarking Comparables .................................................................................... 96
Appendix B-8: Beta Regression Analysis ......................................................................................... 97
Appendix C-1: ALD1910 Development Project Timeline ............................................................... 99
Appendix C-2: Epidemiology Model of ALD1910 ........................................................................ 100
Appendix D-1: EV-to-Equity Bridge: Treatment of Select Items .................................................. 102
Appendix D-2: Dilution .................................................................................................................. 103
Appendix E-1: Market Model Regressions ..................................................................................... 104
List of Abbreviations

Alder Alder BioPharmaceuticals Inc.

approx. approximately
bn billion
biotech biotechnology
BLA Biologics License Application
BSM Black-Scholes-Merton
ca. circa
capex capital expenditures
CAR cumulative abnormal returns
cf. confer
CHMP Committee on Human Medicinal Products
CGRP calcitonin gene-related peptide
CMO contract manufacturing organisation
CNS central nervous system
CVR contingent value right
D&A depreciation & amortisation
DCF discounted cash flow (to firm) analysis
EMA European Medicines Agency
EU European Union
EV enterprise value
FCFF unlevered free cash flow to firm
ICER Institute for Clinical and Economic Review
i.e. id est (Latin expression for “that is”)
IND Investigational New Drug Application
LOA Cumulative likelihood of approval
Lundbeck H. Lundbeck A/S
mAb monoclonal antibody
M&A mergers and acquisitions
MAD market asset disclaimer
mn million
MOA mechanism of action
NBI NASDAQ Biotechnology Index
NDA New Drug Applications
NOL net operating loss
NPV net present value
PACAP pituitary adenylate cyclase-activating polypeptide
P,P&E property, plant and equipment
R&D research & development
SEC U.S. Securities and Exchange Commission
SOTP sum-of-the parts (valuation)
TCJA U.S. Tax Cuts and Jobs Act
U.S. / USA United States of America
U.S. FDA U.S. Food and Drug Administration
USD United States dollar
vs. versus
WACC weighted average cost of capital
List of Tables

Table 1 – Overview of Valuation Methodologies ........................................................................................... 12

Table 2 – Option Parameters ........................................................................................................................... 16

Table 3 – Overview of Selected Drug Development Cost Estimates .............................................................. 23

Table 4 – Overview of Selected Drug Development Success Rates ............................................................... 24

Table 5 – Overview of anti-CGRP Inhibitors .................................................................................................. 30

Table 6 – Eptinezumab Approval vs. WACC Sensitivity Analysis ................................................................ 43

Table 7 – Sensitivity Analyses on the Peak Sales for ALD1910 .................................................................... 49

Table 8 – Overview of Volatility Estimates .................................................................................................... 54

Table 9 – Real Options Valuation – Sensitivity Analysis ............................................................................... 57

Table 10 – Overview of Event Study Results.................................................................................................. 67

List of Figures

Figure 1 – U.S. Drug Development Process Stages .......................................................................................... 5

Figure 2 – Outline of Analysis Approach .......................................................................................................... 8

Figure 3 – Cox, Ross & Rubinstein Binomial Tree (Two-Periods) ................................................................ 19

Figure 4 – Pipeline of Alder BioPharmaceuticals ........................................................................................... 26

Figure 5 – Share Price Performance of ALDR ................................................................................................ 27

Figure 6 – Total (Preventive) CGRP Sales Forecast ....................................................................................... 36

Figure 7 – Relative Market Shares of Forecasted CGRP Drugs ..................................................................... 36

Figure 8 – Weighted Average Cost of Capital (WACC) ................................................................................. 42

Figure 9 – Operating Model & FCFF Calculation (Eptinezumab) .................................................................. 44

Figure 10 – DCF & NOL Valuation (Eptinezumab - Standalone) .................................................................. 45

Figure 11 – NOL Valuation (Eptinezumab - Acquisition Case) ..................................................................... 46

Figure 12 – Operating Model & FCFF Calculation (ALD1910) ..................................................................... 52

Figure 13 – Static DCF & NOL Valuation (ALD1910) .................................................................................. 53

Figure 14 – Binomial Tree of Peak Sales Estimate for ALD1910 .................................................................. 55

Figure 15 – Real Options Valuation Tree for ALD1910 ................................................................................. 57

Figure 16 – Implied Sum-of-the-Parts Valuation of Alder (Standalone) ........................................................ 59

Figure 17 – Implied Sum-of-the-Parts Valuation of Alder (Acquisition Case) ............................................... 61

Figure 18 – Price vs. Value Comparison ......................................................................................................... 62

Figure 19 – Cumulative Abnormal Return in the Event Window ................................................................... 68

1. Introduction

“Price is what you pay; value is what you get”

(Warren E. Buffett – February 27, 2009)1

This basic principle of successful investing applies to many personal as well as institutional decisions every
day. Yet nowhere has this been more important than in the modern biopharmaceutical industry. For years,
well-known global biopharmaceutical companies have actively sought to replenish their drug pipelines, often
engaging in considerable mergers and acquisitions (M&A) activity with younger, innovative biotechnology
(biotech) firms in order to secure their survival, increase profit growth and create shareholder value.

In the process of doing so, company executives, venture and private equity investment professionals as well
as public market investors all need to evaluate investments in this promising but complicated industry. The
outcome of such an investment decision is often very much ‘binary’ in that a potential drug candidate is either
approved or fails at any stage of its tiered development process. However, most of these practitioners tend to
use standard valuation methodologies that face considerable difficulties when applied to the unique business
nature of biotech ventures. They do so for various reasons which often put practical convenience above
theoretical substance as shown by various scholars.2 On the other side, academic valuation theorists have long
claimed that so-called real option approaches can offer a remedy when assessing potential investments – be
that for example individual research & development (R&D) projects, M&A or licensing opportunities.3

In the pursuit to maximise shareholder value, management teams need to make informed decisions. In
particular, within pipeline-driven M&A, the valuation question thus becomes outright crucial – and risky.
Usually, one can observe how acquirors pay very significant premiums for (listed) biotech firms that have yet
to generate any positive cash flow or even revenues. To the outside observer, these premia in particular can
hardly be explained using conventional valuation approaches. If one assumes markets to be (semi-) efficient
as famously proposed by Fama (1970), it leads to the question of how big pharma can justify paying such steep
premia above the (public) market value for refuelling their pipelines. One recent example of this phenomenon
is Lundbeck’s acquisition of pre-revenue stage Alder BioPharmaceuticals in September 2019.

1 Quote from Berkshire Hathaway’s 2009 Annual Letter to the Shareholders (p. 5).
2 See for example Holmberg & Jeppsson (2010); Block (2007); Plumacher (2018); Puran (2005); Copeland & Tufano (2004) and Imam,
Barker & Clubb (2008).
3 See for example Trigeorgis (1993 & 2006); Copeland & Antikarov (2003); Van Putten & MacMillan (2004); Bogdan & Villiger

(2005a) and Plumacher (2018).

1
1.1 Problem Definition

This thesis seeks to investigate this question in H. Lundbeck A/S’ (Lundbeck) recent USD 1.95 billion (bn)
acquisition of pre-revenue biotech firm Alder BioPharmaceuticals Inc. (Alder). When analysing whether such
transactions are truly value-adding, traditional valuation approaches such as net present value (NPV) or
multiples that are widely used within the financial industry face critical weaknesses. For example, whereas
NPV approaches do not allow for managerial flexibility throughout time and sees the future project path as
fixed, multiples rely on (positive) value drivers such as revenue and the concept of business comparability.
These assumptions are hardly applicable for biotechnology firms. Hence, for an investment decision such as
Lundbeck’s more specialised knowledge is required to obtain an adequate understanding of ‘value’.

1.2 Research Question & Hypotheses

First and foremost, this thesis seeks to answer the following research question:

Can Lundbeck justify the price paid for Alder to its shareholders and can real options
help understand this? Has value been created?

To assess this problem, first the theoretical background around discounted cash flow analysis (DCF) as well
as real options analysis will be presented before discussing advantages and disadvantages of both
methodologies. Then, the existence of common real options within the drug development process will be
investigated before forming an opinion on whether or not a respective valuation approach can meaningfully
enhance our understanding of Lundbeck’s rationale for acquiring Alder. Finally, the value of Alder is
investigated using discounted cash flow as well as real option techniques and compared to the offer
consideration from Lundbeck. In the end, an event study is conducted.

The following hypotheses represent a milestone roadmap for answering the above research question:

I. Real option valuation provides clear theoretical advantages over the traditional approach of the
discounted cash flow analysis
II. Real options are well-suited to capture the dynamics of drug development processes
III. A real options approach is well-suited to investigate the value of Alder
IV. A real options approach can help explain Lundbeck’s willingness to pay a premium of almost 100%
V. Lundbeck’s offer consideration for Alder can be considered ‘fair value’ under reasonable assumptions
VI. Lundbeck’s stock exhibited a statistically significant positive abnormal return on announcement day

In the end, we seek to form an opinion on how much ‘value’ Lundbeck is getting for the price they paid.

2
1.3 Relevance

Research has shown that analysts tend to favour standard valuation methods when evaluating biotechnology
investments (Holmberg & Jeppsson, 2010). Meanwhile, academics have long argued that the (real) options
approach to valuation is superior to traditional methods (Trigeorgis, 1993 & 2006; Copeland & Antikarov,
2003). However, despite extensive theoretical coverage on (real) options, even the pharmaceutical industry
rarely uses real options approaches (Hartmann & Hassan, 2006). Therefore, this research piece is relevant to
both practitioners and academics. For the former group, it showcases how to apply real options valuation in a
concrete case example. For the latter, it provides an applied example of real options within life sciences from
a transactional standpoint of view, thus opening up further avenues for future research. Furthermore, in general,
biotechnology remains one of the more opaque niches in (public) equity markets. Investors thus require an
understanding of medical science as well as the healthcare sector on top of finance in order to make informed
investment decisions. To aid their thought process, this thesis sets out to showcase a sophisticated way of
examining a biotech M&A transaction.

1.4 Background to the Biotechnology Sector

The biotechnology industry is highly specialised and requires knowledge within both medical sciences as well
as healthcare regulation and commercialisation. In 1.4.1, an introduction to the industry is given. Afterwards,
1.4.2 and 1.4.3 will explain to the reader how the drug development process works in the U.S. and Europe.

1.4.1 Introduction to Biotechnology

Within the healthcare industry, which can be seen to consist of sub-sectors such as healthcare services, medical
technology and life sciences, pharmaceutical firms develop, produce and market medication drugs. These
products can be divided into classic chemically-synthesized small molecules such as for example Aspirin® as
well as the newer – and strongly growing – branch of biologics, which are complex molecules stemming from
biological sources. The latter is the focus area of biotechnology as we think of it today and encompasses
innovations such as vaccines, immunotherapies, allergenics, gene therapy and recombinant therapeutic
proteins, amongst others (U.S. FDA, 2018d).

Biologics are usually created using cutting edge technology and often either represent the most effective
treatment option or the only one available, often for various orphan conditions (Kuick Research, 2014). The
high growth in biologics is interrelated to its pricing dynamics. IQVIA, a healthcare data science provider,
showed that biologics accounted for 93% of growth in U.S. net drug spending over 2014 to 2017 (IQVIA,
2018). Comparable to generic versions of small molecules, after patent and / or market exclusivity expiration,

3
biologic drugs usually face competition from so-called biosimilars (U.S. FDA, 2017). Regardless, the world’s
highest selling drug in 2019 remains Humira, a biologic monoclonal antibody (mAb) used to treat conditions
such as arthritis, psoriasis and Crohn’s disease, amongst others, and worldwide net sales of almost USD 20.0
billion (bn) (AbbVie, 2020).

The huge potential financial rewards from biopharmaceutical innovations are counterweighted by incredibly
risky and long-lasting development and trial processes. A 2013 study by the Tufts Center for the Study of Drug
Development found that it requires an average of around eight years and USD 2.6 billion pre-tax (until first
approval) to develop a candidate (DiMasi, Grabowski & Hansen, 2016). And while this estimate includes costs
for abandoned candidates as well, the general probability of success for an individual candidate entering the
first clinical trial stage has been estimated at around 9.6% to 12% in the U.S. according to the Biotechnology
Innovation Organisation (BIO, 2016) and the Pharmaceutical Research and Manufacturers of America
(PhRMA, 2019).

The biotechnology industry is highly concentrated in North America due to various factors such as massive
American healthcare spending, lower regulations (especially with regards to pricing), and a far better venture
capital environment. Even more so, hopeful European biotechs looking to raise capital via an IPO often decide
to do so on NASDAQ as it offers a deeper and less risk-averse capital market for these risky equities
(McKinsey, 2019).

1.4.2 Drug Development in the U.S.

The drug development process differs depending on jurisdiction. As explained above, the U.S. is by far the
dominant playground for biotechs. Often, new drugs are first brought to market there before companies seek
marketing authorisations in for example the European Union.

Developing a new drug in the U.S. requires a sequential process of research, clinical trials and regulatory
reviews in order to convince the U.S. FDA that a pipeline candidate qualifies for being allowed to enter the
market. Figure 1 showcases the standard stages of the U.S. regulatory drug development process.

Below, we can see that any new potential drug candidate generally originates from ongoing basic research,
whereby diverse scientific actors from the private sector, from academia, foundations as well as from
government contribute via studies and their expertise to the global understanding of diseases. During drug
discovery, researchers at a biotech firm (or a contract research organization) identify and select certain drug
candidates that seem promising to address a certain disease via a so-called mechanism of action (MOA).

4
 Figure 1 – U.S. Drug Development Process Stages

Source: PhRMA, 2019.

Note: IND = Investigational New Drug Application, NDA = New Drug Application, BLA = Biologic License Application.

After selecting a certain candidate, extensive pre-clinical trials including animal experiments are conducted to
get a first view on efficacy and safety before starting human testing.4 Up to this point, the timeline can vary
somewhat even though pre-clinical trials usually take two to three years’ time according to Flynn et al. (2017).
To initiate human clinical trials, an Investigational New Drug (IND) application to the U.S. FDA is required.
Then, a phase I trial can be started to first and foremost assess safety of the new drug candidate in question.
Furthermore, phase I trials are intended to assess pharmacokinetics as well as pharmacodynamics.56 These can
last several months and typically encompass a test group of less than one hundred subjects (U.S. FDA, 2018a).
Following (mostly) positive data readouts in phase I, a phase II trial can be constructed to further assess safety
but also in particular efficacy and potential side effects. Here, we often see a duration of up to two years and
several hundreds of study subjects (U.S. FDA, 2018b). Hence, we see a shifting focus to the drug candidate’s
risk-and-reward profile. Phase III clinical trials then seek to confirm and substantiate findings from phase II
with regards to efficacy and safety via a lager study involving many hundreds to several thousands of subjects
over a longer horizon such as one to four years. Even more so, phase III trials are so-called pivotal studies that
provide most of the data for an eventual regulatory approval decision because they are likelier to detect long-
term or rare adverse reaction (U.S. FDA, 2018c). If positive, the results from phase III also guide the scope of

4 Efficacy denotes the maximum response that can be achieved by applying a certain therapeutic’s mechanism of action (Holford &
Sheiner, 1981).
5 Pharmacokinetics refers to the body’s effect on the drug (M. Nordberg, J. Duffus und D. M. Templeton, 2004).
6 Pharmacodynamics refers to the drug‘s effect on the body (M. Nordberg, J. Duffus und D. M. Templeton, 2004).

5
the so-called label, i.e. the officially approved indications and safety warnings for each marketed drug (Flynn
et al., 2017).

Following positive results in phase III, the company can file for an U.S. FDA review by submitting either a
New Drug Application (NDA) for small molecule-drugs or a Biologic License Application (BLA) for biologic
drugs (PhRMA, 2019). The FDA then aims to issue a decision around ten months later at the so-called PDUFA
action date.7 Finally, the FDA either approves the drug application or reverts with a complete response letter
(CRL) outlining why an approval was denied (U.S. FDA, 2018b; Flynn et al., 2017).

After the approved drug starts being marketed, so-called post-approval safety monitoring is conducted. This
refers to the ongoing monitoring of adverse reactions (also known as pharmacovigilance) and continuous
researching of an already approved drug to keep control over longer-term side effects and investigate its
potential for other indications. Obtaining approval for further indications requires a supplemental application
(sBLA / sNDA) to the FDA (U.S. FDA, 2018c).

1.4.3 Drug Approval in the European Union

Developing a drug candidate in the EU requires to navigate a different regulatory procedure than in the U.S.
In general, there are two possible paths. Companies can either choose the (most common) centralised
procedure via the European Medicines Agency (EMA) through which they apply for a marketing authorisation
in the entire European Union (EU) as well as Norway, Liechtenstein and Iceland. This also yields a unified
label in these countries. This procedure is generally mandatory for new drugs aimed at treating “cancer, human
immunodeficiency virus (HIV), diabetes, neurodegenerative diseases, auto-immune and other immune
dysfunctions and viral diseases, as well as therapeutics derived from biotechnology (e.g. genetic engineering),
so-called advanced-therapy medicines (e.g. gene therapy) and rare diseases” (EMA, 2020). The EMA’s
Committee on Human Medicinal Products (CHMP) makes a recommendation which is then adopted and
translated into legislation by the European Commission, if approved (EMA, 2020).

The other path are the national procedures which are used by producers of generic products and non-
prescription drugs. The standard is the national authorisation in a single country of the EU. This yields a
marketing authorisation with a particular label solely in the respective country following the country’s own
regulatory process. Other variants include the mutual recognition procedure, which seeks to provide a
marketing authorisation for a certain member state for a product that is already approved in another member
state, and the decentralised procedure, whereby an applicant seeks simultaneous approvals in several EU
member states for a product that is not covered by the EU-wide centralised procedure (EMA, 2020).

7PDUFA refers to the U.S. Prescription Drug User Fee Act of 1992 which normed how the U.S. FDA can charge fees to fund the drug
approval process (PDUFA, 1992)

6
2. Methodology

In the following chapter, the methodology of this thesis will be presented. Section 2.1 will provide the
underlying philosophy of science that this research effort follows. In section 2.2, the analysis approach will be
presented alongside a discussion on why it has been selected. Section 2.3 then provides information on the
data used and its sources.

2.1 Philosophy of Science Statement

Basic philosophical assumptions inform our research design, research questions and the methods by which we
seek to answer these. Following the research ‘onion’ as proposed by Saunders, Lewis and Thornhill (2016, p.
124), this paper’s philosophy of science can be seen as pragmatism.

On an ontological level, the reality of corporate (equity) valuation is complex. Previous scholars have proposed
that valuation can be seen as part science, part art and part craft due to diverse noise that is inevitable in each
valuation (Fabozzi, Focardi & Jonas, 2018). Valuation reality is thus a practical consequence of the ideas of
the professional. On a sub-level, this work relies heavily on objectivism, as we generally deal with numbers
and facts where possible.

With regards to epistemology, this thesis seeks to investigate practical implications of applying options theory
to the valuation problem of pre-revenue biotech firms. Given this specific context, it therefore focuses on
“problems, practices and successful actions” (Saunders, Lewis & Thornhill, 2018, p. 137).

The role of values (axiology) can be seen as value-driven research. This thesis seeks to “contribute practical
solutions that inform future practice” (p. 143). It will do so via a practical case example and utilise a mix of
quantitative (i.e. DCF and option calculations) as well as qualitative methods (i.e. analysis of expected market
share).

This work follows a deductive approach by starting off with relevant valuation theory before examining a
specific problem in this context. Financial data is used to evaluate our hypotheses from section 1.2, and
ultimately, it seeks to verify the existing theory around real options within pharmaceutical R&D.

Hence, with regards to the methodological choice, this thesis is both an evaluative but also exploratory study
using a mono-quantitative method, i.e. a valuation model, to assess its hypotheses.

7
2.2 Outline of Analysis Approach

Firstly, the relevant theoretical valuation background for this thesis is presented. Afterwards, select required
empirical findings are presented which will be required for the following valuation analysis.

Alder BioPharmaceuticals has two pipeline drugs: Eptinezumab and ALD1910. The former is a late-stage
candidate awaiting U.S. PDUFA approval while the latter is an early-stage candidate straight out of discovery
stage. As a result, Eptinezumab is valued using a risk-adjusted DCF while both a risk-adjusted DCF and a real
options approach are used to capture the suspected significant flexibility value in ALD1910. Hence, these two
assets will be modelled separately before combining them to arrive at a total enterprise value for Alder.
Afterwards, the obtained value of Alder will be compared to the price Lundbeck offered. There, Lundbeck’s
strategic rationale will be touched upon before concluding with an event study on Lundbeck’s share price
reaction to assess if value has likely been created or destroyed. Figure 2 below provides an illustration of the
structure of this thesis.

Figure 2 – Outline of Analysis Approach

Source: Author.

8
2.2.1 Company Overview, Market & Strategic Environment

Before diving into the asset valuations, a company introduction is given in 5.1.1 alongside a market
introduction in 5.1.2. In addition, in 5.1.3 a strategic analysis following Porter’s Five Forces (Porter, 1979)
will be conducted to set the stage on which to value Alder’s assets and the competitive environment. Emphasis
is put on Eptinezumab as the near-commercialisation asset which is suspected to represent the majority value
of Alder as of September 2019. Significantly more information is available for Eptinezumab than for
ALD1910.

2.2.2 Eptinezumab

Given Eptinezumab’s near-commercialisation status, it will be modelled from 2019 to 2040 using a risk-
adjusted DCF to capture the whole expected lifecycle of the drug, if approved. In the beginning, the total anti-
CGRP market for migraine prevention is modelled using point-in-time analyst consensus estimates from
EvaluatePharma as well as polynomial regressions to estimate the increasing part of the expected sales
adoption curve for Eptinezumab and its three competitors up to their patent expirations. Emphasis is put on
benchmarking Eptinezumab to its competitors to reflect on the sales curve and build a defensible forecast.
Afterwards, the patent cliff and the following sales decline will be estimated using a standard penetration curve
as provided by the Bogdan and Villiger (2010). Then, the assumptions for the components of the expected
unlevered free cash flows are explained and justified in detail before discounting them. To do so, an appropriate
weighted-average cost of capital discount rate is estimated using historical data from Alder (NASDAQ:
ALDR) and the Nasdaq Biotechnology Index (NASDAQ: NBI) following a discussion of different beta
regression analyses of Alder and the S&P500 as well as the NBI. Particular focus is laid upon valuing the tax
shield from Alder’s accumulated net operating losses (NOLs), which are valued separately in order to precisely
depict Alder’s business operations.

Results are obtained for two cases: (1) Alder’s value standalone and (2) Alder’s value to an acquiror such as
Lundbeck.

2.2.3 ALD1910

ALD1910 is an early-stage candidate with very limited publicly available information. To account for the
uncertainty and thus managerial flexibility, a real options approach to valuation is applied in addition to a DCF.

After constructing a project schedule based on Bogdan and Villiger (2010), the potential market for ALD1910
as an early-stage drug candidate is sketched out by developing a simple bottom-up epidemiology model for
the main markets U.S. and the EU as suggested by Deardorff et al. (2018). Empirical survey findings by Puran

9
(2005) confirm this as the most prominent approach in practice. The result will also be cross-checked with the
median peak sales estimates of comparable drugs as proposed by Bogdan and Villiger (2010). Also here,
emphasis is put on understanding the value proposition and market potential to derive a defensible peak sales
estimate.

The value of ALD1910 is then estimated using the real options approach proposed by Bogdan and Villiger
(2010, p. 122) whereby focus is laid on the typical abandonment option within R&D (cf. 3.4.1.2). Upon
obtaining an educated peak sales estimate, a binomial tree is constructed for peak sales upon approval in order
to capture the high uncertainty for this very early-stage asset.

For the tree, the volatility is the critical, unknown input parameter. Unfortunately, its estimation is subject to
much debate as touched upon with the MAD assumption in 3.4.3. For the purpose of valuing Alder, one has
to weight the benefits of using a decisively more complex and computationally expensive approach such as
estimating volatility using Monte Carlo simulations and the benefits of having a parsimonious model, i.e. a
simpler model that accomplishes an appropriate level of explanation. As ALD1910 likely only has a fraction
of the value of Eptinezumab given its early-stage status, it was decided to opt for using Alder’s stock price
volatility as a project proxy as suggested by Damodaran (2005). While this ‘blended volatility’ is surely heavily
influenced by the market’s assessment of Eptinezumab’s near-term commercial prospects, it is somewhat
reasonable in that ALD1910 and Eptinezumab are similar biologics with effectively the same target market.

Following construction of the tree, the value of the project is estimated in each end node using a static DCF
analysis before working back through the tree to obtain a real options valuation. Decision points were installed
at the beginning of phase I, phase II, phase III, and at U.S. PDUFA approval.

2.2.4 Price vs. Value Analysis

This research effort aims to understand if Lundbeck’s decision to acquire Alder for almost USD 2.0 bn in
September 2019 can be justified using a prudent set of assumptions. The price vs. value balance of this
acquisition will be assessed by contrasting purchase consideration and the sum-of-the-parts (SOTP) valuation
of the acquired assets from an acquirors viewpoint. In addition, the contingent value right of USD 2.00 per
share and its strategic role will be discussed in-depth as it takes a potentially crucial role in this transaction.

2.2.5 Lundbeck Event Study

Lastly, an event study on Lundbeck’s share price reaction upon and around the announcement is conducted to
obtain the market’s immediate view on the transaction as either confirming or contradicting evidence for the
verdict of this thesis’ valuation analysis.

10
Pursuant to Fama’s (1970) efficient market hypothesis, one can expect that, given rational market participants,
the new information of a corporate transaction announcement should be instantaneously reflected in stock
price.

An event study following the methodology suggested by Benninga (2014) is conducted. The test hypothesis is
constructed as

𝐻0 : 𝐶𝐴𝑅 = 0

where it is assumed that the event has no impact on the stock price.

For this methodology, Jaroszek (2019) summarises the event study steps as follow: First the event is defined
and data is selected. Then, realised returns and abnormal returns are defined before the expected returns around
the event are predicted using the common market model as suggested by Benninga (2014). In this work, both
single abnormal returns as well as cumulative abnormal returns are investigated in order to capture the entire
event window.

The estimation window is set at 252 trading days in compliance with Benninga (2014) in order to obtain
reliable model parameters of the stock. The event window is set to start five days before the 16th of September
2019 in order to capture any potential leaks of information. The event window ends five days afterwards.
Regarding the length of the window one has to trade-off capturing more trading days versus less reliable
expected return estimates.

The statistical significance of the estimated (cumulative) abnormal returns are then tested using both the
distribution-reliant t-test as well as the non-parametric rank test. This controls for the necessary assumption of
the t-test that returns are normally distributed (MacKinlay, 1979; Corrado, 1989).

2.3 Data and Sources

This work uses solely publicly available information to maintain transparency and showcase how an individual
outside investor can apply DCF and real options valuation to a case such as Lundbeck’s acquisition of Alder.
All securities data has been sourced from Bloomberg. As a best practice, the adjusted closing price is used to
account for potential dividends and potential (reverse) stock splits. For matters of consistency, the valuation
date has been set to 15th of September 2019. Therefore, the last unpolluted stock price before the announcement
was observed on 13th of September 2019. Importantly, as this thesis investigates Lundbeck’s rationale for
acquiring Alder at a price of ca. USD 2.0 bn, a pro-forma effective transaction closing date was assumed to be
the 1st of November 2019 as per Lundbeck press release (Lundbeck, 2019a).

11
3. Theoretical Valuation Background

Before diving into the analysis of how Lundbeck can justify its valuation of Alder and how real options can
help us to understand this, we revisit why valuation is central to Lundbeck’s actions and provide the relevant
theoretical foundations of discounted cash flows analysis and real options valuation.

3.1 On the Critical Importance of Valuation

It is widely accepted that the overall goal of corporations should be the maximisation of their own – and thus
the shareholders’ – value. Therefore, according to Brealey, Myers and Allen (2014), management teams
frequently need to make investment decisions that aim to increase firm value. They hereby face a trade-off
between increasing capital expenditures and returning capital to shareholders. Should the firm elect to reinvest
capital into promising projects, then these should at the very least achieve the so-called opportunity cost of
capital as they deprive their shareholders of the opportunity to reinvest the cash themselves in other areas of
the financial markets (Brealey, Myers & Allen, 2014). Hence, to be able to make such decisions with informed
confidence, managers need to be well-versed in the theory of valuing investment projects. In the following, we
will revisit the most important ones for our purposes.

3.2 Overview of Valuation Methodologies

With regards to valuation, there does not exist just one “correct” approach. In fact, there are several different
genres of methodologies that tackle the valuation question from different angles. Table 1 below aims to provide
an overview of the most common approaches in corporate finance.

Table 1 – Overview of Valuation Methodologies

Present Value Approaches Relative Valuation Net Asset Value

• DCF-to-firm/equity • Enterprise value multiples • Liquidation value
• Adjusted present value • Equity value multiples • Replacement cost
• Economic value added • (Precedent transaction • Book value approaches
• Real options approaches multiples)
• Dividend discount model
• Residual income model
Source: A.T. Christiansen, 2018.
Note: Precedent transaction analysis in brackets as it simply represents enterprise / equity value multiples in a different way.

We can discover three branches of valuation methodologies. The (net) present value category is well-known

12
due to the widely used discounted cash flow (DCF) method. Present value approaches can be distinguished
into enterprise and equity valuation methods. The DCF is often used on an enterprise value (EV) level but can
also result in an equity valuation, if adjusted correctly. The adjusted present value approach, in contrast,
handles the interest tax shield separately while also arriving at an enterprise valuation. The economic value
added model is a so-called excess return model that separates book value and returns above what can be
expected given a required rate of return and derives an enterprise value. In similar fashion, the residual income
model derives at an equity valuation. Finally, the dividend discount model is another equity value model that
focuses on dividends from the business. Importantly, in theory, all of the aforementioned methods should yield
the same valuation result (Christiansen, 2018).

In relative valuation, we utilise the concept of so-called multiples which essentially derive valuation as a
multiplier of a certain value driver such as EBITDA for example. Depending on the value driver, this will
result in an enterprise or equity valuation. Practitioners also frequently spin the concept and look at
“comparable” transactions to obtain an indication of what “the market” was willing to pay for a – hopefully –
comparable business (Christiansen, 2018).

Finally, the net asset value category takes a “cost” viewpoint and is in particular used during situations such
as liquidations. In essence, it considers book values as well as replacement costs of for example property, plant
and equipment to derive a valuation that is less focused on potential future upside but the state of the asset
today, i.e. a cost approach instead of an income approach (Christiansen, 2018).

Due to the unique nature of the asset in question, i.e. a pre-revenue stage biotech, this thesis will utilise
discounted cash flow analysis as well as real options theory to analyse Lundbeck’s investment into Alder.

3.3 Discounted Cash Flow Valuation

Under the concept of (net) present value, any investment opportunity can be valued as the sum of its expected
future cash flows discounted at an appropriate opportunity cost of capital to compensate investors for risk and
time value of money (Brealey, Myers & Allen, 2014). In its essence, the DCF model can then be summarized
as
𝑇
𝐹𝐶𝐹𝑡 𝑃𝑉(𝑇𝑒𝑟𝑚𝑖𝑛𝑎𝑙 𝑉𝑎𝑙𝑢𝑒)
𝑃𝑟𝑒𝑠𝑒𝑛𝑡 𝑉𝑎𝑙𝑢𝑒 = ∑ 𝑡
+
(1 + 𝑟) (1 + 𝑟)𝑇
𝑡=1

where 𝐹𝐶𝐹 denotes (mean) expected free cash flow, 𝑟 denotes the project’s opportunity cost of capital and 𝑡
denotes time during the forecasting period. The so-called terminal or horizon value here refers to the
continuation value of the business after our forecasting period has concluded (Brealey, Myers & Allen, pp. 24
& 94, 2014).

13
The free cash flows are generally forecasted using an operational model of the asset built upon assumptions
regarding this asset. For the purposes of this research, the above FCF generally refers to the unlevered free
cash flow-to-firm (FCFF). A derivation can be found in appendix A-1. The appropriate discount rate 𝑟 for the
FCFF is the weighted average cost of capital (WACC) of the firm which represents the required return for
investors to risk their debt or equity capital in the firm of question instead of other businesses with similar risk
profiles Wessels, Goedhart and Koller (2015). It is constructed as

𝐸 𝐷
𝑟𝑊𝐴𝐶𝐶 = ∗ 𝑟𝑒𝑞𝑢𝑖𝑡𝑦 + ∗𝑟 ∗ (1 − 𝑀𝑇𝑅)
𝐸+𝐷 𝐸 + 𝐷 𝑑𝑒𝑏𝑡

where 𝐸 represents the (market) equity value of the company, 𝐷 the (market) financial indebtedness, i.e. the
capital structure and 𝑀𝑇𝑅 refers to the marginal corporate tax rate. The cost of equity 𝑟𝑒𝑞𝑢𝑖𝑡𝑦 is usually
determined using the capital asset pricing model (CAPM) developed by Treynor (1961, 1962; cf. French,
2003), Sharpe (1964), Lintner (1965a, 1965b) and Mossin (1966) as

𝑟𝑒𝑞𝑢𝑖𝑡𝑦 = 𝑟𝑓 + 𝛽 ∗ (𝑟𝑚 − 𝑟𝑓 )

where 𝑟𝑓 denotes the risk-free rate, 𝛽 denotes the measure of relative systemic market risk and 𝑟𝑚 the expected
return on an appropriate market.8 Depending on the asset in question, the cost of debt 𝑟𝑑𝑒𝑏𝑡 can be obtained in
several ways such as for example a debt CAPM, simply from book values (unless we handle distressed firms)
or from the weighted average yield to maturity of the collective financing agreements (Brealey, Myers & Allen,
pp. 490, 2014).

The terminal value, if applicable, can be estimated using a trifecta of methods including the dividend discount
model, an exit multiple of a comparable asset in a mature state or the so-called value driver method which
seeks to improve on the dividend discount model by accounting for more investment needs under higher
growth.

Subsequently, after discounting all expected future cash flows at an appropriate weighted-average cost of
capital, we arrive at an enterprise value valuation that can then be transformed into an implied equity value via
the EV-to-equity bridge whereby we subtract net financial indebtedness as well as other appropriate items
(Wessels, Goedhart & Koller, 2015).

Moreover, the DCF toolbox can be extended by adjusting cash flows for their probability of occurring, thus
yielding the risk-adjusted DCF / NPV approach.

8Treynor (1961, 1962), Sharpe (1964), Lintner (1965a, 1965b) and Mossin (1966) all independently developed the CAPM. Treynor’s
manuscripts "Market Value, Time, and Risk" (1961) and "Toward a Theory of Market Value of Risky Assets" (1962) were never
published (French, 2003).

14
3.4 Real Options Valuation

3.4.1 Introduction to Real Options

DCF analysis, while allowing for different cases as well as detailed projections tailored to the specific asset in
question, has an inherent weakness: it assumes that the project’s course throughout time is fixed on its valuation
date, i.e. managers act passively and do not interfere with the project’s development. This is hardly applicable
in the real world. According to Brealey, Myers and Allen (2014), the “DCF does not reflect the value of
management” as the “opportunity to make such decisions clearly adds value whenever project outcomes are
uncertain” (p. 561).

The managerial flexibility to make such real decisions is generally referred to as a real option. Managers can
create considerable tangible value by taking advantage of good outcomes while limiting losses in unforeseen
setbacks. A variety of real options exists which will be discussed in section 3.4.1.1.

3.4.1.1 Properties of Options

An option is the right, but not the obligation, to buy or sell something for a pre-specified price on or during a
pre-specified period of time. In congruence with the more prominent financial option theory, we can
distinguish between European and American options. European options can only be exercised at option
maturity whereas American options can be exercised at any time prior to option maturity. A (financial) call
option contract gives the holder the right to buy a security at a certain price on a certain date or during a certain
time. In contrast, a put option gives the right to sell something (Hull, 2018). According to Hull (2018), options
can be priced based on a set of six defining characteristics.

Firstly, the price 𝑆 of the underlying asset, be that a stock price for financial options or the asset value for real
options, determines the payoff of the option when exercised at a certain strike price. Thus, an increasing asset
value increases the value of a call option whereas a decreasing asset value increases the value of a put option.
Secondly, the strike price 𝐾 determines the option payoff together with the asset price and represent for
example the initial investment if we exercise an option to expand into a new market. In a call, the payoff
function becomes

𝑚𝑎 𝑥(𝑆𝑇 − 𝐾, 0) and in a put option 𝑚𝑎𝑥 (𝐾 − 𝑆𝑇 , 0)

The time to maturity 𝑇 denotes the time until the option expires. Longer time to maturity will naturally result
in more valuable options.

The volatility 𝜎 of the asset price represents uncertainty about the future development path of the asset. As
options have limited downside but theoretically unlimited upside, the value of both calls and puts increases as
underlying volatility increases given large price movements become more likely.

15
The risk-free interest rate 𝑟𝑓 affects option pricing in a much more opaque way. Hull (2018) explains how an
increase in interest rates, ceteris paribus, leads to a higher expected return from investors in the asset while, on
the other hand, higher rates will pressure the value of all other cash flows to the option holder, thus increasing
call values but decreasing put values.

Lastly, if the underlying asset pays out dividends, these will impact the asset price from the ex-dividend date
so that call values decrease but put options increase in value.

In a real option context with a project as the underlying asset, the parameters can be understood as seen below
in table 2.

Table 2 – Option Parameters

Financial Options Real Options

Current value of underlying Present value of underlying cash flows

Exercise price Investment cost

Time to maturity Time until opportunity disappears

Observed price volatility of traded asset Project value uncertainty

Risk-free interest rate Risk-free interest rate

Source: Haahtela, 2012.

3.4.1.2 Types of Real Options

Brealey, Myers and Allen (2014) consider four main types of managerial options: (I) the option to expand, (II)
the option to wait and learn, (III) the option to abandon, and (IV) the option to alter the mix of output or
production method. In addition, Hull (2018) also mentions (V) the option to extend an asset’s life and (VI) the
contraction option.

(I) The option to expand is a constantly occurring theme for most businesses. Often management decides
to invest into projects today in order to secure future opportunities. There lies value in the option to make
additional investments if uncertainty is resolved and the future turns out as expected or better than expected.
The option to expand can usually be understood as an American call option on the increased capacity with a
strike of the investment needed to expand (Hull, 2018). An example would be the acquisition of oil extraction
concessions in a specific basin even if current oil prices imply a negative NPV of operating the oil wells. If the
price outlook improves, we then can quickly start or scale-up production.

(II) The option to wait and learn is similar to the previous case in that we wait for uncertainty to resolve.
Deferring investments into positive NPV projects and thus losing initial cash flows today can make sense if

16
the value of the option to defer is higher than the present values of the initial cash flows (also known as
dividends with financial options on stocks) (Brealey, Myers & Allen, 2014). This is equivalent to an American
call on the project value (Hull, 2018). An example would be the decision to enter a particular new market
where we could either enter today with a first mover advantage or wait one year and see how for example the
market’s competition or regulatory regime develops.

(III) The abandonment option is the opposing side of the option to expand. Should the market development
take a turn for the worse, then management can decide to shut down or sell the project to prevent further costs.
This becomes desirable if the gain from abandonment is larger than the value of the unexercised option
(Brealey, Myers & Allen, 2014). Abandoning a project equals an American put option on the entire asset’s
value with a strike equal to the sale or liquidation value excluding costs incurred in closing down (Hull, 2018).
An example would be a pharmaceuticals firm that awaits the outcome of a critical trial.

(IV) The option to alter the output mix or production method can take different forms. In essence, its value
comes from having the opportunity to alter output according to market conditions or from having the capability
to switch production to a different product without having to invest into new property, plant and equipment or
human resources. This can be represented as an American call option. An example could be a chemicals firm
that can quickly adjust its production lines to produce disinfectant during a pandemic.

(V) The option to extend an asset’s life refers to the case whereby we can prolong the use of an asset for
a fixed payment. This is equivalent to a European call option on the underlying’s future value (Hull, 2018).
An example could be any kind of licensing agreement to potentially market a product after its initial life.

(VI) The contraction option refers to the flexibility in scaling down an undertaking. It can be seen as an
American put on the present value of lost capacity where the strike price equals the savings in future
expenditures at exercise (Hull, 2018).

Furthermore, the existence of several different real options within a company can easily lead to
interdependency, thus creating a decisively more complex problem to solve. For example, the option to expand
is only valid if the option to abandon has not yet been exercised while the option to abandon likely changes its
value depending on whether the option to expand has been exercised or not (Maier, 2017).

3.4.2 Overview of Different Approaches

There are several different approaches for valuing an option on a stochastically-evolving underlying. In
general, we can differentiate between discrete and continuous time approaches. In a discrete time setting, we
use lattice models such as the binomial, trinomial or even higher order trees. Here, value changes can only
occur at deterministic points in time. In contrast, continuous time models seek to model that changes can occur
at any time and can come in closed-from approaches such as the Black-Scholes-Merton (BSM) model for

17
pricing European options (Black & Scholes, 1973; Merton, 1973). Importantly, they yield the same result if
we let the length of the discrete time steps approach 0, i.e. the number of periods approach the continuous limit
(Haahtela, 2010).

In the following, the recombining binomial tree and the Black-Scholes-Merton formula will be presented in
3.4.2.1 and 3.4.2.2. Other methods will briefly be touched upon in 3.4.2.3.

3.4.2.1 The Binomial Tree Model

The classic two-way lattice model for financial options was institutionalised by Cox, Ross and Rubinstein
(1979) as well as Rendleman and Bartter (1979) following early work by Sharpe (1978).

Its basic idea is that the price 𝑆 of an underlying asset follows a random walk whereby it can increase by a
factor 𝑢 or decrease by a factor 𝑑 in each period in a discrete time setting, thus being recombinant. Assuming
the principle of no arbitrage holds, we can construct a hedged portfolio with an identical payoff in either state
by taking a long position of ∆ ∗ 𝑆0 in the underlying asset and a short position in an option on that asset9.
Hence, as this portfolio is risk-less, it is to be discounted at the risk-free rate to arrive at its value today.
Importantly, this implies that the underlying asset’s expected return based on real-world probabilities does not
matter when pricing the option. In fact, we price the option in relative terms to the underlying instead of
absolute terms. This is known as risk-neutral valuation, i.e. valuing a derivative in a world with risk-neutral
investors that only demand a compensation equivalent to the risk-free rate. As the real-world probabilities 𝑝
of the stock’s expected movements are already incorporated into its current price, it can in fact be shown that
the option’s price must be identical in both a risk-neutral and a (mostly) risk-averse world, hence eliminating
the need for a (difficult to find) risk-adjusted discount rate (Hull, 2018).

Subsequently, the Cox, Ross and Rubinstein (1979) model can be extended for further periods to increase
accuracy. In practice, Hull (2018) explains a typical number of periods of 30 or more. In case of valuing an
American option, it then becomes necessary to check between exercising early and keeping the option at each
node before maturity, while working backwards through the tree.

Given the risk-neutral pricing approach, we can construct a tree from the following parameters: ∆ ∗ 𝑡 as the
length of the discrete time periods, the factors 𝑢 and 𝑑 from above, as well as the risk-neutral probability of an
increase 𝑞 defined as

𝑒 𝑟𝑓∗∆𝑡 − 𝑑
𝑞=
𝑢−𝑑

9 𝑓𝑢 −𝑓𝑑
Where ∆= .
𝑆0 ∗𝑢−𝑆0 ∗𝑑

18
We obtain the risk-neutral probability 𝑞 by first matching the underlying asset’s volatility with the factors 𝑢
and 𝑑 so that the expected return using 𝑞 becomes the risk-free rate. Cox, Ross and Rubinstein (1979) propose
the following definitions of the factors 𝑢 and 𝑑

𝑢 = 𝑒 𝜎∗√∆𝑡 𝑑 = 𝑒 −𝜎∗√∆𝑡

Figure 3 below illustrates a typical Cox, Ross and Rubinstein (1979) binomial tree.

Figure 3 – Cox, Ross & Rubinstein Binomial Tree (Two-Periods)

Source: Own Illustration based on Hull (2018, p. 305).

Finally, when increasing the number of time steps to infinity, the price of a European option in the binomial
setting converges to the exact price as given by the Black-Scholes-Merton closed-form solution as it has
identical underlying assumptions (cf. next paragraph).

3.4.2.2 The Black-Scholes-Merton Model for European Options

In 1973, a closed-form model for pricing European options was invented by Black, Scholes and Merton (1973).
The BSM model’s backbone is the BSM partial differential equation that defines the price of any derivative
on an underlying stock via

𝜕𝑓 𝜕𝑓 1 2 2
𝜕2𝑓
+ 𝑟𝑓 ∗ 𝑆 ∗ + ∗ 𝜎 ∗ 𝑆 ∗ 2 = 𝑟𝑓 ∗ 𝑓
𝜕𝑡 𝜕𝑆 2 𝜕𝑆

The BSM partial differential equation and the (original) closed-form formula are based on numerous
assumptions:

19
(I) Stock prices follow a geometric Brownian motion process defined as 𝑑𝑆/𝑆 = 𝜇𝑑𝑡 + 𝜎𝑑𝑧 ;10
(II) Short selling of securities and using full proceeds is permitted;
(III) No transaction costs or taxes and all securities are perfectly divisible;
(IV) No arbitrage principle holds;
(V) Trading happens continuous;
(VI) The risk-free rate is constant and identical for all maturities.

Therefore, in particular, stock prices are assumed to be distributed log-normally while returns follow a normal
distribution. Then, the price of an European call 𝐶 or put 𝑃 on a non-dividend-paying stock can be solved as

𝐶 = 𝑆0 ∗ 𝑁(𝑑1 ) − 𝐾 ∗ 𝑒 −𝑟∗𝑇 ∗ 𝑁(𝑑2 ) 𝑃 = 𝐾 ∗ 𝑒 −𝑟∗𝑇 ∗ 𝑁(−𝑑2 ) − 𝑆0 ∗ 𝑁(−𝑑1 )

with

𝑆 𝜎2 𝑆 𝜎2
𝑙𝑛 ( 0 ) + (𝑟 + ) ∗ 𝑇 𝑙𝑛 ( 0 ) + (𝑟 − ) ∗ 𝑇
𝐾 2 𝐾 2
𝑑1 = 𝑑1 = = 𝑑1 − 𝜎 ∗ √𝑇
𝜎 ∗ √𝑇 𝜎 ∗ √𝑇

where 𝑁(𝑑2 ) is the cumulative distribution function of the standard normal distribution and corresponds to the
probability of exercise (Hull, 2018).

3.4.2.3 Other Option Pricing Approaches

In addition to the binomial and the BSM model, there exists a variety of literature on different approaches to
solve in particular American options as well as more complicated exotic options.11 Moreover, there are also
diverse methods to price options on other underlyings such as futures, interest rates, bonds, swaps, etc.

One prominent area are Monte Carlo approaches as proposed by Boyle (1977) and further explored by Boyle,
Broadie and Glasserman (1997) as well as Broadie, Glasserman and Jain (1997). A popular framework is given
by Longstaff and Schwartz (2001) who value American options via a least squares regression Monte Carlo
simulation. They use a least squares regression to estimate a conditional expected payoff from continuing
versus exercising an option. The key advantage of Monte Carlo simulations is generally that they can handle
path-dependency and multi-factor situations. On the other side, they tend to be computationally expensive
(Hull, 2018).

Another major area are so-called finite difference approaches whereby one directly uses a partial differential
equation such as the BSM version from above. In essence, the partial differential equation is solved by iteration
of several difference equations. By working backwards from the maturity date, finite difference methods share

10 Where z denotes a Wiener process.

11 The term ‘exotic’ option was coined by Rubinstein and Reiner (1992) and includes non-standard payoff structures such as for example

Bermudan, Asian, barrier, rainbow or basket options.

20
certain characteristics with tree methods. In particular, explicit finite differences are effectively the same as a
trinomial tree according to Hull (2018).

Importantly, real options differ in some regards from financial options as they represent non-standardised and
non-securitised managerial choices. This has significant implications with regards to which methods can be
applied to value these real options and will be discussed in section 3.4.3.

3.4.3 MAD assumption

While the value of the underlying asset today is readily available in a financial option context, it is most likely
not in a real option context. This has profound implications as it indicates that there exists no replicating
portfolio, i.e. the use of the risk-free rate / risk-neutral probabilities to discount in the binomial tree as seen in
3.4.2.1 becomes theoretically inconsistent. Some real option advocates such as Copeland and Antikarov
(2003), Brandao, Dyer and Hahn (2005), Smith (2005), Datar et al. (2007) or Schneider et al. (2008) have
proposed variations of the so-called market asset disclaimer (MAD) approach as a remedy. The MAD
argument is essentially that the best estimate for the underlying asset value today is the present value of its
cash flows instead of the (non-existing) replicating portfolio. Effectively, the assumption is that there is a
hypothetical ‘liquid’ market for each (private) firm or project where the market price is agreed upon via a
universally known DCF. This is also somewhat supported by empirical evidence such as for example Kaplan
and Ruback (1995). The MAD assumption is prominently required when running Monte Carlo simulations to
estimate the project volatility.

Other scholars such as Bogdan and Villiger (2010) argue that the MAD assumption and even the underlying
assumptions of the BSM model simply do not hold in reality and that, hence, risk-neutral valuation should not
apply to real options that value real business opportunities. Subsequently, real options valuation becomes a
“risk-adjusted expectation” (p. 64).

3.5 Advantages & Disadvantages of DCF vs. Real Options

Traditionally, capital budgeting has been dominated by the concept of NPV / discounted cash flow analysis.
Given its assumptions, it intuitively provides the academically correct value of a project or firm in a straight-
forward and easily comprehensible format. It therefore remains the most prominent tool for financial managers
(Hartmann & Hassan, 2006; Block, 2007).

Despite its widespread use, Myers (1984) points out several weaknesses of standard DCF analyses such as the
difficulty of determining an appropriate discount rate, the difficulty of properly forecasting (far out) cash flows,
the difficulty of anticipating and modelling cross-sectional relationships between projects, and the challenge

21
of correctly estimating the project’s impact on the firm’s future investment opportunities. Myers (1984)
emphasises that while DCF analysis is comparatively appropriate for safe and stable investments, it can hardly
handle firms with plenty of (uncertain) growth opportunities. While scenario and sensitivity analysis can aid
the thought process of the DCF analyst, they do not cover for the shortcoming that DCF analysis effectively
implies a passive role of management, i.e. it ignores managerial flexibility. In addition, Copeland and Tufano
(2004) criticise that the concept of weighted-average cost of capital as the appropriate discount rate is widely
accepted, yet its inherent assumptions, i.e. that the firm maintains a constant capital structure and that the
riskiness of the projected cash flows does not change over time, remain far from reality (Myers, 1984;
Copeland & Tufano, 2004).

In particular, Myers (1984) singles out that investments with mostly intangible assets such as R&D are
effectively dominated by option value. In support of this, Bogdan & Villiger (2005b) argue that around 30%
of pharmaceutical R&D projects are abandoned for economic reasons. As laid out in section 3.4, real options
can more appropriately capture the diverse multi-staged uncertainty of such investments (Copeland & Tufano,
2004; Bogdan & Villiger, 2005b).

However, despite being predicted as the coming revolution of corporate decision-making (Trigeorgis, 1993 &
2005; Copeland & Antikarov, 2003), real options are also subject to criticism for several reasons. Firstly,
corporate managers often complain that real-life options are decisively more complicated than can be depicted
by the approaches laid out in 3.4.1. It therefore induces error potential in that it overly simplifies complex
situations, especially if applying the standard Black-Scholes-Merton closed-form model (Copeland & Tufano,
2004). Copeland and Tufano (2004) counter these claims by proposing the use of the binomial tree which can
be customised to nearly all situations. Secondly, they concede that a material drawback of real options
valuation remains the “disconnect between the way managers value options and the way they manage them”
(Copeland & Tufano, 2004), i.e. the assumption that options are exercised optimally does not necessarily hold
in reality. Another major drawback is represented by the ongoing debate about the MAD assumption.

Regardless of the MAD discussion, Bogdan & Villiger (2006) summarise that “real options solve the
shortcomings of the DCF at the cost of increased complexity” (p.180-181) while Myers (1984) concludes that
the combined use of DCF and real options valuation models can yield tangible benefits for corporate decision-
making as it promises to – and can – far better capture managerial flexibility amid multi-stage uncertainty.

As a consequence, hypothesis (I), namely that real options provide clear theoretical advantages over the (risk-
adjusted) DCF method, can be confirmed. However, the benefits from its application will likely vary between
industries and situations.

22
4. Empirical Industry Background

In this chapter, we assess why the biopharmaceutical industry might be well-suited for a non-standard, tailored
valuation approach such as real options. The following section 4.1 will provide necessary empirical findings
on the biotech-specific drug development process (cf. section 1.4) which will be crucial in chapter 5 later on.

4.1 Drug Development Costs

As we have seen in section 1.4, the drug development process has several phases and takes around ten years
to bring a product to market.

To properly model early-stage pipeline projects, a cost plan is needed as traditional approaches such as
percentages of sales do not apply to biotech firms without any revenue for example. The most cited cost
estimates stem from DiMasi et al. at the Tufts Centre for Drug Development. Results of select empirical
investigations can be observed in table 3 below.

Table 3 – Overview of Selected Drug Development Cost Estimates

Mean cost

Period Phase I Phase II Phase III

Study Focus
(Inflation Year) (2019 Prices) (2019 Prices) (2019 Prices)

DiMasi et al. 1990-2010 USD 25.3 mn USD 58.6 mn USD 255.4 mn

Pharmaceuticals
(2016) (2013) (USD 27.9 mn) (USD 64.7 mn) (USD 282.0 mn)

DiMasi et al. 1990-2003 USD 32.3 mn USD 37.7 mn USD 96.1 mn

Biologics
(2007) (2005) (USD 41.5 mn) (USD 48.4 mn) (USD 123.5 mn)

DiMasi et al. 1980-1999 USD 15.2 mn USD 23.5 mn USD 86.3 mn

Pharmaceuticals
(2003) (2000) (USD 21.9 mn) (USD 33.8 mn) (USD 124.2 mn)
Note: The studies usually rebase costs to the inflation year using the GDP implicit price deflator.

Above, one can observe how costs increase with each stage in the clinical trial process as trials become more
sophisticated and enlist larger numbers of individuals. Also, the mean costs have tended to increase above
inflation over time.

23
4.2 Empirical Data on Clinical Trial Success Rates

Each drug development phase has a binary decision point at each node. Appendix A-2 illustrates this as a
decision tree (Rogers, 2002).

In order to appropriately value a biotech company like Alder, we need to risk-adjust our cash flows with these
trial success rates to capture the underlying dynamics of drug development. Several scholars have investigated
this empirically with regards to different stages (DiMasi, Feldman, Seckler & Wilson, 2010; DiMasi,
Grabowski & Hansen, 2016) as well as different disease areas (Hay, Thomas, Craighead, Economides &
Rosenthal, 2014). The most recent and comprehensive study to date was performed by Hay, Thomas, Burns,
Audette, Carroll and Dow-Hygelund (2016) which updates Hay et al. (2014). In particular, they use a
significantly larger data set compared to other scholars. Table 4 below provides an overview of selected results
on their estimated success probabilities depending on respective stage, disease area and molecular type which
will become relevant in chapter 5.

Table 4 – Overview of Selected Drug Development Success Rates

No. of Obs. Pre-clinical Phase I Phase II Phase III NDA/BLA LOAa

All indications 9,985 - 63.2% 30.7% 58.1% 85.3% 9.6%

Biologics (all) 2277 - 66.0% 34.3% 57.2% 88.4% 11.5%

Neurology (all) 1304 - 59.1% 29.7% 57.4% 83.2% 8.4%

mAbsb 329 - 70.1% 38.1% 60.7% 86.8% 14.1%

Source: Hay et al. (2016), Hay et al. (2014).
Time period covered: 2006-2015. (2003-2011 for mAbs).
a LOA = cumulative probability of U.S. FDA approval for a drug entering phase I.
b mAb statistics as per Hay et al. (2014).

Above, one can observe how new investigational drugs face their biggest hurdle when moving from phase II
to III. Importantly, even if phase III trials succeed, a U.S. FDA approval of a NDA / BLA is still subject to a
review where between 10 to 20% of candidates fail. Overall, around 10% of all candidates starting a phase I
trial eventually achieved FDA approval over 2006 to 2015. The whole study shows that significant differences
between different disease areas (e.g. neurology) and molecule types (e.g. mAbs) do exist.

Furthermore, there also exists empirical evidence on the subsequent success rate of marketing authorisation
applications in major markets such as the European Union following an approval by for example the U.S.
FDA. Kashoki et al. (2019) find that EMA and the U.S. FDA agreed in more than 90% of the cases over 2014
to 2016.

24
Overall, the staged, binary uncertainty in the drug development process significantly impacts our expected
cash flows from a new R&D project and both allows for as well as demands managerial flexibility. The value
of these intangible assets can likely best be captured via (real) option valuation approaches as implied by Myers
(1984).

Therefore, hypothesis (II), can be confirmed. The real options approach is the only method that can accurately
depict the multi-stage uncertainty of drug development. However, the earlier the development stage the higher
the benefit of a real options view is likely to be.

4.3 Sales Forecasting & The Patent Cliff

The ‘tricky’ part of pharmaceutical sales forecasting is the so-called drug sales curve and the impact of the
patent cliff. When a firm starts a new drug development process for a promising compound, it will apply for a
patent to protect its intellectual property. In the U.S., patents are initially granted for 20 years, excluding any
possible extensions (Title 35 of the United States Code). As drug development phases frequently take up to 10
years, the marketed period is usually around 12 years according to Bogdan and Villiger (2010). Subsequently,
one can usually observe an upwards sloping sales uptake that slows down towards the so-called ‘peak sales’
year. After patent expiry, sales start to decline with a “sudden” strong drop upon entry of cheaper generics /
biosimilars which take advantage of the high prices for branded drugs (Calo-Fernández & Matínez-Hurtado,
2012). Usually, old drugs are discontinued if severe cheap generic competition takes hold. A typical example
curve can be seen in appendix B-3.

25
5. Analysis & Discussion

In this chapter, a valuation of Alder BioPharmaceuticals will be conducted before comparing the resulting
value estimate to the price Lundbeck paid in 2019. Before deep-diving into the numbers, a qualitative analysis
of the fundamentals of Alder’s business as well as its competitive situation are presented to familiarise the
reader with the object of this study.

5.1 Company Overview, Market & Strategic Environment

As outlined in section 1.4, the business model of biopharmaceutical firms entails the – comparatively risky –
development of new innovative medicines against significant upfront investments with the promise of sizeable
financial rewards a couple of years down the road.

5.1.1 Company Overview

Alder BioPharmaceuticals Inc. is a Bothell, Washington, USA based clinical-stage biopharmaceutical firm
specialising in therapeutic monoclonal antibodies to treat central nervous system (CNS) diseases. Its lead
pipeline candidate is Eptinezumab, a monoclonal antibody (mAb) that is intended to prevent migraine in adults.
It is envisaged to be administered via a quarterly 30-minute intravenous infusion and will attempt to suppress
calcitonin gene-related peptide (CGRP), which is suspected to mediate and initiate migraine (Alder, 2019).
According to its acquiror H. Lundbeck A/S, a Denmark-based firm focusing on CNS conditions, it will be the
first intravenous therapy of its kind, if approved (Lundbeck, 2019a). In addition, Alder’s pipeline also features
ALD1910, a mAb inhibitor that seeks to block pituitary adenylate cyclase-activating polypeptide (PACAP)
which has been shown to stimulate migraine, too. Alder’s pipeline can be summarised as follows in figure 4.

Figure 4 – Pipeline of Alder BioPharmaceuticals

Source: Alder BioPharmaceuticals, 2019 (website via web archive).

26
For Eptinezumab, Alder filed a so-called biologics license application (BLA) with the U.S. Food and Drug
Administration (U.S. FDA) in early 2019 and got allocated a PDUFA committee on 21 st February 2020
(Lundbeck, 2019a). ALD1910, on the contrary, started its phase I trial at the end of 2019 (Hoffman, 2019).

With regards to hypothesis (III), it can be inferred that the real options approach offers a well-suited alternative
to the DCF for investigating the value of ALD1910 due to its very early-stage status. For Eptinezumab, the
real options approach is likely of limited use compared to a risk-adjusted DCF.

Alder went public on NASDAQ on 8th May 2014. Its share price performance can be observed in figure 5
below.

Figure 5 – Share Price Performance of ALDR

Since IPO Last Year before Delisting

60 80 25 60

45 60 20 45
U.S. Dollars

Millions
30 40 15 30

15 20 10 15

5
May/14 Sep/15 Jan/17 Jun/18 Oct/19 Oct/18 Jan/19 Apr/19 Jul/19 Oct/19
Volume (m) Share price (close, adj.) Volume (m) Share price (close, adj.)

Source: Bloomberg.

Above, one can observe how Alder’s market capitalisation has fluctuated widely throughout the years before
trending down to a share price of USD 10.06 before the Lundbeck takeover announcement. One can also
observe significant volume spikes in March 2016 and August 2017 which were likely due to the release of trial
results.

27
5.1.2 The Migraine Prevention Market

Migraine is estimated as one of the top five leading causes of ‘years lived with disability’ globally by the
Global Burden of Diseases, Injuries, and Risk Factors Study 2016.12 In 1998, its economic impact due to loss
in productivity has been estimated at USD 13.0 bn yearly in the United States alone (Hu, Markson, Lipton et
al., 1999).

The Institute for Clinical and Economic Review (ICER, 2018) describes migraine as “a common, recurrent
headache disorder” with attacks that “are characterized by moderate-to-severe pain and other symptoms, which
may persist between attacks” (p. 1). Migraine comes in several forms but can broadly be identified as episodic
if a patient experiences fewer than 15 headaches per month, or as chronic if a patient experiences more than
15 migraine days per month, over a three month period. ICER (2018) estimates that migraine affects
approximately 20% of U.S. women and 6 to 10% of U.S. men, i.e. around 37 million Americans according to
the American Migraine Foundation (2020). 10% of U.S. migraine cases take the chronic form (ICER, 2018).

A study by Buse et al. (2016) found that chronic migraine in particular affects daily life. More than half of
survey participants reported missing family or other important social events due to migraine attacks while one
third of migraineurs worry about their financial security due to their headaches impacting their work
performance.

While legacy medications such as Botox exist, recent approvals of CGRP inhibitors are seen as a major
innovation that has “opened up the treatment of migraine” (p. 1) and will leapfrog legacy therapies, most of
which “were poorly tolerated or ineffective” (Bertels & Pradhan, 2019, p. 51). Bertels and Pradhan (2019)
point out that “over 50% of chronic migraine patients were dissatisfied with their treatment” (p. 51). The
approval of anti-CGRPs thus creates a new sizeable market opportunity. ICER (2018) sees the main use case
of anti-CGRP medication in chronic migraine patients for which previous prevention efforts have failed.

The second major (promising) innovation in migraine prevention is pituitary adenylate cyclase-activating
polypeptide (PACAP), another suspected cause of both episodic and chronic migraine attacks (Bertels &
Pradhan, 2019). However, for anti-PACAP therapy, no products are currently approved, and to date, only two
candidates are in early-to-mid-stage development (Do, Guo & Ashina, 2019).

5.1.2.1 CGRP & PACAP mAbs explained

CGRP inhibitors work as follows: If a person develops a migraine attack, they show a higher concentration of
CGRP peptide in their nervous systems. These signalling molecules (“ligands”) have been shown to trigger /

12‘Years lived with disability’ measures the burden of a disease as the product of its prevalence and its estimated loss of health. Global
coordinated research study by numerous authors referred to as ‘GBD 2016 Disease and Injury Incidence and Prevalence Collaborators’
and published in The Lancet (2017).

28
further inflammation reactions that occur in migraine pain upon them binding to a specific cell receptor.
Eptinezumab and its competitors then block this mechanism by either attaching to the receptor or the ligand
(Bigal, Walter & Rapoport, 2015).

PACAP inhibitors work analogously to CGRP inhibitors.

An intuitive illustration of the mechanisms of action can be found in appendix B-2.

5.1.3 Strategic Environment

Biopharmaceutical firms face three major risks according to Flynn et al. (2017). Firstly, development risk
during the (pre-)clinical trial phase. Afterwards, regulatory approval is the next major hurdle. Section 4.2 has
shown how the regulator can decide to reject applications despite favourable phase III evidence. If the drug
makes it to market, then commercialisation becomes the biggest risk. Here, the drug needs to be accepted –
and prescribed – by physicians and the biopharmaceutical firm needs to organise an appropriate insurance /
payor reimbursement scheme to ensure affordability for citizens.

Alder’s competitive position within the migraine field can be captured via Porter’s Five Forces framework
(Porter, 1979) which is illustrated in appendix B-1.

5.1.3.1 Rivalry among existing competitors

Alder’s lead candidate Eptinezumab is one of four viable anti-CGRP inhibitors for migraine prevention. Its
three competitors are already being marketed since 2018 by the larger firms Amgen (in ex-US collaboration
with Novartis), Eli Lilly & Co. and Teva Pharmaceuticals. An overview of these products, i.e. the market for
anti-CGRP inhibitors for migraine prevention, can be seen in table 5 below.

29
 Table 5 – Overview of anti-CGRP Inhibitors

Aimovig Emgality Ajovy

eptinezumaba
(erenumab) (galcanezumab) (fremanezumab)
Company Alder BioPharma. Amgen / Novartis Eli Lilly & Co. Teva Pharmaceuticals
(Ticker) (ALDR) (AMGN / NOVN) (LLY) (TEVA)
Net revenue ‘18A
- USD 23.8 / 51.9 bn USD 21.5 bn USD 18.3 bn
(Company)

U.S. market entry Q2 ‘20 June ‘18 Sept. ‘18 Sept. ‘18

Episodic migraine Episodic migraine Episodic migraine Episodic migraine

Approved
Chronic migraine Chronic migraine Chronic migraine Chronic migraine
indications
(Cluster headache)

Dosing frequency Quarterly Monthly Monthly Monthly / Quarterly

Intravenous Subcutaneous Subcutaneous Subcutaneous

Administration
Infusion Injection Injection Injection

Autoinjector - ✓ ✓ ✓

Initial U.S. list price $575 $575 $575

?
(yearly equiv.) b,c ($6,900) ($6,900) ($6,900)
1. 100 & 300mg 1. 70mg ffffffff 1. 120 & 240mg 1. 225 & 675mg
Phase III studies 2. 100 & 300mg 2. 70 & 140mg 2. 120 & 240mg 2. 225 & 675mg
(3. 300mg)……...
1. -0.7 & -1.1 days 1. -1.1 days / -1.9 1. -1.9 & -1.8 days 1. -1.5 & -1.2 days
Efficacy
2. -2.1 & -2.6 days 2. -1.4 & -1.9 days 2. -2.0 & -1.9 days 2. -2.1 & -1.8 days
(vs. placebo)d
(3. -3.5 per week)…
Injection-site &
Respiratory / urinary Fatigue, influenza,
Safety abdominal pain, Injection-site pain,
tract infections, cold, nausea,
(side effects) respiratory tract dizziness, fatigue
fatigue back pain
infections
Source: Cairns (2018); Young, Merle & Yuan (2017); Amgen, Eli Lilly & Co. & Teva Pharmaceuticals (2020).
a Eptinezumab’s brand name was announced as Vyepti after its U.S. PDUFA in 2020.
b List price is also known as wholesale acquisition cost, i.e. the cost at which a firm sells to wholesaler. The retail price can vary

depending on reimbursement route but is usually significantly lower. U.S. pricing data for competitors as of 2019. Eptinezumab
assumed unknown as of September 2019.
c Net prices can be approximated by applying an industry-wide average discount of 27% for branded drugs as per the Institute of

Clinical and Economic Review (2018).

d Efficacy refers here to the average reduction in monthly migraine days vs. the reduction for the placebo group.

We can observe several important facts in table 5 above. Firstly, unless Alder finds a sizeable licensing partner,
they will be competing against much larger rivals as indicated by revenue. In fact, its competitors are some of
the largest biopharmaceutical firms in the world with most likely far larger and better connected sales forces.
These competitors have already been marketing their products since 2018.

Importantly, Eptinezumab has a different value proposition. According to Lundbeck (Armstrong, 2019), its
quarterly intravenous infusion approach fits into the scheduled quarterly physician meetings of migraine
patients. Lundbeck also emphasises the 100% bioavailability of the infusion approach, i.e. Eptinezumab will

30
show its results as early as 30 minutes afterwards. In contrast, its competitors can be self-administered at home
via a subcutaneous injection ‘pen’ but can take several weeks to show meaningful reductions in migraine days.
It remains to be seen which route proves more attractive to physicians and patients.

With regards to efficacy, all four products have showed similar therapeutic gains as measured by the average
reduction in migraine days compared to the placebo group (Do, Guo & Ashina, 2019). Perhaps logically, the
competitors are currently marketed at identical (U.S. wholesale) prices. This can indicate a high price elasticity
for CGRP inhibitors given similar phase III efficacies and thus implies that rivalry is strong in the fight for
market share.

Overall, one can summarise that Alder is facing very strong rivalry in the anti-CGRP mAb space.

5.1.3.2 Threats from new entrants

Besides the strong competition as explaining in the previous paragraph, Alder and its competitors are also
facing certain indirect threats from new entrants in the future (Do, Guo & Ashina, 2019).

As of September 2019, both Allergan and Biohaven Pharmaceuticals have small molecule anti-CGRP
candidates with new mechanisms of delivery in phase III trials and / or filing stage. For migraine prevention,
Allergan has Atogepant and Biohaven its Rimegepant in phase III trials, both of which are taken in oral form
(Do, Guo & Ashina, 2019; Armstrong, Brown & Fagg, 2019).

Hence, there are potentially new entrants coming to market in 2020 to 2021. The oral administration in tablet
form requires more frequent (i.e. daily) application but eliminates the need for injection, thus people with
“needle fear” might be more inclined to choose one of these new competitors (if approved). Furthermore,
Cowan (2018) draws a comparison to the emergence of triptans in the 1990s stating that oral delivery systems
proved to be much more successful than injectables. Armstrong, Brown and Fagg (2019) also point out how
“the bar is also fairly low” with regards to efficacy when taking the existing four in table 5 as a benchmark.
The EvaluatePharma sell-side consensus forecast as provided by Armstrong, Brown and Fagg (2019) predicts
that these new entrants will represent approximately USD 1.2 bn in sales by 2024.

Besides other CGRP candidates, two PACAP candidates are currently being developed and can potentially
also pose a threat to Eptinezumab’s future sales. Amgen is developing AMG-301 in phase II while Alder is
developing its ALD1910 in phase I (Do, Guo & Ashina, 2019). Lundbeck (2019a) sees ALD1910 as a possible
alternative treatment for patients who react unsatisfyingly to other migraine therapies.

In summary, new entrants, if approved, could become a serious threat to Alder in the medium term.

31
5.1.3.3 Supplier bargaining power

The manufacturing of biologics involves more value-add than for traditional small molecule compounds
(Gooch, Cordes, Bressau & Berk, 2017). It seems very likely that Alder would outsource the procurement of
its product to a contract manufacturing organisation (CMO) unless it finds a larger licensing partner.
Otherwise, significant capital expenditures for biologics processing equipment would be required which is
hardly economic for a small biotech such as Alder. The supply of bioprocessing equipment is dominated by a
few large players such as Sartorius, Thermo Fisher Scientific, Danaher and Merck. Also, compared to Alder’s
larger rivals such as Amgen, Alder likely faces comparatively more supplier bargaining power from CMOs
due to lower overall order volume and smaller economies of scale.

5.1.3.4 Threats from Substitutes

For migraine prevention, a couple of alternatives to CGRPs (and PACAP) exist. Besides some anti-
depressants, beta-blockers and anti-seizure drugs, Botox is the most prominent substitute that can be prescribed
against chronic migraine (ICER, 2018). If patients are being prescribed Botox and it is working as intended,
then ICER (2018) sees “insufficient” evidence of a net added health benefit from prescribing CGRPs.
However, according to ICER (2018), a sizeable portion of people that receive such existing treatments (i.e.
Botox) do not see effective improvements or face serious side-effects, thus yielding tangible health benefits
for these cases.

On the other hand, after patent expiry for the CGRP blockers in the mid-2030s, biosimilar (i.e. “biologic
generics”) competition is likely to be strong (Calo-Fernández & Matínez-Hurtado, 2012). Even more so
because biologics pre-patent expiry have higher profit margins than small molecules (Gooch, Cordes, Bressau
& Berk, 2017), thus creating an attractive opportunity for biosimilar developers.

Alder’s product is addressing a very large market of a disease that has diverse causes and is notoriously difficult
to treat. Hence, it will likely be able to establish its niche until at least patent expiry. In addition, non-CGRP
substitutes should pose only a limited threat to Eptinezumab in the short-to-medium term due to advantageous
efficacy for the large number of cases where previous prevention therapy has failed. Biosimilars will quickly
reduce sales after patent expiry, however.

5.1.3.5 Buyer bargaining power

In 5.1.3.1, it was discovered how nearly identical pricing among the three marketed products can hint at high
price elasticity. Generally, biologics such as mAbs have significantly higher prices than (branded) small
molecules. At the same time, they offer innovative treatment options for patients that would otherwise suffer
from chronic conditions such as migraine which significantly reduces their quality of life. Usually, the
reimbursement structure, however, helps to alleviate a large part of the costs for the end-consumer.

32
While insurance systems and the medical reimbursement structure differ from jurisdiction to jurisdiction, one
can infer that CGRP inhibitors are not at the top-end in terms of pricing. In the U.S., ICER (2018) estimates
that, after applying an average industry discount of 27%, the CGRP inhibitors “meet commonly accepted
thresholds for cost-effectiveness” (p. 5) and “aligns with the benefits they provide to patients” (p. 5). Hence,
ICER (2018) confirms the affordability for the U.S. healthcare system unless the yearly number of
prescriptions are significantly exceeded.

Therefore, one can argue that the bargaining power of Alder’s buyers, i.e. the distributors, drug retailers and
the patients, is not critically high as the pricing (for now) seems reasonable based on ICER’s assessment.
Nevertheless, future price increases (above inflation adjustments) are not particularly likely as “reimbursement
by payers remains the major limiting factor to growth” (p. 1) due to the significant cost of biologics / mAbs
(Scolnik, 2009). Sacco et al. (2019) also mention that due to the significant price not all migraine patients will
be able to afford CGRP mAb treatments.

An assessment of buyer bargaining power for PACAP is more nebulous given its early- / medium-stage. No
pricing indication or independent economic assessment is available to date.

33
5.2 Valuation of Alder BioPharmaceuticals

In the following, Alder will be valued using a two-step approach whereby both Eptinezumab and ALD1910
are valued separately as assets and then combined to arrive at an enterprise valuation.

5.2.1 Eptinezumab – DCF Assumptions & Model Construction

To evaluate a drug development project, Flynn et al. (2017) focus on two main questions: (1) Will it work?
and (2) How big is the market?

Table 5 in section 5.1 has revealed that three similar CGRP agents have already been approved by the U.S.
FDA while Alder’s phase III trials of Eptinezumab have shown similar efficacy and similar adverse events
(“side effects”). Hence, one can suspect that Eptinezumab will likely be approved at its U.S. PDUFA in
February 2020, not least also because Eptinezumab offers a different schedule and mechanism of delivery with
its quarterly intravenous infusion, thus offering tangible clinical benefit to patients that do not respond well to
the other CGRP inhibitors.

5.2.1.1 Revenue forecast

With regards to the market size, several approaches are possible. Deardorff, Baylan, Trzcinska and Eskow
(2018) suggest to derive revenue forecasts either via a prescription-based model (for mature markets) or via
an epidemiology model (for novel products). Both approaches require specialised knowledge with regards to
their parameters and can quickly become error-prone. Research analyst consensus estimates, if available, offer
another (aggregated) way.

For the purpose of this thesis, it was decided to resort to (interpolated) analyst consensus estimates for
Eptinezumab due to a trifecta of reasons: (1) it has been assumed that dedicated sell-side industry analysts
following Alder’s stock since IPO in 2014 have a better understanding of Eptinezumab’s revenue prospects,
(2) available consensus estimates for the competing CGRPs allow us to model the entire (preventive) CGRP
market, i.e. perform different cross-checks, and (3) the three direct CGRP competitors have been on the market
since 2018 and thus offer several quarters of prescription uptake data which further lend credibility to the
consensus estimates.

Hence, this thesis models the entire (preventive) CGRP market, whereby ‘market’ here refers to cumulative
sales of currently approved CGRP inhibitors as well as oral solutions Rimegepant and Atogepant (both yet to
be approved, i.e. we assume approvals in both cases), in order to construct an informed, defensible sales
forecast for Eptinezumab.

34
Publicly available recent consensus revenue estimates were obtained from industry data provider
EvaluatePharma for 2019, 2020, 2022 and 2024 for Eptinezumab, Aimovig, Emgality, Ajovy as well as
Rimegepant and Atogepant.13,14

After careful consideration, annual 2021 and 2023 revenue for Eptinezumab and its approved competitors were
interpolated using a polynomial equation of 3rd order as given by

𝑓(𝑥) = 𝑎𝑥 3 + 𝑏𝑥 2 + 𝑐𝑥 + 𝑑

which was fitted using simple least squares regression. This approach best captures the typical drug sales cycle
curve with an increasing acceleration after start of marketing and a decreasing acceleration towards peak sales.
This approach was also deemed reasonable as enough historical data and estimates for Alder and its
competitors were available to regress the polynomial equation. Aimovig, Emgality and Ajovy started
marketing in 2018 with Aimovig being the first-mover. Their quarterly sales performance since launch is
exhibited in appendix B-4.

For Rimegepant and Atogepant, however, sales estimates for 2021 and 2023 were interpolated using a standard
drug sales curve as provided by Bogdan & Villiger (2010, p. 109), which is exhibited in appendix B-3. There
were simply not enough data points available to justify the same approach as above. Hence, it was deemed that
the only defensible assumption would be that of a standard drug sales cycle.

All six products were then further extrapolated using the sales curve from before to capture peak sales
development as well as the patent cliff upon patent expiry. Hence, all products are assumed to reach their peak
sales after 12 years on the market. Figures 6 and 7 below showcase aggregated results. The complete overview
can be found in appendices B-5 and B-6.

13 https://www.evaluate.com/products-services/pharma/evaluatepharma
14 Publicly available estimates were retrieved via Evaluate Vantage industry coverage articles (https://www.evaluate.com/vantage)

35
 Figure 6 – Total (Preventive) CGRP Sales Forecast

8,387 Atogepant
8,153 8,333 8,234
8,332
Total Sales (USD millions)

7,853 Rimegepant
7,384 7,098
6,744 Eptinezumab
5,935 6,127 Ajovy
5,079 5,037 Emgality
3,851 3,778 Aimovig
2,642 2,551
1,628 1,634
810 912
488 219
127 65 8

'18A '20E '22E '24E '26E '28E '30E '32E '34E '36E '38E '40E '42E

Note: Values refer to combined total sales of all six forecasted drugs.

Figure 7 – Relative Market Shares of Forecasted CGRP Drugs

100%

75%

50%

25%

0%
'18A '20E '22E '24E '26E '28E '30E '32E '34E '36E '38E '40E '42E

Source: Quarterly and annual reports of Amgen / Novartis, Eli Lilly & Co. and Teva Pharmaceuticals. Retrieved via SEC EDGAR
and company websites. 2019, 2020, 2022 and 2024 revenue analyst consensus estimates as per EvaluatePharma (Gardner, 2019;
Armstrong, Brown & Fagg, 2019). Bogdan & Villiger (2010).
Note: Smoothed lines for better illustration but identical underlying data as above.

Intuitively, the obtained forecast fits well into the schedule of patent expiries which were all assumed to be
without any potential extensions. For example, Amgen’s patent on Aimovig in Europe expires already in 2029,
thus slowly starting to put downwards pressure on all three subcutaneous CGRP products as these are quite
similar and will all inevitably suffer from biosimilars of Aimovig. Meanwhile, Alder’s Eptinezumab, due to
its unique mechanism of delivery and somewhat different clinical profile is estimated to reach peak sales in
2031. This is one year before Alder’s patent expiry but was assumed because much cheaper biosimilars based
on for example Aimovig will undoubtedly place a drag on Eptinezumab’s sales given its broader clinical profile

36
similarity. The two oral CGRPs were assumed to peak even later as their differentiated, perhaps preferable,
mechanism of delivery should sustain sales (cf. Cowan, 2018).

For Eptinezumab, a peak sales estimate of USD 1,207.8 mn in 2031 corresponding to a market share of 14.5%
was derived. This is at the lower end of 20% to 30% which Alder estimates in a market of four products (Alder,
2018b). This is however in line with what Lundbeck’s CEO said on the post-announcement conference call,
where she states that they consider Eptinezumab a “potential blockbuster” (p. 6), i.e. a drug with sales of more
than USD 1.0 bn (Lundbeck, 2019b). In the market model in appendix B-5, the total peak sales of the six
CGRP drugs was estimated at USD 8,387.4 mn in 2030. For perspective, Lundbeck (2019b) estimates a market
size including the two yet to be approved competitors of more than USD 7.0 bn by 2027. The market model in
this thesis estimates a size of USD 7,853.4 mn in the same year. Maladwala (2017) estimates a migraine market
size of USD 8.7 bn by 2026.

Importantly, as can be observed in appendices B-5 and B-6, Eptinezumab is assumed to show a faster ramp-
up than its three existing competitors. Bauer and Fischer (2000) show that (novel) pharmaceutical early-mover
products often face a longer growth phase as they first need to overcome physicians’ risk-aversity via extensive
promotion while late-mover products exhibit a quicker ramp-up (cf. also figure 2 in Bauer & Fischer, 2000).
However, an early-mover such as Aimovig also commands a “clear market share bonus” (Bauer & Fischer,
2000, p. 714). The six derived market shares at the time of first patent expiry / peak sales of Aimovig in 2029
are basically in line with the findings of Kalyanaram (2008) who investigated the order of entry penalty within
prescription pharmaceuticals (cf. table III of Kalyanaram, 2008).

5.2.1.2 Cost of sales

Following the derivation of Eptinezumab’s sales curve, the remaining income statement items need to be
forecasted. As Alder is a pre-revenue biotech, expense forecasting in particular becomes more difficult than
with standard firms. The widely used percentage of sales approaches using a peer-group often become invalid
as the peer-group might not have sales either. In the case of Alder, its competing products are all marketed by
far larger conglomerate players such as Amgen or Eli Lilly & Co., for example. This renders any comparison
to them difficult as well as no expenses for individual drugs such as Aimovig are broken out.

For cost of sales, one has to consider whether or not Alder would manufacture this biologic drug itself or
outsource it to a CMO as touched upon in 5.1.3.3. On a standalone basis, Alder would most likely outsource
the procurement. If Alder would partner with a larger firm (or in example sell itself to Lundbeck), one would
need to consider their manufacturing capabilities as well. Manufacturing of biologics is more complicated and
expensive than small molecules while at the same time a much higher price is charged for each unit, thus
making standard cost of sales as a percentage of sales approximations of limited use (Gooch et al., 2017). On
its post-announcement call, Lundbeck (2019b) reveals that Alder is the first major step towards biologics for

37
them apart from some early-stage projects. In addition, they state their intent to also outsource Eptinezumab’s
procurement.

Regardless of Lundbeck’s intention, this was also assumed for Alder on a standalone basis. Basu, Suresh,
Joglekar, Rai and Vernon (2008) find empirical evidence for average biotech cost of goods sold of 14% for
large cap biotechs including Amgen (cf. figure 1 in Basu et al., 2008). In a more recent analysis, Gooch et al.
(2017) investigate cost of goods sold on a business unit level. They estimate cost of sales at around 10 to 15%
of sales for biologics. It was decided to use 18.9% of sales for Alder’s products, which is the average between
15% in Gooch et al. on a business unit level and 22.7% of the large cap peer group. For simplicity, 2019 cost
of sales for Eptinezumab was assumed as zero. 18.9% is in the middle of the large cap peer group in appendix
B-7. However, there exists clear variation between firms.

5.2.1.3 Selling, general & administrative (S,G&A) expenses

For a biotech’s operating costs excluding R&D, Basu et al. (2008) find an average percentage of sales of 21%
over 1994 to 2005. The estimate of this study was derived from large caps and is in line the benchmarking peer
group in appendix B-7.

In addition, one has to consider the implications of the launch costs and the marketing ramp-up for
Eptinezumab in the first years. Elevated costs are a certainty and need to be considered separately here.
Davitian, Fitzpatrick, Perkins, Breton, Denzler, and Giovanniello (2018) find that, on average, single product
companies spend between USD 125 to 160 mn in cumulated S,G&A expenses in the three years leading up to
a launch. In comparison, Bauer and Fischer (2000) estimate USD 225 mn in after-tax launch costs for a globally
sold product. The former estimate was chosen for 2019 as it represents single product firms.

Alder is expecting to launch Eptinezumab in 2020. For 2019, Alder’s CEO guided dedicated pre-launch
expenses of around USD 300 mn (Dunn, 2019). These are reduced by the S,G&A estimates from Davitian et
al. (2018) to avoid double-counting and accounted for as ‘Other operating expenses, net’. For 2020 and 2021,
an average of 2019 / 2020 and 2022 is used to smoothen out the high launch expenses.

Furthermore, based on Myers and Howe (1997), Kellogg and Charnes (2000) provide assumptions of 11.1%
of sales for general and administrative expenses as well as separate marketing expenses of 100% in the first
year, 50% in the second year, 25% in years 3 and 4, and 20% thereafter. As Kellogg and Charnes (2000)
attempt to value a smaller biotech in their paper, these estimates were used for Alder’s product launches.

In addition, due to an earlier lawsuit settlement regarding Eptinezumab’s patent, Alder has an obligation to
pay Teva milestone fees of USD 25 mn upon approval, USD 75 mn upon hitting USD 1.0 bn in sales and
another USD 75 mn upon hitting USD 2.0 bn in sales (Cairns, 2018). For illustrative simplicity, these were
also classified as ‘Other operating expenses, net’ and can also be understood as licensing fees.

38
5.2.1.4 R&D expenses

For 2019, R&D has been estimated as simply the average of the two preceding years because Alder is finishing
up one of its phase III trials for Eptinezumab and faces further expenses due to the BLA process.

With regards to 2020 and after, one can expect that Eptinezumab will not require significant R&D investments
once it has been approved and starts marketing. However, there will need to be post-approval monitoring of
adverse events as well as post-approval studies given the novelty of CGRP mAbs. Cowan (2018) points out
that while the safety profile of CGRPs has been favourable in a clinical trial setting, there are many examples
where only the wider circulation post-approval has revealed long-term side effects.

Basu et al. (2008) find an average R&D percentage of sales of 26% for large cap biotechs which usually always
have several development projects as well as marketed drugs at once. This is hardly applicable to Alder’s
Eptinezumab asset.

To the best of the author’s knowledge, there exist no explicit empirical assessments of post-approval R&D
costs as a percentage of sales. Hence, it was resorted to the following rough approximation: DiMasi, Hansen
and Grabowski (2016) find that capitalised post-approval R&D costs within their large cap dataset represent
10.9% of the capitalised total R&D expenses.15 Applying this relationship to Basu et al. (2008) would translate
into 2.8% of sales which seems reasonable for Eptinezumab.

For simplicity, no R&D beyond the terminal horizon is assumed.

5.2.1.5 Taxation & NOLs

Alder as a pre-revenue stage biotech has not yet turned a profit. As of Alder’s 2018 annual report, significant
NOLs of USD 914.7 mn and R&D tax credits of USD 18.3 mn have been accumulated over time and can be
used to offset taxes in the past and the future.16 The accumulated losses have been a result of the development
of both Eptinezumab and ALD1910. As Eptinezumab is far closer to revenue generation, it was assumed that
the entire accumulated NOL balance will be used to shield Eptinezumab’s pre-tax earnings as this will increase
NPV. Due to Alder’s consistent history of losses, it was assumed that all NOLs will be carried forward, if
possible.

Following Wessels, Goedhart and Koller (2015), it was decided to value Alder’s NOLs separately (cf. p. 410).
Alder is based in Bothell, Washington, which does not impose a state-level corporate income tax but a gross
receipts tax (Cammenga, 2020). As taxes are notoriously difficult to project, it was decided to use the marginal
rate as a conservative approach. For simplicity, it was decided to use solely the U.S. federal marginal corporate

15 DiMasi, Hansen & Grabowski (2003) capitalize their R&D costs to the date of approval in order to make them comparable throughout
time.
16 Note that Alder only publishes yearly NOL balances as per 31 st December in accordance with the tax year cycle.

39
income tax rate of 21% as per the U.S. Tax Cuts and Jobs Act (TCJA) of 2017. Alder’s Q2 2019 report states
that its pre-2017 NOLs will start expiring by 2024 (Alder 2019a). As a worst case assumption this was assumed
for the entire pre-2017 balance.

With regards to taxation from an acquiror’s point of view (i.e. Lundbeck), one has to take into account the
limiting factor of the Internal Revenue Code (IRS) Section 382, which caps yearly use of acquired NOLs to

𝑒𝑞𝑢𝑖𝑡𝑦 𝑝𝑢𝑟𝑐ℎ𝑎𝑠𝑒 𝑝𝑟𝑖𝑐𝑒 ∗ 𝑙𝑜𝑛𝑔 𝑡𝑒𝑟𝑚 𝐼𝑅𝑆 𝑡𝑎𝑥 𝑒𝑥𝑒𝑚𝑝𝑡 𝑟𝑎𝑡𝑒 𝑓𝑜𝑟 𝑜𝑤𝑛𝑒𝑟𝑠ℎ𝑖𝑝 𝑐ℎ𝑎𝑛𝑔𝑒𝑠

with a long-term IRS tax-exempt rate of 1.77% as of October 2019 (IRS, 2019).17 In addition, the TCJA
changed the deductibility of post-2017 NOLs to just 80%. In general, this work follows Baker Tilly’s
guidelines as articulated by Dillon & Hobbs (n.d.). Further details on the NOL valuation can be obtained in
figures 10 and 11.

5.2.1.6 Depreciation & amortisation (D&A)

To derive unlevered free cash flows, we require an estimate of D&A. Basu et al. (2008) find an average of
depreciation as a percentage of sales of 9% from 2000 to 2005 for biotech firms. This estimate excludes
amortisation however. The stable, large cap biopharma peer group in appendix B-7 exhibits a median 7.0%
over the last five years. In comparison, Lundbeck’s D&A as a percentage of sales stands at a median 8.6%
over the same period. It was decided that 7.0% is reasonable given the limited manufacturing footprint due to
outsourcing.

5.2.1.7 Net working capital (NWC)

To focus on current operating assets, working capital was defined as

𝐶𝑢𝑟𝑟𝑒𝑛𝑡 𝑎𝑠𝑠𝑒𝑡𝑠 𝑙𝑒𝑠𝑠 𝑐𝑎𝑠ℎ & 𝑐𝑎𝑠ℎ 𝑒𝑞𝑢𝑖𝑣𝑎𝑙𝑒𝑛𝑡𝑠 − 𝐶𝑢𝑟𝑟𝑒𝑛𝑡 𝑙𝑖𝑎𝑏𝑖𝑙𝑖𝑡𝑖𝑒𝑠 𝑙𝑒𝑠𝑠 𝑠ℎ𝑜𝑟𝑡 𝑡𝑒𝑟𝑚 𝑓𝑖𝑛𝑎𝑛𝑐𝑖𝑎𝑙 𝑑𝑒𝑏𝑡

and net working capital subsequently as the difference between the previous and the current year. As working
capital is notoriously difficult to forecast, it was opted to apply the five year median NWC / sales ratio of the
large cap biopharma peer group in appendix B-7. This corresponds to a percentage of sales of 2.9%.

5.2.1.8 Capital expenditures (Capex)

Capital expenditures for Alder’s soon to be marketed asset should be comparatively low as only limited further
investment is needed, especially given the assumption of outsourcing the biologics production. While some
part of the USD 300mn in launch preparation costs (cf. 5.2.1.3) should likely be capex, it is difficult to isolate

17Note that the use of the equity purchase price will result in a calculation circularity. IRS Section 382 mandates the use of the highest
rate out of the month of the acquisition and the two prior months. Rulings are published one month in advance.

40
that based on publicly available information. In general, one has to balance capex and D&A in a meaningful
manner to maintain internal consistency.

Using the large cap biopharma peer group’s 6.4% of sales in appendix B-7 fulfils this criteria. In particular, it
implies that D&A will eventually completely salvage the remaining property, plant and equipment (P,P&E) as
it exceeds capex. This is intuitive for a project that has a limited life-time due to patent expiry. A more detailed
approach would be a D&A, capex and P,P&E schedule. However, this is not feasible on a project level.

5.2.1.9 Weighted average cost of capital (WACC)

After deriving unlevered free cash flows, an appropriate cost of capital is required to arrive at an enterprise
valuation for the expected cash flows of Alder’s assets.

According to Wessels, Goedhart and Koller (2015), WACC has to take into account all capital sources of the
company. In Alder’s case, its latest Q2 2019 balance sheet reveals convertible senior notes as well as preferred
stock. Subsequently, the applied WACC formula from section 3.3 becomes

𝐸 𝑃 𝐷
𝑟𝑊𝐴𝐶𝐶 = ∗ 𝑟𝑒𝑞𝑢𝑖𝑡𝑦 + ∗ 𝑟𝑝𝑟𝑒𝑓𝑒𝑟𝑟𝑒𝑑 + ∗𝑟 ∗ (1 − 𝑀𝑇𝑅)
𝐸+𝑃+𝐷 𝐸+𝑃+𝐷 𝐸 + 𝑃 + 𝐷 𝑑𝑒𝑏𝑡

with P representing the preferred stock component. As the financing and the preferred stock was not publicly
listed, book values were used to approximate. In its 2018 10-K, Alder provides a preferred dividend yield of
5.0% as well as an effective interest rate on the debt component of the convertible of 10.6%.

The cost of equity was determined by the standard CAPM model presented in section 3.3. To reduce
complexity, it was decided to use a single 20-year U.S. treasury bond as a proxy for the risk-free rate. 20 years
represents a compromise between liquidity (i.e. price finding) and the maturity of Alder’s project cash flows.18

Following Wessels, Goedhart and Koller (2015), the market risk premium was determined by computing the
historical return on the market index less the risk-free rate, where it was decided to use the NASDAQ
Biotechnology Index (NBI) as ‘the market’. The reasoning here is three-fold: (1) Biotech is one of the most
specialised sectors of all and requires sophisticated knowledge for (rational) investment decisions, hence the
shareholder base of a biotech is usually somewhat different than for a household firm such as Apple, i.e. the
‘marginal investor’ of Alder will likely be someone focused on that field. (2) Appendix B-8 showcases
different beta estimates for different time intervals and horizons using both the NBI as well as the broader
diversified S&P500 index. We can observe how the returns on the S&P500 perform worse at explaining the
returns of Alder over time and over different intervals. (3) From an acquisition standpoint, Lundbeck can be

18Generally, the 10-year U.S. treasury bond is far more liquid than its 20-year and in particular 30-year equivalents. For a more detailed
analysis, individual cost of capitals for each year using zero-coupon treasury bonds can be computed, the effect of which is usually
seen as neglectable in times of near-zero rates (Wessels, Goedhart and Koller, 2015).

41
seen as a biopharma-only investor, i.e. its risk-return expectations will be oriented towards the NBI and not a
broader all-sector index. The geometric average yearly market return was computed over the maximum
possible period of 1994 to 2019 as suggested by Wessels, Goedhart and Koller (2015).

As was observed in 5.1.1, Alder has exhibited massive price swings since IPO, possibly due to volatility from
trial results regarding Eptinezumab. With biotech in particular, it is likely best practiced to choose a recent
window in order to depict the ‘current’ riskiness of the business as suggested by Benninga (2014, p. 333) for
event studies. Based on appendix B-8, a beta of 1.24 using weekly returns in the year prior to the 13th September
2019 was chosen. This mitigates noise from trading breaks which often plagues regressions using daily data
(Damodaran, n.d.). An 𝑅 2 of 39.2% informs us of the portion of Alder’s risk that can be attributed to the
variation in market returns as represented by the NBI.

Figure 8 below summarises the cost of capital derivation.

Figure 8 – Weighted Average Cost of Capital (WACC)

Risk-free rate (20-Year Treasury Bond) 2.2% Convertible senior notes, net (book value) 188.3
Cost of debt (effective rate) 10.6%
Market return (NBI, geometric) 12.4%
Equity risk premium 10.2% Equity value (market capitalisation) 841.4
Beta 1.24 Total assets 1,136.3
Cost of equity 14.8%
Marginal tax rate 21.0%
Convertible preferred stock (book value) 106.5
Cost of preferred stock 5.0% Weighted average cost of capital (WACC) 12.8%

Source: Raw market data from Bloomberg. Financial data from Alder 10-Q Quarterly Report (June 2018).

Above, we obtain a WACC of 12.8% for Alder. This is precisely in line with surveys and empirical assessments
of biotech- and pharma-specific discount rates in several papers such as DiMasi et al. (2016), Harrington
(2012), DiMasi and Grabowski (2007), Puran (2005) and Myers and Shyam-Sunder (1996).

When discounting, a mid-year adjustment is applied to more precisely reflect the actual receipt of cash flows.

5.2.2 Eptinezumab – Valuation Results

Finally, following the DCF model calibration, the operating model and the FCFF is obtained as exhibited
below in figure 9. Figure 10 shows the DCF valuation result of USD 1,368.7 mn on a risk-adjusted basis as
well as the standalone NOL tax shield valuation of USD 136.5 mn. Thus, Eptinezumab has been valued at
USD 1,505.2 mn on a risk-adjusted basis.

42
If we consider the acquisition case for the NOL balance, we need to account for IRS Section 382 which caps
the yearly deductibility of acquired NOLs. In this case, for Lundbeck (and disregarding any potential
synergies), the acquired NOL tax shield could be valued at USD 113.3 mn with a total valuation of USD
1,482.0 mn. Figure 11 showcases the acquisition case for the NOLs.

As the U.S. PDUFA approval probability is of integral importance to the valuation of Eptinezumab, a
sensitivity table can be found below in table 6.

Table 6 – Eptinezumab Approval vs. WACC Sensitivity Analysis

U.S. PDUFA Approval Probability

1,482.0 75.0% 80.0% 86.8% 90.0% 95.0%

8.8% 1,801.9 1,928.1 2,099.7 2,180.5 2,306.6
10.8% 1,506.9 1,613.4 1,758.3 1,826.4 1,932.9
WACC
12.8% 1,268.2 1,358.8 1,482.0 1,539.9 1,630.5
14.8% 1,073.3 1,150.9 1,256.4 1,306.1 1,383.7
16.8% 912.9 979.8 1,070.7 1,113.5 1,180.4

Source: Own calculations.

43
 Figure 9 – Operating Model & FCFF Calculation (Eptinezumab)
I. Income Statement

FYE - 31st Dec 2018A 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E
Revenue, net — — 50.0 189.8 382.0 588.9 773.0 905.9 1,015 1,099 1,160 1,196
% Y-o-Y growth 279.7% 101.2% 54.2% 31.3% 17.2% 12.0% 8.3% 5.5% 3.1%
% of peak sales 4.1% 15.7% 31.6% 48.8% 64.0% 75.0% 84.0% 91.0% 96.0% 99.0%
Cost of sales — — (9.4) (35.8) (72.1) (111.1) (145.9) (171.0) (191.5) (207.4) (218.8) (225.7)
% of sales 18.9% 18.9% 18.9% 18.9% 18.9% 18.9% 18.9% 18.9% 18.9% 18.9%
Gross profit — — 40.6 154.0 309.9 477.7 627.1 734.9 823.1 891.7 940.7 970.1

Marketing expenses — — (50.0) (94.9) (95.5) (147.2) (154.6) (181.2) (202.9) (219.8) (231.9) (239.1)
% of sales 100.0% 50.0% 25.0% 25.0% 20.0% 20.0% 20.0% 20.0% 20.0% 20.0%
Research & development costs (239.1) (246.0) (1.4) (5.4) (10.8) (16.6) (21.8) (25.6) (28.7) (31.1) (32.8) (33.8)
% of sales 2.8% 2.8% 2.8% 2.8% 2.8% 2.8% 2.8% 2.8% 2.8% 2.8%
General & administrative expenses (47.5) (90.0) (66.0) (54.0) (42.0) (64.8) (85.0) (99.6) (111.6) (120.9) (127.5) (131.5)
% of sales 132.0% 28.5% 11.0% 11.0% 11.0% 11.0% 11.0% 11.0% 11.0% 11.0%
Other operating expenses, net (210.0) (25.0) — — — — — (75.0) — — —
Income (loss) from operations / EBIT (286.6) (546.0) (102) (0.3) 161.6 249.1 365.6 428.5 404.9 519.9 548.4 565.6
% margin (203.7%) (0.2%) 42.3% 42.3% 47.3% 47.3% 39.9% 47.3% 47.3% 47.3%

II. Unlevered Free Cash Flow

FYE - 31st Dec 2018A 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E
EBIT (286.6) (546.0) (101.9) (0.3) 161.6 249.1 365.6 428.5 404.9 519.9 548.4 565.6
Marginal Corporate Tax on EBIT — — — — (33.9) (52.3) (76.8) (90.0) (85.0) (109.2) (115.2) (118.8)
NOPAT (286.6) (546.0) (101.9) (0.3) 127.7 196.8 288.8 338.5 319.9 410.7 433.3 446.8
add back: D&A 2.6 2.8 3.5 13.3 26.9 41.4 54.3 63.7 71.3 77.3 81.5 84.1
less: Increase (decrease) in NWC (3.4) (2.4) (1.4) (4.0) (5.6) (6.0) (5.3) (3.8) (3.1) (2.4) (1.7) (1.0)
less: Capital expenditure (3.1) (3.2) (3.2) (12.1) (24.3) (37.4) (49.1) (57.6) (64.5) (69.9) (73.7) (76.0)
Unlevered Free Cash Flow (290.6) (548.8) (103.0) (3.1) 124.7 194.8 288.7 340.7 323.6 415.7 439.3 453.8

Transition probabilities 86.8% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
Cash flow probabilities 100.0% 86.8% 86.8% 86.8% 86.8% 86.8% 86.8% 86.8% 86.8% 86.8% 86.8%

I. Income Statement

FYE - 31st Dec 2030E 2031E 2032E 2033E 2034E 2035E 2036E 2037E 2038E 2039E 2040E 2041E
Revenue, net 1,208 1,208 1,196 1,172 1,147 869.6 640.1 446.9 289.9 157.0 72.5 12.1
% Y-o-Y growth 1.0% — (1.0%) (2.0%) (2.1%) (24.2%) (26.4%) (30.2%) (35.1%) (45.8%) (53.8%) (83.3%)
% of peak sales 100.0% 100.0% 99.0% 97.0% 95.0% 72.0% 53.0% 37.0% 24.0% 13.0% 6.0% 1.0%
Cost of sales (228.0) (228.0) (225.7) (221.1) (216.6) (164.1) (120.8) (84.3) (54.7) (29.6) (13.7) (2.3)
% of sales 18.9% 18.9% 18.9% 18.9% 18.9% 18.9% 18.9% 18.9% 18.9% 18.9% 18.9% 18.9%
Gross profit 979.9 979.9 970.1 950.5 930.9 705.5 519.3 362.5 235.2 127.4 58.8 9.8

Marketing expenses (241.6) (241.6) (239.1) (234.3) (229.5) (173.9) (128.0) (89.4) (58.0) (31.4) (14.5) (2.4)
% of sales 20.0% 20.0% 20.0% 20.0% 20.0% 20.0% 20.0% 20.0% 20.0% 20.0% 20.0% 20.0%
Research & development costs (34.1) (34.1) (33.8) (33.1) (32.4) (24.6) (18.1) (12.6) (8.2) (4.4) (2.0) (0.3)
% of sales 2.8% 2.8% 2.8% 2.8% 2.8% 2.8% 2.8% 2.8% 2.8% 2.8% 2.8% 2.8%
General & administrative expenses (132.9) (132.9) (131.5) (128.9) (126.2) (95.7) (70.4) (49.2) (31.9) (17.3) (8.0) (1.3)
% of sales 11.0% 11.0% 11.0% 11.0% 11.0% 11.0% 11.0% 11.0% 11.0% 11.0% 11.0% 11.0%
Other operating expenses, net — — — — — — — — — — — —
Income (loss) from operations / EBIT 571.3 571.3 565.6 554.2 542.7 411.3 302.8 211.4 137.1 74.3 34.3 5.7
% margin 47.3% 47.3% 47.3% 47.3% 47.3% 47.3% 47.3% 47.3% 47.3% 47.3% 47.3% 47.3%

II. Unlevered Free Cash Flow

FYE - 31st Dec 2030E 2031E 2032E 2033E 2034E 2035E 2036E 2037E 2038E 2039E 2040E 2041E
EBIT 571.3 571.3 565.6 554.2 542.7 411.3 302.8 211.4 137.1 74.3 34.3 5.7
Marginal Corporate Tax on EBIT (120.0) (120.0) (118.8) (116.4) (114.0) (86.4) (63.6) (44.4) (28.8) (15.6) (7.2) (1.2)
NOPAT 451.3 451.3 446.8 437.8 428.8 325.0 239.2 167.0 108.3 58.7 27.1 4.5
add back: D&A 84.9 84.9 84.1 82.4 80.7 61.1 45.0 31.4 20.4 11.0 5.1 0.8
less: Increase (decrease) in NWC (0.3) — 0.3 0.7 0.7 8.0 6.6 5.6 4.5 3.8 2.4 1.7
less: Capital expenditure (76.8) (76.8) (76.0) (74.5) (72.9) (55.3) (40.7) (28.4) (18.4) (10.0) (4.6) (0.8)
Unlevered Free Cash Flow 459.1 459.5 455.2 446.4 437.2 338.9 250.2 175.6 114.8 63.6 30.0 6.3

Transition probabilities 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
Cash flow probabilities 86.8% 86.8% 86.8% 86.8% 86.8% 86.8% 86.8% 86.8% 86.8% 86.8% 86.8% 86.8%

Source: Own calculations and assumptions as explained in 5.2.1.

44
 Figure 10 – DCF & NOL Valuation (Eptinezumab - Standalone)
III. Discounted Cash Flow Analysis

FYE - 31st Dec 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E
Unlevered Free Cash Flow (548.8) (103.0) (3.1) 124.7 194.8 288.7 340.7 323.6 415.7 439.3 453.8
Period discounted (EOY) 0.17 1.17 2.17 3.17 4.17 5.17 6.17 7.17 8.17 9.17 10.17
Participation in yearly cash flow 16.7% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
WACC 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8%
Discount factor (mid-year adjusted) 99.0% 92.3% 81.8% 72.4% 64.2% 56.9% 50.4% 44.7% 39.6% 35.1% 31.1%
Discounted uFCF (90.8) (95.0) (2.5) 90.3 125.0 164.3 171.8 144.6 164.6 154.1 141.1
Enterprise Value (if approved) 1,590.7

FYE - 31st Dec 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E
Unlevered Free Cash Flow (risk-adj.) (548.8) (89.4) (2.7) 108.2 169.1 250.6 295.8 280.8 360.8 381.3 393.9
Discounted uFCF (risk-adj.) (90.8) (82.5) (2.2) 78.4 108.5 142.6 149.1 125.5 142.9 133.8 122.5
Enterprise Value (risk-adj.) 1,368.7

IV. Net Operating Losses

Case (S = Standalone, A = Acquisition) S

FYE - 31st Dec 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E
Taxable income (553.2) (109.0) (7.5) 154.4 241.9 358.4 421.3 397.7 512.7 541.3 558.4
NOL used — — — (154.4) (241.9) (341.1) (337.0) (263.4) — — —
Taxable income (shielded) (553.2) (109.0) (7.5) — — 17.3 84.3 134.3 512.7 541.3 558.4
Tax rate 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0%
NOL tax shield — — — 32.4 50.8 71.6 70.8 55.3 — — —
Cash taxes due — — — — — 3.6 17.7 28.2 107.7 113.7 117.3
Discounted NOL tax shield — — — 23.5 32.6 40.8 35.7 24.7 — — —
PV(NOL tax shield) 157.3

Discounted NOL tax shield (risk-adj.) — — — 20.4 28.3 35.4 31.0 21.5 — — —
PV(NOL tax shield, risk-adj.) 136.5

III. Discounted Cash Flow Analysis

FYE - 31st Dec 2030E 2031E 2032E 2033E 2034E 2035E 2036E 2037E 2038E 2039E 2040E 2041E
Unlevered Free Cash Flow 459.1 459.5 455.2 446.4 437.2 338.9 250.2 175.6 114.8 63.6 30.0 6.3
Period discounted (EOY) 11.17 12.17 13.17 14.17 15.17 16.17 17.17 18.17 19.17 20.17 21.17 22.17
Participation in yearly cash flow 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
WACC 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8%
Discount factor (mid-year adjusted) 27.6% 24.4% 21.6% 19.2% 17.0% 15.1% 13.3% 11.8% 10.5% 9.3% 8.2% 7.3%
Discounted uFCF 126.5 112.2 98.5 85.6 74.3 51.0 33.4 20.8 12.0 5.9 2.5 0.5

FYE - 31st Dec 2030E 2031E 2032E 2033E 2034E 2035E 2036E 2037E 2038E 2039E 2040E 2041E
Unlevered Free Cash Flow (risk-adj.) 398.5 398.8 395.1 387.5 379.5 294.1 217.1 152.4 99.7 55.2 26.1 5.5
Discounted uFCF (risk-adj.) 109.8 97.4 85.5 74.3 64.5 44.3 29.0 18.0 10.4 5.1 2.1 0.4

IV. Net Operating Losses

FYE - 31st Dec 2030E 2031E 2032E 2033E 2034E 2035E 2036E 2037E 2038E 2039E 2040E 2041E
Taxable income 564.1 564.1 558.4 547.0 535.5 404.1 295.6 204.2 129.9 67.1 27.1 (1.5)
NOL used — — — — — — — — — — — —
Taxable income (shielded) 564.1 564.1 558.4 547.0 535.5 404.1 295.6 204.2 129.9 67.1 27.1 (1.5)
Tax rate 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0%
NOL tax shield — — — — — — — — — — — —
Cash taxes due 118.5 118.5 117.3 114.9 112.5 84.9 62.1 42.9 27.3 14.1 5.7 —
Discounted NOL tax shield — — — — — — — — — — — —

Discounted NOL tax shield (risk-adj.) — — — — — — — — — — — —

Source: Own calculations and assumptions as explained in 5.2.1.

45
Figure 11 – NOL Valuation (Eptinezumab - Acquisition Case)

Note: Assumes equity purchase price equal to fully diluted equity value in the acquisition case (cf. 5.3.3).
IV. Net Operating Losses
Case (S = Standalone, A = Acquisition) A
FYE - 31st Dec 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E 2030E 2031E 2032E 2033E 2034E 2035E 2036E 2037E 2038E 2039E 2040E 2041E
Taxable income (546.0) (101.9) (0.3) 161.6 249.1 365.6 428.5 404.9 519.9 548.4 565.6 571.3 571.3 565.6 554.2 542.7 411.3 302.8 211.4 137.1 74.3 34.3 5.7
NOL used — — — (161.6) (249.1) (365.6) (133.5) (37.4) (37.4) (37.4) (37.4) (37.4) (10.1) — — — — — — — — — —
Taxable income (shielded) (546.0) (101.9) (0.3) — — — 294.9 367.5 482.5 511.1 528.2 533.9 561.2 565.6 554.2 542.7 411.3 302.8 211.4 137.1 74.3 34.3 5.7
Tax rate 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0%
NOL tax shield — — — 33.9 52.3 76.8 28.0 7.9 7.9 7.9 7.9 7.9 2.1 — — — — — — — — — —
Cash taxes due — — — — — — 61.9 77.2 101.3 107.3 110.9 112.1 117.8 118.8 116.4 114.0 86.4 63.6 44.4 28.8 15.6 7.2 1.2
Discounted NOL tax shield — — — 24.6 33.6 43.7 14.1 3.5 3.1 2.8 2.4 2.2 0.5 — — — — — — — — — —
PV(NOL tax shield) 130.5
Discounted NOL tax shield (risk-adj.) — — — 21.3 29.2 37.9 12.3 3.0 2.7 2.4 2.1 1.9 0.5 — — — — — — — — — —
PV(NOL tax shield, risk-adj.) 113.3
Acquisition Case
Equity purchase price 2,112 2112.323
IRS long-term tax-exempt rate 1.77%

46
Yearly cap on NOL use 37.4

Source: Own calculations and assumptions as explained in 5.2.1.

Historical NOL (acquired)
NOL balance (BOY) 914.7 914.7 914.7 914.7 765.2 271.8 234.4 197.0 159.7 122.3 84.9 47.5 10.1 — — — — — — — — — —
NOL used — — — (149.5) (37.4) (37.4) (37.4) (37.4) (37.4) (37.4) (37.4) (37.4) (10.1) — — — — — — — — — —
of which: pre-2018 — — — (149.5) (37.4) — — — — — — — — — — — — — — — — — —
of which: post-2017 @ 80% — — — — — (37.4) (37.4) (37.4) (37.4) (37.4) (37.4) (37.4) (10.1) — — — — — — — — — —
NOL expired (pre-2018 tranche) — — — — (456.0) — — — — — — — — — — — — — — — — — —
NOL balance (EOY) 914.7 914.7 914.7 914.7 765.2 271.8 234.4 197.0 159.7 122.3 84.9 47.5 10.1 — — — — — — — — — — —
of which: pre-2018 642.9 642.9 642.9 642.9 493.4 — — — — — — — — — — — — — — — — — — —
remaining life-time (worst case) 5.0 4.0 3.0 2.0 1.0 — — — — — — — — — — — — — — — — — — —
of which: post-2017 271.8 271.8 271.8 271.8 271.8 271.8 234.4 197.0 159.7 122.3 84.9 47.5 10.1 — — — — — — — — — — —
Unused yearly cap balance (BOY) — 37.4 74.8 112.1 — — — — — — — — — — — — — — — — — — —
Carryforward cap created 37.4 37.4 37.4 — — — — — — — — — — — — — — — — — — — —
Carryforward cap used — — — (112.1) — — — — — — — — — — — — — — — — — — —
Unused yearly cap balance (EOY) — 37.4 74.8 112.1 — — — — — — — — — — — — — — — — — — — —
New NOL (ongoing)
NOL balance (BOY) — 546.0 647.9 648.2 636.1 424.4 96.1 — — — — — — — — — — — — — — — —
NOL added 546.0 101.9 0.3 — — — — — — — — — — — — — — — — — — — —
NOL used — — — (12.1) (211.7) (328.2) (96.1) — — — — — — — — — — — — — — — —
NOL expired — — — — — — — — — — — — — — — — — — — — — — —
NOL balance (EOY) — 546.0 647.9 648.2 636.1 424.4 96.1 — — — — — — — — — — — — — — — — —
5.2.3 ALD1910 – Static DCF Assumptions & Model Construction

ALD1910 is an investigational molecule that started its phase I in the second half of 2019 (Hoffman, 2019).
So far, only very limited information on ALD1910 is available. In particular, there is considerable uncertainty
if the molecule works, and if so, how large the addressable market size really is.

5.2.3.1 Project schedule

At start, a project schedule is built using standard development duration estimates for the different phases from
Bogdan and Villiger (2010, p. 76). The patented period was assumed to be 12 years and the sales cycle will
follow a standard average sales curve as provided by Bogdan & Villiger (2010). Cash flows are modelled out
until 2049 with peak sales assumed by 2038. An exhibit of the schedule can be found in appendix C-1.

5.2.3.2 Peak sales estimation

Following Bogdan & Villiger (2010), a peak sales estimate as of today is obtained using a basic epidemiology
model for the U.S. and Europe. Modelling early-stage investigational candidates requires extensive knowledge
and resources. To simplify our analysis, we only regard the U.S. and the European Union as the two main
developed markets for migraine mAbs. Furthermore, we assume identical epidemiological assumptions
regarding both regions as well as a constant future USD/EUR exchange rate of 1.1074.19

As of September 2019, there have not been any released test results from either ALD1910 or Amgen’s
candidate AMG-301 (Do, Guo & Ashina, 2019).20 Alder’s CEO points out that ALD1910 holds promise for
treating non-responders to Eptinezumab (Alder, 2018b). Hence, it was decided that the best possible
approximation are the current CGRPs. As PACAP is a similar agent compared to CGRP (cf. appendix B-2),
the epidemiological model was based on ICER’s assessment of the current CGRP drugs (2018) as well as their
phase III studies. We assume an application of ALD1910 in both episodic and chronic migraine cases of which
44.0% qualify for preventive therapy and of which 19.7% are eligible for CGRP prescription. Subsequently,
ICER (2018) estimates 1.7% of the entire U.S. population as eligible for CGRP application. This is in line with
Amgen’s estimate of 2.0% according to ICER (2018).

We assume and extrapolate population size and growth as per U.S. Census and EUROSTAT until 2049 (U.S.
Census Bureau, 2017; Eurostat, 2020).21 Using ICER’s estimate from above, we obtain the expected eligible
number of patients in each year. Do, Guo and Ashina (2019) and Alder (2018b) describe PACAP as an
alternative treatment path to CGRP. Given that the CGRPs will have established themselves with doctors by

19 Note that we disregard fluctuations in the Danish krone (which is pegged anyway) and the Swedish krone after considering limited
gains in accuracy vs. the parsimony principle. Foreign exchange data from https://fred.stlouisfed.org/series/DEXUSEU.
20 Note that Amgen has delayed the release of results on AMG-301 which were initially expected in mid-2019.
21 Eurostat only releases 5-year interval forecasts. Thus we apply the same approach as for Eptinezumab’s sales, i.e. a polynomial (of

2nd order).

47
the time of ALD1910 market launch, it is reasonable to assume that PACAP inhibitors will mainly be
prescribed when a patient does not respond (well) to the CGRP therapies. Hence, we first derive a continuous
annual CGRP patient base before applying their non-responder rate to obtain a PACAP patient base
approximation. The CGRP patient base is estimated via the following logical steps:

(1) Out of the eligible patient base, a patient first needs to be diagnosed. This is influenced by various
factors including which type of physician they see, the general access to healthcare, etc. Maladwala (2017)
estimates diagnosis at roughly 40 to 50% for episodic and chronic migraine. As a worst-case assumption, 40%
was chosen.
(2) A patient then needs to actually start a prescription proposed by the doctor. For traditional preventive
therapy such as Botox (which requires two dozen injections at once), this is quite low (Mavridis, Koniari,
Fakas & Mitsikostas, 2019). For CGRPs with their monthly or quarterly cycle, a recent survey of doctors and
patients by Cowan, Cohen, Rosenman and Iyer (2019) implies a prescription start rate of 65.3%. This is
supported by Gottschalk (2019) for Eptinezumab in particular due to the infusion showing almost instant
results.

(3) Furthermore, a patient then needs to show adherence / compliance to the prescribed therapy, i.e. they
need to continue taking their drugs for at least a year to see if they work properly. Previously, compliance has
been very low due to various factors (Hepp, Bloudek & Varon, 2014; Hepp et al., 2015; American Headache
Society, 2018; Diken, Kosak, Aydinlar & Kocaman, 2018). The study by Cowan, Cohen, Rosenman and Iyer
(2019) implies an adherence rate of 66.2% for the CGRP dosing schedules. A higher compliance is supported
by Scuteri et al. (2019) and Mavridis, Koniari, Fakas and Mitsikostas (2019).

(4) Following application as intended, patients then either do respond or do not.22 Do, Guo and Ashina
(2019) summarise how this responder rate ranges between 48 to 62% for the four current CGRPs. As a worst-
case assumption, 62% was chosen.

Subsequently, the PACAP patient base was estimated by assuming that all CGRP non-responders are then
prescribed with PACAP. Here, the only defensible estimate of the responder rate can be the CGRP equivalent
as no clinical trial data is publicly available as of September 2019. A worst-case scenario of the lower end of
the CGRPs, i.e. 48% was selected.

Next, the pricing of a PACAP inhibitor is needed. The only defensible assumption can be that it will – on
average – be similarly (i.e. identically) priced to the current CGRP products (cf. table 5). The average (initial)
monthly wholesale list prices of USD 575 in the U.S. and EUR 494 in the Euro countries were extrapolated by

22The response rate is usually defined as the percentage of people with more than a 50% reduction in migraine frequency (American
Headache Society, 2018; Bendtsen et al., 2018).

48
the average (expected) inflation in both regions over 2010 to 2019 (IMF, 2020).23 Finally, an U.S. industry
average gross-to-net discount of 27% as per ICER (2018) was applied to arrive at net prices. An identical
gross-to-net discount is assumed in Europe. This is a rather fragile assumption due to the very different
reimbursement structures in Europe compared to the U.S. but has been made as these discounts are usually
confidential and differ among EU countries (Vogler & Martikainen, 2014). While apparently no studies on an
average EU discount exist, overall, one can expect it to be lower than in the U.S. where wholesale list prices
are far less regulated. However, as a worst-case assumption, 27% has been assumed as it was the only available
reference point.

Finally, the standard sales curve from appendix B-3 is applied and a total peak (market) revenue estimate for
the U.S. and the EU computed. As Amgen’s AMG-301 is ahead in development, a peak market share of 42%
as estimated by Kalyanaram (2008) for second-mover prescription drugs was assumed and a peak revenue
estimate for ALD1910 of USD 1,202.1 mn is obtained. An overview of the epidemiology model can be found
in appendix C-2. Table 7 below provides sensitivity analyses of the inputs for the epidemiology model with
regards to their impact on the total peak sales estimate.

Table 7 – Sensitivity Analyses on the Peak Sales for ALD1910

Eligible % of Populationa Adherence Rate

1,202.1 1.2% 1.5% 1.7% 2.0% 2.2% 1,202.1 56.2% 61.2% 66.2% 71.2% 76.2%
30.0% 639 770 902 1,033 1,165 55.3% 864 941 1,018 1,095 1,172
Diagnosis 35.0% 745 898 1,052 1,205 1,359 Prescription 60.3% 942 1,026 1,110 1,194 1,278
Rate 40.0% 851 1,027 1,202 1,377 1,553 Rate 65.3% 1,021 1,111 1,202 1,293 1,384
45.0% 958 1,155 1,352 1,550 1,747 70.3% 1,099 1,196 1,294 1,392 1,490
50.0% 1,064 1,284 1,503 1,722 1,941 75.3% 1,177 1,281 1,386 1,491 1,595

PACAP Responder Rate Peak Market Share

1,202.1 37.7% 42.7% 47.7% 52.7% 57.7% 1,202.1 32.0% 37.0% 42.0% 47.0% 52.0%
CGRP 28.0% 832 942 1,053 1,163 1,273 37.0% 790 914 1,037 1,161 1,284
Gross-to-
Non- 33.0% 891 1,009 1,127 1,246 1,364 32.0% 853 986 1,120 1,253 1,386
Net
Responder 38.0% 950 1,076 1,202 1,328 1,454 27.0% 916 1,059 1,202 1,345 1,488
43.0% 1,009 1,143 1,277 1,411 1,545
Discount 22.0% 979 1,132 1,284 1,437 1,590
Rate
48.0% 1,068 1,210 1,352 1,493 1,635 17.0% 1,041 1,204 1,367 1,530 1,692

Source: Own calculations.

a For CGRPs according to ICER (2018).

Above, one can infer a comparatively strong sensitivity to the ICER estimate of the eligible percentage of the
population that is eligible for CGRPs and the diagnosis rate as well as the peak market share and the gross-to-
net discount. Overall, the values of the four quadrants range between USD 639 mn and USD 1,941 mn.

23 The average European list price was assumed to be equivalent to the German list price as of May 2020.

49
Importantly, one can also compare the obtained result to the median peak sales estimate of the six CGRPs
modelled earlier, which serves as another estimation approach and cross-check for a new investigational
compound (Bogdan & Villiger, 2010). The median peak sales estimate of the Eptinezumab peer-group is USD
1,290.2 mn, and hence the estimate for ALD1910 is in line with what can reasonably be expected given the
very limited information available.

5.2.3.3 Operating assumptions (Development period)

For the development phase from 2019 to 2027, we require dedicated absolute R&D costs, general &
administrative expenses as well as D&A, net working capital and capital expenditure numbers.

For R&D costs, the out-of-pocket cost estimates of DiMasi and Grabowski (2007) on biotechnology projects
were deemed most relevant as they stem from a sub-sample of mainly mAbs and recombinant protein
candidates (cf. table 3). We adjust for the time difference using the GDP deflator from 2005 to 2019 as
suggested by DiMasi and Grabowski (2007). The costs per phase are then distributed in equal increments over
the relevant years.

For general & administrative (and marketing) we first apply the pre-launch estimate of Davitian et al. (2018).
In addition, in 2027, we assume identical ‘Other operating expenses, net’ of USD 300 mn as guided by Alder’s
CEO for Eptinezumab’s launch but adjust for the USD 90 mn entailed in Davitian et al. (2018). Prior to that
we simply assume identical general & administrative costs to Alder’s financial history over 2011 to 2018 when
it was developing mainly Eptinezumab. Given the similarity of the product, this should comprise a reasonable
approximation. Any other ‘best guess’ was deemed insufficient to be defensible.

In order to project D&A expenses, we resort to the same approach and select a median D&A expense of USD
1.3 mn based on Alder’s financial history over 2011 to 2018. Identical approximation are used for the absolute
change in net working capital as well as capital expenditures.

5.2.3.4 Operating assumptions (Marketed period)

For ALD1910, only very limited public information exists. Hence, one can only resort to the most defensible
assumptions. In this case, the marketed period from 2028 to 2049 is forecasted using identical assumptions to
Eptinezumab. Any other assumptions would hardly be defensible given the very similar underlying value
drivers of both Eptinezumab and ALD1910 as well as the limited information on ALD1910.

5.2.3.5 Taxation & NOLs

The taxation assumptions are identical to 5.2.1.5. Regarding the historic NOL balance, it was assumed that the
tax shield is credited against Eptinezumab’s sales. Hence, none will be available for ALD1910. However,

50
(new) NOLs incurred from the development of ALD1910 will in fact be utilised to shield ALD1910’s future
pre-tax earnings.

5.2.3.6 Weighted average cost of capital (WACC)

The same cost of capital as derived in 5.2.1.9 is used for discounting expected FCFF from Alder’s ALD1910.
This assumption of a constant discount rate throughout all development phases remains debatable. From a
theoretical standpoint, scholars have pointed out that the cost of capital for R&D investments actually declines
with each phase by means of a “risk-return staircase” (Myers & Shyam-Sunder, 1996; Myers & Howe, 1997).
However, estimating such as stage-dependent cost of capital would require significantly more time and
resources. Puran (2005) finds that only a few practitioners actually use stage-dependent discount rates.

51
5.2.4 ALD1910 – Static Valuation Results

First, we compute the static DCF value of ALD1910. An overview of the operating model and FCFF derivation
for ALD1910 can be obtained below in figure 12. The static DCF valuation is exhibited in figure 13 below.
Disregarding managerial flexibility, ALD1910 can be valued at a negative USD (47.9) mn, thus suggesting
abandonment of the project. The NOL tax shield present value is estimated at USD 5.0 mn, hence leaving the
total value of ALD1910 negative.

Figure 12 – Operating Model & FCFF Calculation (ALD1910)

C.I. Income Statement

FYE - 31st Dec 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E 2030E 2031E 2032E 2033E 2034E

Revenue, net — — — — — — — — — 60.1 228.4 432.8 613.1 769.4 901.6 1,009.8

% Y-o-Y growth 280.0% 89.5% 41.7% 25.5% 17.2% 12.0%
% of peak sales
Cost of sales — — — — — — — — — (11.3) (43.1) (81.7) (115.7) (145.2) (170.2) (190.6)
% of sales 18.9% 18.9% 18.9% 18.9% 18.9% 18.9% 18.9%
Gross profit — — — — — — — — — 48.8 185.3 351.1 497.4 624.2 731.4 819.2

Marketing expenses — — — — — — — — — (60.1) (114.2) (108.2) (153.3) (153.9) (180.3) (202.0)

% of sales 100.0% 50.0% 25.0% 25.0% 20.0% 20.0% 20.0%
Research & development costs (20.7) (20.7) (16.1) (16.1) (16.1) (41.1) (41.1) (41.1) (4.9) (1.7) (6.5) (12.2) (17.3) (21.7) (25.5) (28.5)
% of sales 2.8% 2.8% 2.8% 2.8% 2.8% 2.8% 2.8%
General & administrative expenses (6.6) (7.2) (7.7) (12.5) (16.7) (26.1) (38.1) (47.5) (90.0) (79.4) (65.0) (47.6) (67.4) (84.6) (99.2) (111.1)
% of sales 132.0% 28.5% 11.0% 11.0% 11.0% 11.0% 11.0%
Other operating expenses, net — — — — — — — — (210.0) — — — — — — —
Income (loss) from operations / EBIT (27.4) (28.0) (23.8) (28.6) (32.9) (67.3) (79.3) (88.6) (304.9) (92.4) (0.3) 183.1 259.3 363.9 426.5 477.6
% margin (153.7%) (0.2%) 42.3% 42.3% 47.3% 47.3% 47.3%

C.II. Unlevered Free Cash Flow

FYE - 31st Dec 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E 2030E 2031E 2032E 2033E 2034E
EBIT (27.4) (28.0) (23.8) (28.6) (32.9) (67.3) (79.3) (88.6) (304.9) (92.4) (0.3) 183.1 259.3 363.9 426.5 477.6
Marginal Corporate Tax on EBIT — — — — — — — — — — — (38.4) (54.5) (76.4) (89.6) (100.3)
NOPAT (27.4) (28.0) (23.8) (28.6) (32.9) (67.3) (79.3) (88.6) (304.9) (92.4) (0.3) 144.6 204.9 287.5 336.9 377.3
add back: D&A 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 1.3 4.2 16.1 30.4 43.1 54.1 63.4 71.0
less: Increase (decrease) in NWC (4.8) (4.8) (4.8) (4.8) (4.8) (4.8) (4.8) (4.8) (4.8) (4.8) (4.9) (5.9) (5.2) (4.5) (3.8) (3.1)
less: Capital expenditure (1.6) (1.6) (1.6) (1.6) (1.6) (1.6) (1.6) (1.6) (1.6) (3.8) (14.5) (27.5) (39.0) (48.9) (57.3) (64.2)
Unlevered Free Cash Flow (32.5) (33.1) (28.9) (33.7) (38.0) (72.4) (84.4) (93.8) (310.0) (96.8) (3.7) 141.6 203.8 288.1 339.1 381.0

Transition probabilities 100.0% 70.1% 100.0% 100.0% 38.1% 100.0% 100.0% 60.7% 86.8% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
Cash flow probabilities 100.0% 100.0% 70.1% 70.1% 70.1% 26.7% 26.7% 26.7% 16.2% 14.1% 14.1% 14.1% 14.1% 14.1% 14.1% 14.1%

C.I. Income Statement

FYE - 31st Dec 2035E 2036E 2037E 2038E 2039E 2040E 2041E 2042E 2043E 2044E 2045E 2046E 2047E 2048E 2049E

Revenue, net 1,093.9 1,154.0 1,190.1 1,202.1 1,202.1 1,190.1 1,166.1 1,142.0 865.5 637.1 444.8 288.5 156.3 72.1 12.0
% Y-o-Y growth 8.3% 5.5% 3.1% 1.0% — (1.0%) (2.0%) (2.1%) (24.2%) (26.4%) (30.2%) (35.1%) (45.8%) (53.8%) (83.3%)
% of peak sales
Cost of sales (206.5) (217.8) (224.6) (226.9) (226.9) (224.6) (220.1) (215.5) (163.4) (120.3) (83.9) (54.5) (29.5) (13.6) (2.3)
% of sales 18.9% 18.9% 18.9% 18.9% 18.9% 18.9% 18.9% 18.9% 18.9% 18.9% 18.9% 18.9% 18.9% 18.9% 18.9%
Gross profit 887.5 936.2 965.5 975.2 975.2 965.5 946.0 926.5 702.2 516.9 360.8 234.1 126.8 58.5 9.8

Marketing expenses (218.8) (230.8) (238.0) (240.4) (240.4) (238.0) (233.2) (228.4) (173.1) (127.4) (89.0) (57.7) (31.3) (14.4) (2.4)
% of sales 20.0% 20.0% 20.0% 20.0% 20.0% 20.0% 20.0% 20.0% 20.0% 20.0% 20.0% 20.0% 20.0% 20.0% 20.0%
Research & development costs (30.9) (32.6) (33.6) (34.0) (34.0) (33.6) (33.0) (32.3) (24.5) (18.0) (12.6) (8.2) (4.4) (2.0) (0.3)
% of sales 2.8% 2.8% 2.8% 2.8% 2.8% 2.8% 2.8% 2.8% 2.8% 2.8% 2.8% 2.8% 2.8% 2.8% 2.8%
General & administrative expenses (120.3) (126.9) (130.9) (132.2) (132.2) (130.9) (128.3) (125.6) (95.2) (70.1) (48.9) (31.7) (17.2) (7.9) (1.3)
% of sales 11.0% 11.0% 11.0% 11.0% 11.0% 11.0% 11.0% 11.0% 11.0% 11.0% 11.0% 11.0% 11.0% 11.0% 11.0%
Other operating expenses, net — — — — — — — — — — — — — — —
Income (loss) from operations / EBIT 517.4 545.9 562.9 568.6 568.6 562.9 551.5 540.2 409.4 301.4 210.4 136.5 73.9 34.1 5.7
% margin 47.3% 47.3% 47.3% 47.3% 47.3% 47.3% 47.3% 47.3% 47.3% 47.3% 47.3% 47.3% 47.3% 47.3% 47.3%

C.II. Unlevered Free Cash Flow

FYE - 31st Dec 2035E 2036E 2037E 2038E 2039E 2040E 2041E 2042E 2043E 2044E 2045E 2046E 2047E 2048E 2049E
EBIT 517.4 545.9 562.9 568.6 568.6 562.9 551.5 540.2 409.4 301.4 210.4 136.5 73.9 34.1 5.7
Marginal Corporate Tax on EBIT (108.7) (114.6) (118.2) (119.4) (119.4) (118.2) (115.8) (113.4) (86.0) (63.3) (44.2) (28.7) (15.5) (7.2) (1.2)
NOPAT 408.8 431.2 444.7 449.2 449.2 444.7 435.7 426.7 323.4 238.1 166.2 107.8 58.4 27.0 4.5
add back: D&A 76.9 81.1 83.7 84.5 84.5 83.7 82.0 80.3 60.8 44.8 31.3 20.3 11.0 5.1 0.8
less: Increase (decrease) in NWC (2.4) (1.7) (1.0) (0.3) — 0.3 0.7 0.7 8.0 6.6 5.6 4.5 3.8 2.4 1.7
less: Capital expenditure (69.5) (73.4) (75.6) (76.4) (76.4) (75.6) (74.1) (72.6) (55.0) (40.5) (28.3) (18.3) (9.9) (4.6) (0.8)
Unlevered Free Cash Flow 413.7 437.3 451.7 457.0 457.3 453.1 444.3 435.1 337.3 249.0 174.8 114.3 63.3 29.9 6.3

Transition probabilities 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
Cash flow probabilities 14.1% 14.1% 14.1% 14.1% 14.1% 14.1% 14.1% 14.1% 14.1% 14.1% 14.1% 14.1% 14.1% 14.1% 14.1%

Source: Own calculations and assumptions as explained in 5.2.3.

52
 Figure 13 – Static DCF & NOL Valuation (ALD1910)
C.III. Discounted Cash Flow Analysis

FYE - 31st Dec 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E 2030E 2031E 2032E 2033E
Unlevered Free Cash Flow (32.5) (33.1) (28.9) (33.7) (38.0) (72.4) (84.4) (93.8) (310.0) (96.8) (3.7) 141.6 203.8 288.1 339.1
Period discounted (EOY) 0.17 1.17 2.17 3.17 4.17 5.17 6.17 7.17 8.17 9.17 10.17 11.17 12.17 13.17 14.17
Participation in yearly cash flow 16.7% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
WACC 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8%
Discount factor (mid-year adjusted) 99.0% 92.3% 81.8% 72.4% 64.2% 56.9% 50.4% 44.7% 39.6% 35.1% 31.1% 27.6% 24.4% 21.6% 19.2%
Discounted uFCF (5.4) (30.5) (23.7) (24.4) (24.4) (41.2) (42.5) (41.9) (122.7) (34.0) (1.1) 39.0 49.8 62.4 65.0
Static EV (if approved) 299.0

FYE - 31st Dec 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E 2030E 2031E 2032E 2033E
Unlevered Free Cash Flow (risk-adj.) (32.5) (33.1) (20.3) (23.6) (26.6) (19.3) (22.5) (25.0) (50.3) (13.6) (0.5) 19.9 28.7 40.5 47.7
Discounted uFCF (risk-adj.) (5.4) (30.5) (16.6) (17.1) (17.1) (11.0) (11.4) (11.2) (19.9) (4.8) (0.2) 5.5 7.0 8.8 9.2
Static EV (risk-adj.) (47.9)

C.IV. Net Operating Losses

Standalone Case
New NOL
NOL balance (BOY) — — — 25.1 55.0 89.2 157.8 238.4 328.3 634.5 731.1 747.5 625.4 452.4 204.6
NOL added 28.7 29.3 25.1 29.9 34.2 68.6 80.6 89.9 306.2 96.6 16.4 — — — —
NOL used — — — — — — — — — — — (122.1) (173.0) (247.9) (204.6)
NOL expired — — — — — — — — — — — — — — —
NOL balance (EOY) — — 25.1 55.0 89.2 157.8 238.4 328.3 634.5 731.1 747.5 625.4 452.4 204.6 —

FYE - 31st Dec 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E 2030E 2031E 2032E 2033E
Taxable income (28.7) (29.3) (25.1) (29.9) (34.2) (68.6) (80.6) (89.9) (306.2) (96.6) (16.4) 152.6 216.2 309.8 363.1
NOL used — — — — — — — — — — — (122.1) (173.0) (247.9) (204.6)
Taxable income (shielded) (28.7) (29.3) (25.1) (29.9) (34.2) (68.6) (80.6) (89.9) (306.2) (96.6) (16.4) 30.5 43.2 62.0 158.5
Tax rate 21.0% 0.2 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0%
NOL tax shield — — — — — — — — — — — 25.6 36.3 52.0 43.0
Cash taxes due — — — — — — — — — — — 6.4 9.1 13.0 33.3
Discounted NOL tax shield — — — — — — — — — — — 7.1 8.9 11.3 8.2
PV of NOL Tax Shield 35.4

Discounted NOL tax shield (risk-adj.) — — — — — — — — — — — 1.0 1.2 1.6 1.2

PV of NOL Tax Shield (risk-adj.) 5.0

C.III. Discounted Cash Flow Analysis

FYE - 31st Dec 2034E 2035E 2036E 2037E 2038E 2039E 2040E 2041E 2042E 2043E 2044E 2045E 2046E 2047E 2048E 2049E
Unlevered Free Cash Flow 381.0 413.7 437.3 451.7 457.0 457.3 453.1 444.3 435.1 337.3 249.0 174.8 114.3 63.3 29.9 6.3
Period discounted (EOY) 15.17 16.17 17.17 18.17 19.17 20.17 21.17 22.17 23.17 24.17 25.17 26.17 27.17 28.17 29.17 30.17
Participation in yearly cash flow 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
WACC 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8% 12.8%
Discount factor (mid-year adjusted) 17.0% 15.1% 13.3% 11.8% 10.5% 9.3% 8.2% 7.3% 6.5% 5.7% 5.1% 4.5% 4.0% 3.5% 3.1% 2.8%
Discounted uFCF 64.7 62.3 58.4 53.4 47.9 42.5 37.3 32.4 28.1 19.3 12.6 7.9 4.6 2.2 0.9 0.2

FYE - 31st Dec 2034E 2035E 2036E 2037E 2038E 2039E 2040E 2041E 2042E 2043E 2044E 2045E 2046E 2047E 2048E 2049E
Unlevered Free Cash Flow (risk-adj.) 53.6 58.2 61.5 63.6 64.3 64.4 63.8 62.5 61.2 47.5 35.0 24.6 16.1 8.9 4.2 0.9
Discounted uFCF (risk-adj.) 9.1 8.8 8.2 7.5 6.7 6.0 5.2 4.6 4.0 2.7 1.8 1.1 0.6 0.3 0.1 0.0

C.IV. Net Operating Losses

Standalone Case
New NOL
NOL balance (BOY) — — — — — — — — — — — — — — — —
NOL added — — — — — — — — — — — — — — — —
NOL used — — — — — — — — — — — — — — — —
NOL expired — — — — — — — — — — — — — — — —
NOL balance (EOY) — — — — — — — — — — — — — — — —

FYE - 31st Dec 2034E 2035E 2036E 2037E 2038E 2039E 2040E 2041E 2042E 2043E 2044E 2045E 2046E 2047E 2048E 2049E
Taxable income 406.6 440.5 464.7 479.3 484.1 484.1 479.3 469.6 459.9 348.5 256.6 179.1 116.2 62.9 29.0 4.8
NOL used — — — — — — — — — — — — — — — —
Taxable income (shielded) 406.6 440.5 464.7 479.3 484.1 484.1 479.3 469.6 459.9 348.5 256.6 179.1 116.2 62.9 29.0 4.8
Tax rate 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0% 21.0%
NOL tax shield — — — — — — — — — — — — — — — —
Cash taxes due 85.4 92.5 97.6 100.6 101.7 101.7 100.6 98.6 96.6 73.2 53.9 37.6 24.4 13.2 6.1 1.0
Discounted NOL tax shield — — — — — — — — — — — — — — — —

Discounted NOL tax shield (risk-adj.) — — — — — — — — — — — — — — — —

Source: Own calculations and assumptions as explained in 5.2.3.

53
5.2.5 ALD1910 – Real Options Assumptions & Model Construction

To estimate peak sales upon approval, a binomial tree following the CRR approach is constructed with a
horizon of nine years, in line with the projected development schedule. A time delta of one year is thus
assumed.

5.2.5.1 Volatility estimation

In this thesis, the project volatility for ALD1910 is assumed to be equivalent to the stock price volatility of
Alder, which was deemed reasonable as ALD1910 and Eptinezumab are somewhat similar candidates and thus
offer the closest observable market estimation of such a biologic migraine prevention candidate.

In harmony with the beta estimation in 5.2.1.9, an estimation window of 252 trading days, equivalent to one
year before the announcement, was selected. Following Benninga (2014) and Hull (2018), this is again justified
because we seek to depict the current information on / riskiness of the business, i.e. ALD1910 entering phase
I in the second half of 2019 and Eptinezumab having completed its phase III trials.

𝑖𝑆
In accordance with Hull (2018), volatility is estimated using log-returns 𝑢𝑖 = ln (𝑆 −1 ) via
𝑖

1 𝑛
𝑠=√ ∗ ∑ (𝑢𝑖 − 𝑢̅)2
𝑛−1 𝑖=1

where 𝑆𝑖 is the stock price at interval 𝑖 and 𝑛 is the number of observations. The annualised volatility is then
computed as 𝜎 = 𝑠 ∗ √𝑁. Table 8 below provides results on Alder and the NBI.

Table 8 – Overview of Volatility Estimates

NASDAQ: ALDR NASDAQ: NBI

No. of Obs. (N) 252 252

Volatility (𝜎) 60.5% 24.7%

Mean absolute daily log-return (𝑢

̅) 2.9% 1.2%

Source: Raw data from Bloomberg. Own calculations.

We can observe that Alder is more volatile than the market as represented by the NBI. This is in line with the
results from the beta regressions in appendix B-8. In particular, a volatility estimate per annum of 60.5% for
Alder is spot on when compared to Alder’s disclosed volatility estimates for its own employee stock options

54
model (Alder, 10-K 2018). However, it nevertheless remains at the very upper end of “observed” project
volatilities according to Bogdan and Villiger (2010). In a sense, it is appropriate as ALD1910 has only
concluded pre-clinical research so far.

5.2.5.2 Growth rate estimation

Following the concerns voiced by Bogdan and Villiger (2010), a growth rate of 0.0% was selected in order to
not run the risk of overestimating the bottom-up peak sales estimate derived in section 5.2.3.2.

5.2.5.3 Creating the binomial tree

Much debate exists on how to appropriately value a real option in the binomial tree (cf. section 3.4.3). While
the MAD assumption is necessary for Monte Carlo approaches, this thesis will use the more parsimonious
approach proposed by Bogdan and Villiger (2010). Following their approach, the risk-free rate is exchanged
with the growth rate (“drift”) when computing the probability of an up- and down-move to remain theoretically
consistent. At the same time, to remain theoretically consistent, the risk-adjusted WACC is used for
discounting the end node enterprise values within the binomial tree back to the root node. This was also deemed
appropriate as the author concluded that the MAD assumption is “shaky” at best.

We assume a time interval length of 1 year and derive 𝑢 and 𝑑 factors of 1.831 and 0.546, respectively. These
yield probabilities 𝑞 and (1 − 𝑞) of 35.3% and 64.7%, respectively. Using the peak sales estimate of USD
1,202.1 mn as obtained in 5.2.3.2, a binomial tree is constructed up to U.S. FDA approval in 2027. Figure 14
below showcases the result.

Figure 14 – Binomial Tree of Peak Sales Estimate for ALD1910

FYE - 31st Dec 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E
Peak sales estimate 1,202 2,201 4,029 7,377 13,505 24,725 45,266 82,873 151,723
657 1,202 2,201 4,029 7,377 13,505 24,725 45,266
359 657 1,202 2,201 4,029 7,377 13,505
196 359 657 1,202 2,201 4,029
107 196 359 657 1,202
58 107 196 359
32 58 107
17 32
10

State probabilities 100.0% 35.3% 12.5% 4.4% 1.6% 0.6% 0.2% 0.1% 0.0%
64.7% 45.7% 24.2% 11.4% 5.0% 2.1% 0.9% 0.4%
41.8% 44.3% 31.3% 18.4% 9.8% 4.8% 2.3%
27.1% 38.2% 33.8% 23.9% 14.7% 8.3%
17.5% 30.9% 32.7% 27.0% 19.1%
11.3% 24.0% 29.7% 27.9%
7.3% 18.1% 25.6%
4.7% 13.4%
3.1%

Source: Own calculations.

Above, the reader can observe how the peak sales estimate can fluctuate to extreme, even unrealistic values.

55
This is due to the comparatively high volatility estimate. It is important to realise that for example the upper
most node has only a probability of 0.02%, i.e. an expected value of USD 37 mn, whereas the initial peak sales
estimate has a probability of 19.1%, i.e. an expected value of USD 229 mn. Moreover, given that no growth
rate was assumed, the cumulative expected peak sales estimate in 2027 remains the same to today’s estimate
of USD 1,202 mn. Bogdan & Villiger (2010) remark how even extreme values are realistic in the sense that
certain drugs can turn out to be a blockbuster for several indications. For ALD1910, its efficacy regarding
episodic as well as chronic migraine has not yet fully crystallised. In addition, the PACAP molecule has also
been associated with other indications such as post-traumatic stress disorder and could therefore expand
ALD1910’s peak sales potential in the future (Ressler et al., 2011).

5.2.5.4 Valuation of the end nodes

The above peak sales binomial tree will now be used to obtain a real options valuation. First, a DCF valuation
is computed for each peak sales estimate in each end node. To do so, one can simply use an adjusted version
of the static DCF model from 5.2.4. For example, for a peak sales estimate of USD 1,202 mn, a DCF valuation
of USD 1,280 mn in 2027 is obtained. The cash flows are risk-adjusted for the final approval probability.

5.2.5.5 Working backwards to the root

After determining the values in each possible end node, backwards induction is used to determine the real
options valuation today. In non-decision nodes, the value can be determined by the following equation

𝑝𝑖 ∗ (𝑞 ∗ 𝐸𝑉𝑢𝑖+∆∗𝑡 + (1 − 𝑞) ∗ 𝐸𝑉𝑑𝑖+∆∗𝑡 )
𝐸𝑉𝑡 = + 𝐶𝐹𝑖
(1 + 𝑟𝑊𝐴𝐶𝐶 )∆∗𝑡

Effectively, the next period’s valuation is being risk- and probability adjusted as well as discounted. In
addition, the required cash flow for the respective period is added.

At decision points, one has to decide if to continue or to abandon the project, thus the equation becomes

𝑝𝑖 ∗ (𝑞 ∗ 𝐸𝑉𝑢𝑖+∆∗𝑡 + (1 − 𝑞) ∗ 𝐸𝑉𝑑𝑖+∆∗𝑡 )
𝐸𝑉𝑡 = max ( + 𝐶𝐹𝑖 , 0)
(1 + 𝑟𝑊𝐴𝐶𝐶 )∆∗𝑡

where 𝑖 denotes the time period and 𝑝𝑖 the transition probability. This is the case in 2019, 2021, 2024 and
2027, i.e. today, following phase I results, following phase II results and upon U.S. FDA approval. An
abandonment value of 0 is hence assumed.

56
5.2.6 ALD1910 – Real Options Valuation Results

Following the procedure laid out in the previous section, a real options valuation is obtained. The final
valuation tree is exhibited in figure 15 below.

Figure 15 – Real Options Valuation Tree for ALD1910

FYE - 31st Dec 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E
Real Options Valuation (USD mn) 28.5 28.8 138 624 1,429 2,986 15,740 31,010 60,673 194,907
(19) 60 292 771 4,439 9,034 17,933 57,972
Value of flexibility (USD mn) 76.4 — (14) 113 1,063 2,475 5,181 17,115
(66) (29) 72 513 1,374 4,925
% of nodes abandoned 13.3% (38) — (43) 231 1,280
— (161) (57) 180
(173) (94) —
(94) —
—

Source: Own calculations based on Bogdan & Villiger (2010).

Note: Grey areas represent decision nodes.

Above, a real options valuation of USD 28.5 mn for ALD1910 is obtained. Crucially, the value is positive as
compared to a negative risk-adjusted DCF valuation of USD (47.9) mn in section 5.2.4, which suggest a
managerial flexibility value of USD 76.4 mn. This significant option value seems reasonable if the
comparatively high volatility is considered as option premiums generally increase with volatility. Even more
so, Bogdan and Villiger (2010) explain how valuation differences between DCF and ROV disappear with a
project that is completely in-the-money as the flexibility to abandon becomes irrelevant. For ALD1910, the
opposite is true. In fact, one can see how the project is abandoned in one out of three peak sales cases in 2021,
in two out of six cases in 2024, and in three out of nine cases upon approval. Interestingly, the relatively high
applied volatility estimate fits the empirical findings by DiMasi et al. (2001) that around one third of
pharmaceutical projects are abandoned due to economic reasons.

Hence, also ex-post, hypothesis (III) that real options valuation is well-suited for valuing (a part of) Alder can
be confirmed.

Select sensitivities of the real options valuation, can be found in table 9 below.

Table 9 – Real Options Valuation – Sensitivity Analysis

Peak Sales Market Growth

31,010.5 962 1082 1202 1322 1443 0.0 -2.0% -1.0% 0.0% 1.0% 2.0%
50.0% — 7.1 19.8 32.4 45.1 50.0% 1.9 10.4 19.8 30.0 41.2
55.0% — 11.1 24.0 36.8 49.7 55.0% 5.9 14.5 24.0 34.3 45.6
Volatility
60.5% 2.9 15.4 28.5 41.5 54.5 60.5% 10.2 18.9 28.5 38.9 50.2
65.0% 6.7 18.9 32.1 45.2 58.4 65.0% 13.5 22.4 32.1 42.6 54.0
70.0% 10.8 22.6 35.9 49.3 62.6 70.0% 17.2 26.1 35.9 46.5 58.1

Source: Own calculations.

57
From table 9, one can recognise that peak sales of smaller ca. USD 950 mn will drive the real options value to
zero for volatility lower than Alder’s historical one. Also, a 1% change in market growth has less impact than
a peak revenue change of 10%.

5.2.7 SOTP Valuation (Standalone Case)

Finally, after computing valuations for both Eptinezumab and ALD1910, the total company valuation can be
obtained.

In essence, the valuations of Eptinezumab and ALD1910 are simply combined. For Eptinezumab, the risk-
adjusted DCF valuation including the present value of the NOL tax shield is chosen while for ALD1910, the
real options valuation is chosen over the static DCF as it most accurately represents the value for the
aforementioned advantages in section 3.5.

To accurately infer a valuation on an (implied) share price level, a so-called “EV-to-equity bridge” is
constructed, whereby our obtained enterprise value from unlevered FCF is adjusted for financing items and
other obligations as well as non-operating (off-)balance sheet items to which only the equity holders maintain
a right. In general, the adjustments were computed following Wessels, Goedhart and Koller (2015). The
exhibits in this thesis only show items that were either standard net debt components or relevant according to
Alder’s last reported financials (Alder, 2019a). Further explanations can be found in appendix D-1.

This work assumes a valuation on a fully diluted basis in order to be comparable to Lundbeck’s transaction
offer structure as per deal announcement (Lundbeck, 2019a). The fully diluted method assumes that all dilutive
in-the-money instruments are exercised thereby increasing the cash balance.

Regarding the applied share count, one then has to take into consideration the (hidden) dilutive equity claims
from Alder’s employee stock options, its convertible preferred stock and its convertible senior notes as per its
most recent 10-Q (Alder, 2019a). This is a reasonable effort as biotechs such as Alder often use mezzanine
financing and offer significant equity compensation. Importantly, one has to take into account how the
convertible instruments would behave in an acquisition situation. In the case of Alder, the indenture of its
convertible senior notes mandates that a liquidation event such as an acquisition will trigger conversion into
common stock while the convertible preferred stock will trigger the issuance of warrants with the right to buy
additional preferred stock (Alder, 2018; Alder, 2019c).

Appendix D-2 provides an overview of the different instruments and their share count dilution on a standalone
basis as well as in an acquisition case. These derivations are required in order to make consistent comparisons.

58
Figure 16 below summarizes the sum-of-the parts (SOTP) valuation results for Alder on standalone basis.

Figure 16 – Implied Sum-of-the-Parts Valuation of Alder (Standalone)

(USD mn) (USD mn, except share and per share data)
DCF Valuation of Eptinezumab 1,505.2 Total Enterprise Value 1,533.7
of which NOLs: 136.5 less: Financial debt —
% of total EV 98.1% less: Convertible preferred stock 106.5
less: Convertible senior notes, net 188.3
Real Options Valuation of ALD1910 28.5 less: Unfunded pension liabilites —
of which NOLs: 5.0 less: Operating lease liabilities 5.0
% of total EV 1.9% plus: Cash & cash-equivalents 88.4
plus: Short-term investments 320.9
Total Enterprise Value of Alder BioPharmaceuticals 1,533.7 plus: Net operating losses —
Diluted equity value 1,643.1
Basic shares outstanding: 83.6
plus: Newly issued shares from ITM options & RSUs 0.8
plus: Newly issued shares from ITM convertible preferred stock —
plus: Newly issued shares from ITM convertible senior notes, net —
Fully diluted shares outstanding 84.4
Implied fully diluted share price 19.46
Implied premium to current share price 93.4%

Source: Latest publicly available financial data as of June 2019 (Alder, 2019a). Own calculations.

The reader shall note above that the present value of the anticipated NOL tax savings is already included in
the definition of the total enterprise value here. Hence, an addition when computing the (diluted) equity value
is not necessary despite the NOLs being a non-operating asset, technically.

Above, one can discover how Alder can be valued at a total enterprise value of USD 1,533.7 mn on a standalone
basis as of 13th September 2019. Somewhat intuitively, late-stage Eptinezumab currently indeed dominates the
value of Alder while ALD1910 only contributes a small fraction of the overall value. After moving from EV
to equity value, one can see how Alder’s equity is valued higher than its EV due to the large cash holdings, i.e.
a negative net debt balance. The above assumes that conversion of the preferred stock and senior notes is not
triggered as we view Alder on a standalone basis, thus leading to an implied fully diluted share price of USD
19.46, which is markedly higher than the current share price of USD 10.03 as of 13 th September 2019. This
implies a potential undervaluation of Alder’s equity by the market.

5.3 Price vs. Value Discussion

One can now compare the purchase price consideration paid by Lundbeck to the estimated value of Alder.
Before diving into the numbers, paragraph 5.3.1 gives a condensed overview of the transaction to provide the
background for 5.3.2 where the structure of Lundbeck’s offer is summarised. In 5.3.3 the acquisition case sum-
of-the-parts valuation of Alder is provided before contrasting it to Lundbeck’s offer in 5.3.4. In particular, the
role of the contingent value right in Lundbeck’s offer is discussed. Lastly, the discussion is closed with an
assessment of Lundbeck’s strategic rationale for the transaction.

59
5.3.1 Transaction Overview at Announcement

On 16th September 2019, Lundbeck announced a definitive agreement to acquire NASDAQ-listed Alder
BioPharmaceuticals at a transaction valuation of up to USD 1.95 bn.24 Pursuant to the agreement, Lundbeck
commenced a public tender offer at USD 18.00 per share in cash plus one non-tradable contingent value right
(CVR) with a pay-out of USD 2.00 per share for all outstanding shares of Alder. The CVR is only paid out if
Alder’s lead candidate Eptinezumab is approved by EMA, in example (i.e.) should Eptinezumab successfully
obtain a marketing authorisation via the centralised procedure touched upon in section 1.4.3 (Lundbeck,
2019a). According to the press release, the “upfront cash consideration represents a 79% premium to Alder’s
shareholders based on the closing price on 13 September 2019 and an approximately 3% discount based on
the 52-week high share price” (Lundbeck, 2019a).

Alder’s board of directors unanimously approved this acquisition and recommended to its shareholders to
accept the tender offer. Lundbeck set a minimum acceptance threshold of 50% of all outstanding shares and –
following regulatory approvals – planned to close the transaction in the last quarter of 2019 (Lundbeck, 2019a).
On 21st October 2019, the tender offer ended with ca. 63.8% of stock tendered. Thus, the transaction closed in
October. Lundbeck funded the transaction using cash as well as bank credit facilities (Lundbeck 2019c).

5.3.2 Purchase Consideration & Structuring

In order to contrast Alder’s valuation with Lundbeck’s offer, one requires a detailed look at the structure of
Lundbeck’s offer consideration.

Lundbeck mentions that “the transaction is valued at up to USD 1.95 bn net of cash, on a fully diluted basis”
in its announcement (Lundbeck, 2019a). While it is not precisely clear to which common metric – EV, equity
value or transaction value including fees etc. – this number refers, it is likely the fully diluted equity value less
acquired cash and cash-equivalents. It was found that this definition comes within the margin of error.

Interestingly, Lundbeck filed an amendment to its ‘Schedule TO’ with the U.S. Securities and Exchange
Commission (SEC) on 27th September 2019 in which it calculates the (equity) transaction valuation in detail
to arrive at a price of USD 2,098.7 mn (Lundbeck, 2019d). The Schedule TO is the official tender offer
document that a prospective acquiror has to file with the SEC. It was decided to use this figure for the final
price vs. value comparison in the next section because it most accurately reflects the equity transaction value

24 Purchase consideration of up to USD 1.95 bn defined as net of cash acquired and on a fully diluted basis (Lundbeck, 2019a).

60
of this transaction. Importantly, it is based on information that was provided by Alder on 12th September 2019
(Lundbeck, 2019d).

Lundbeck is paying USD 18.00 per share in upfront cash as well as USD 2.00 per share in the form of a
contingent value right, which will pay out cash-settled if Eptinezumab is approved by the EMA in 2020.
Assuming a successful approval of Eptinezumab in Europe, this implies that the total offer to Alder’s
shareholders is 90% secure upfront cash and 10% contingent cash.

5.3.3 SOTP Valuation (Acquisition Case)

For the price vs. value comparison, the acquisition case is selected to reflect the valuation impact of the NOLs
correctly. Figure 17 below provides an overview of Alder’s estimated value from a M&A perspective.

Figure 17 – Implied Sum-of-the-Parts Valuation of Alder (Acquisition Case)

(USD mn) (USD mn, except share and per share data)
DCF Valuation of Eptinezumab 1,482.0 Total Enterprise Value 1,510.4
of which NOLs: 113.3 less: Financial debt —
% of total EV 98.1% less: Convertible preferred stock —
less: Convertible senior notes, net —
Real Options Valuation of ALD1910 28.5 less: Unfunded pension liabilites —
of which NOLs: 5.0 less: Operating lease liabilities 5.0
% of total EV 1.9% plus: Cash & cash-equivalents 285.7
plus: Short-term investments 320.9
Total Enterprise Value of Alder BioPharmaceuticals 1,510.4 plus: Net operating losses —
Diluted equity value 2,112.0
Basic shares outstanding: 83.6
plus: Newly issued shares from ITM options & RSUs 11.0
plus: Newly issued shares from ITM convertible preferred stock 8.4
plus: Newly issued shares from ITM convertible senior notes, net 14.2
Fully diluted shares outstanding 117.2
Implied fully diluted share price 18.02
Implied premium to current share price 79.1%

Source: Latest publicly available financial data as of June 2019 (Alder, 2019a). Own calculations.
Note: Change in cash balance due to assumed exercise of employee stock options and issuance of warrant shares following fully
diluted method.

Figure 17 above showcases the most relevant perspective for our transaction analysis, i.e. the acquisition case
where the value of the acquired NOLs is significantly lower. Given publicly available information, a defensible
estimate of Alder’s enterprise value has been derived as USD 1,510.4 mn. Following the fully diluted method,
which assumes that all in-the-money instruments are converted to shares, the share count is witnessing
significant dilution due to the acquisition while the debt component of the convertible preferred stock and the
convertible senior notes disappear from the bridge and cash is increased from exercise of dilutive instruments.
The implied fully diluted share price thus can be approximated as USD 18.02 per share, representing an implied
premium of 79.1%.

61
As it turns out, the overall result of this analysis is in fact almost in line with Lundbeck’s offer consideration.
The diluted equity value is estimated higher at USD 2,112.0 mn compared to the stated (suspected) equity
value of USD 1.95 bn but in line with the filed Schedule TO amendment value of USD 2,098.7 mn. The reader
shall note that several likely reasons for a gap exist: (1) the assumptions underlying the valuation of either
Eptinezumab or ALD1910, or both, likely differ from Lundbeck’s, (2) Lundbeck has most likely assigned a
different value to the NOL balance following its tax due diligence, (3) a (slightly) different diluted share count
was used by Lundbeck due to ongoing stock option activity, potential treasury buybacks, etc., and (4) the cash
and cash-equivalent balances will likely have changed significantly in August and September due to the launch
preparations for Eptinezumab.

5.3.4 Price vs. Value Comparison

Finally, the purchase consideration offered by Lundbeck can be compared to the estimated value of Alder.
Paragraph 5.3.4.1 will compare price to value before a dedicated analysis of the contingent cash component is
conducted in 5.3.4.2.

5.3.4.1 Price vs. value assessment

Figure 18 below showcases the result of this paper’s valuation work to the purchase consideration paid by
Lundbeck.

Figure 18 – Price vs. Value Comparison

Price Value

2,099 2,099 2,112

10% CVR
ALD1910 Eptinezumab
Net casha 602
28.5 1,510
(USD millions)

90%
USD 18.00 1,510 1,482
per share EV
(upfront)

Total Equity Offer Structure Implied Implied

Purchase Price Equity Value Enterprise Value

Source: Latest publicly available financial data as of June 2019 (Alder, 2019a). Own calculations.
a
Negative net debt. Also including operating leases. Excluding convertible instruments due to assumed conversion.

62
Above, it can be discovered that the purchase consideration that Lundbeck offered for Alder can indeed be
seen as ‘fair value’. It assumes, however, that the CVR component is paid out, i.e. an approval of Eptinezumab
in the European market. This is in fact a critical assumption and warrants closer examination in the next
paragraph.

Hypothesis (IV), namely that the real options approach can help explain Lundbeck’s offer of almost 100%,
can be confirmed in that it reveals a significant flexibility value compared to a static DCF which, in our case,
yielded actually a negative value. However, the value of ALD1910 is comparatively small in relation to the
total implied enterprise valuation.

Hypothesis (V), that Lundbeck paid ‘fair value’ for Alder, can also be confirmed judging from figure 18.

5.3.4.2 Contingent value right

It can be learned in figure 18 above how Eptinezumab most likely represents the far majority of value in this
transaction. While one could be optimistic regarding the approval of Eptinezumab due to its distinguished
clinical profile and the earlier approvals of relatively similar CGRP drugs, it is not a certainty. A rejection or
even just a delay of the approval in February 2020 would have serious consequences for the value of Alder. It
remains unknown why a CVR structure was only agreed upon for the European market and not the U.S.
approval as well. One could suspect that Lundbeck might have tried to negotiate a CVR structure for the U.S.
approval of Eptinezumab as well but Alder drove a hard bargain. Eventually, the parties agreed upon a CVR
for the European procedure only. This is still of value for Lundbeck because empirical evidence exists for the
EMA rejecting previously U.S. FDA approved drugs – even though in only less than 10% of cases (Kashoki
et al., 2019).

Hence, Lundbeck here effectively managed to transfer a part of the uncertainty (i.e. risk) for the lead asset to
the selling shareholders. Somewhat misleading, the transaction announcement states that “the terms of the
CVR payment reflect the parties' agreement over the sharing of potential economic upside benefits from such
approval” (Lundbeck, 2019a). While this is true, one has to remind himself that it also requires sharing the
downside risks. There is no such thing as a free lunch as Friedman (1975) said.

Therefore, the hypothetical USD 2.00 per share are only achieved in the good state and not in the bad state.
Thus, the expected value today is less than USD 2.00 per share but will converge to USD 2.00 per share if the
EMA procedure goes ahead as planned. When valuing the CVR, one also has to account for time value of
money which reduces the value today as well. Kirman and Goldfeld (2011) furthermore point out that the CVR
is effectively a deferred financing for Lundbeck as it is only paid out upon approval in 2020. At the same time,
Lundbeck opens itself up to possible litigation due to the complex contractual nature of CVRs. In addition,
Kirman and Goldfeld (2011) state that especially non-tradable CVRs such Lundbeck’s can suffer from an

63
illiquidity discount which can lead the target’s shareholders to place less value on the purchase consideration
offered. Lastly, when using a CVR structure, the target shareholders are effectively being exposed to the credit
risk of the acquiror (Kirman & Goldfeld, 2011).

Effectively, a CVR is a binary option in itself. Overall, one can suspect that Lundbeck has gained from the
CVR structure, i.e. by selling a binary call to Alder’s shareholders, not least because it might have enabled the
deal by bridging disagreements around uncertainty. In fact, Fosheim and Langerød (2017) estimate a
probability increase in deal completion of 13.9 to 22.1% when a CVR is used.

5.3.5 Lundbeck’s strategic rationale

So far, this chapter has concluded that Lundbeck paid ‘fair value’ under the assumption of this analysis.
However, even if a purchase price of ca. USD 2.0 bn can indeed be deemed ‘fair’, an acquiror still faces
considerable risks stemming from deal execution, post-merger integration and lifting potential synergies,
amongst others. In particular, biotech M&A is a particularly risky endeavour due to the binary outcome of
drug development. This poses the question of why Lundbeck chose to acquire Alder for such a massive
premium.

Based on press coverage at the time, Lundbeck was facing a dismal outlook in its pipeline following years of
strong sales driven by price increases on older legacy products, which at the time accounted for almost 60%
of revenue. In 2019, following several late-stage failures, Lundbeck’s CEO pronounced her interest in
acquiring outside innovation for up to USD 4.0 to 5.0 bn, not least also in order to counter the looming expiry
of further patents (Urquhart, 2019).

Urquhart (2019) further explains how the CNS landscape generally had few suitable assets for Lundbeck at
the time given its “limited” M&A firepower. Regardless, Lundbeck announced its acquisition of Abide
Therapeutics for USD 250 mn in May 2019 which strengthened its Tourette syndrome pipeline but does not
immediately offer a solution to Lundbeck’s looming revenue drop-off (Brown, 2019). Subsequently, the case
for Alder must have come into focus given its supposed ‘blockbuster’ potential, its near-commercialisation
status and a recently depressed, downwards-trending share price development (cf. figure 4). Migraine fits well
into Lundbeck’s neurology focus even though Eptinezumab will be the first migraine product for Lundbeck.

To summarise, Lundbeck’s rationale for spending ca. USD 2.0 bn on Alder can likely be explained by the
dismal prospect of its own pipeline, which can be regarded as the classic pipeline replenishment theme. In
accordance with Bauer and Fischer (2000), Alder could be a “bridge deal” for Lundbeck to secure short-term
cash flow while it continues to search for an own blockbuster candidate. In addition, Alder posed the
opportunity to diversify by acquiring migraine capabilities which could become a major revenue driver over
the next years with increasing adoption of CGRPs. However, Eptinezumab is a clear late-mover in this market

64
and will require Lundbeck to efficiently use its existing CNS sales force in order to establish Eptinezumab
against its competitors, all of which have shown similar cross-trial efficacy. On the other hand, its distinguished
clinical profile will likely enable it to establish its niche. Moreover, Lundbeck is also acquiring ALD1910, a
promising early-stage compound which will support is long-term pipeline.

5.4 Event Study of Lundbeck’s Share Price Reaction

This thesis argues that Lundbeck has likely paid ‘fair value’ for Alder. As this verdict is based on a single
person’s model with various sensitive underlying assumptions, the market’s perspective is necessary to
substantiate – or contradict – the results of this paper. This can be done by investigating the acquiror’s stock
price reaction around the deal announcement. Alder’s stock price is of less interest here given the very clear
reaction to the massive premium offered.

Lundbeck exhibited a negative stock price return of (0.6%) on 16th September 2019. This might not necessarily
have been due to the deal announcement but just by chance. To further assess the reaction, statistical testing is
required.

5.4.1 Event study construction

Using the event study technique, we test the hypothesis

𝐻0 : 𝐶𝐴𝑅 = 0 against 𝐻𝐴 : 𝐶𝐴𝑅 ≠ 0

In the estimation window, expected returns are estimated using the common single index market model which
is a straightforward linear regression of the individual stock returns 𝑟𝑖,𝑡 onto the market’s returns 𝑟𝑚,𝑡 via

𝑟𝑖,𝑡 = 𝛼𝑖 + 𝛽𝑖 ∗ 𝑟𝑚,𝑡 + 𝜀𝑖,𝑡

where 𝛼 is the intercept, 𝛽 the relative movement to the market and 𝜀 the residual error component (Stapleton
& Subrahmanyam, 1983). The OMX Copenhagen 25 index was chosen over the NBI as the market index for
Lundbeck because it had better explanatory power for Lundbeck’s stock price returns in the estimation window
with a 𝑅 2 of 0.163 versus 0.015 for the NBI. It needs to be noted, that Lundbeck’s stock exhibits relatively
dispersed returns and thus only limited explanatory power exists with both indices. In fact, one outsized
negative return of (26.6%) on 25th October 2018 was deleted for Lundbeck to improve the parameter
regressions. This likely related to the announcement that its lead phase III schizophrenia drug Lu-AF35700
failed (Lundbeck, 2018). It was deemed appropriate in order to predict appropriate trading behaviour with the
market model parameters. The two market model regression can be seen in appendix E-1.

The event study is constructed with event window length 𝑇, in which abnormal returns are then defined as

65
 𝑎𝑏𝑛𝑜𝑟𝑚𝑎𝑙 𝑟𝑒𝑎𝑙𝑖𝑠𝑒𝑑 𝑒𝑥𝑝𝑒𝑐𝑡𝑒𝑑 𝑟𝑒𝑎𝑙𝑖𝑠𝑒𝑑
𝑟𝑖,𝑡 = 𝑟𝑖,𝑡 − 𝑟𝑖,𝑡 = 𝑟𝑖,𝑡 − (𝛼𝑖 + 𝛽𝑖 ∗ 𝑟𝑚,𝑡 )

with 𝑟𝑡 as the (simple) stock price return and cumulative abnormal returns (CAR) subsequently being defined
as
𝑇
𝑎𝑏𝑛𝑜𝑟𝑚𝑎𝑙
𝐶𝐴𝑅 = ∑ 𝑟𝑖,𝑡
𝑡=1

With regards to testing, the applied two-tailed t-test statistic is defined as

𝐶𝐴𝑅
𝑡𝐶𝐴𝑅 =
𝜎𝐶𝐴𝑅

with CAR standard deviation 𝜎𝐶𝐴𝑅 , while the non-parametric rank test statistic 𝑇𝑟𝑎𝑛𝑘 following Corrado (1989)
is defined as

1
(𝑁 ∗ ∑𝑁 ̅
𝑖=1(𝐾𝑡 − 𝐾 )
𝑇𝑟𝑎𝑛𝑘 =
𝜎(𝐾)
𝑎𝑏𝑛𝑜𝑟𝑚𝑎𝑙 ̅ over the entire
with 𝑁 observations, 𝐾𝑖,𝑡 as the rank of abnormal return 𝑟𝑖,𝑡 , the average rank 𝐾
observations and 𝜎(𝐾) as the standard deviation of 𝐾 defined as

𝑁 2
1
𝜎(𝐾) = √( ∗ ∑(𝐾𝑖,𝑡 − 𝐾
̅ ))
𝑁
𝑖=1

An overview of the results can be found in the subsequent section.

66
5.4.2 Event study results

Following the previously described statistical procedures, selected results are presented in table 10 below.

Table 10 – Overview of Event Study Results

t-test Rank test

Interval Days CAR t-statistic p-value Rank statistic Rank p-value

CAR [-5,0] 6 (1.96%) (0.4914) 0.6236 0.5819 0.5611

CAR [-1,0] 2 (2.13%) (0.9252) 0.3558 0.9706 0.3327

CAR [0] 1 0.04% 0.0226 0.9820 (0.0858) 0.9317

CAR [0,1] 2 (2.70%) (1.1747) 0.2412 1.0546 0.2926

CAR [0,4] 5 3.20% 0.8795 0.3800 (0.7466) 0.4560

Source: Raw data from Bloomberg. Own calculations.

Note: CAR [0] anchors the event day, i.e. the 16th of September 2019.
* / ** / *** indicate significance at 10 / 5 / 1 percent, respectively.

Above, one can infer that none of the selected CARs can be deemed significant. Hence, we cannot reject the
null hypothesis 𝐻0 : 𝐶𝐴𝑅 = 0 in favor of the alternative. There has thus not been any statistically significant
movement above what can be expected given the model in use. This is also substantiated by identical results
from the non-parametric rank test.

In particular, the above event study finds that no statistically significant impact on the stock took place during
the week prior to the announcement as well as the week following the announcement. As a visualisation, the
following figure 19 depicts the CAR throughout the event window.

67
 Figure 19 – Cumulative Abnormal Return in the Event Window

6.0%

4.0%
Cumulative abnormal return

2.0%

(2.0%)

(4.0%)

(6.0%)
(5) (4) (3) (2) (1) 0 1 2 3 4

Timeline of Event Window

Source: Raw data from Bloomberg. Own calculations.

Note: The event day was anchored to the 16th of September 2019. Dotted line marks event day.

Above, it can be observed how Lundbeck’s CAR seems to oscillate in the weeks before and after the
announcement. However, none of these individual abnormal returns as well as CARs were statistically
significant.

As a result, the event study does not contradict the findings of the valuation analysis in this paper but it also
does not necessarily confirm that Lundbeck paid an appropriate or even attractive price. For example, it could
be possible that additional information was priced in on the announcement day which counterweighted a
potential impact from the deal announcement.

Subsequently, hypothesis (VI), i.e. that Lundbeck’s stock exhibited a statistically significant positive reaction
upon the deal’s announcement can be rejected as Lundbeck’s stock has seemingly not moved significantly in
either direction. Under the assumptions in this event study and its inherent prediction model, this implies that
Lundbeck has not created additional value for its shareholders (during the event window at least). This also
connects logically to the finding that Lundbeck paid ‘fair value’ for Alder.

68
7. Conclusion

In September 2019, the management of H. Lundbeck A/S decided to spend ca. USD 2.0 bn on the acquisition
of Alder BioPharmaceuticals. This research work has focused on the question if Lundbeck has paid ‘fair value’
for Alder and if real options can help to understand this.

7.1 Summary of Results

Following a detailed investigation of Alder’s assets and their respective market / revenue potential, this paper
valued Alder at an EV of USD 1,534 mn on a standalone basis and USD 1,510 mn from an acquiror’s
perspective. In the latter, using a risk-adjusted DCF model, Eptinezumab was valued at USD 1,482 mn
including an estimated present value of a tax shield from (acquired) NOLs of USD 113.3 mn. Alder’s early-
stage asset ALD1910, on the other hand, was valued at an EV of USD (47.9) mn using a static risk-adjusted
DCF and at USD 28.5 mn using a real options valuation approach, implying a managerial flexibility value of
USD 76.4 mn. This translates into an implied equity value of USD 2,112 mn and is in line with Lundbeck’s
fully diluted equity purchase price of USD 2,098.7 mn. Hence, it was argued that Lundbeck paid ‘fair value’.
This result was not contradicted by an event study on Lundbeck’s share price reaction upon announcement.

Overall, this thesis set out to investigate the research question if Lundbeck can justify the price paid for Alder
to its shareholders and how real options can help understand this. This problem was investigated using several
hypotheses.

Revisiting the hypothesis roadmap from the beginning, it was confirmed that real options provide advantages
over the DCF analysis, especially regarding settings with multi-staged uncertainty. In particular, real options
are well-suited to capture the dynamics of the drug development process. In the case of Alder, real options
were deemed as the appropriate valuation tool for determining a valuation of its early-stage asset ALD1910
due to large uncertainty.

The analysis has subsequently shown that a real options approach can help explain Lundbeck’s willingness to
pay a premium of close to 100% as it unveiled a significant flexibility value for ALD1910 that turned its DCF
value from negative to positive. However, in a case like Alder with a lead near-commercialisation asset as the
focus of the transaction, even a real options valuation of an early-stage drug will likely be comparatively small
in relation to the total valuation. To a certain extent, this can limit the “practical return” from applying real
options valuation. In such situations, a parsimonious real options model should therefore be prioritised over
more advanced models like simulation procedures for example.

69
Moreover, it was reasoned that Lundbeck’s offer can be considered ‘fair value’. Lundbeck can justify the price
to its shareholders as it secures short-term cash flow for further R&D following several recent setbacks.
However, while the market has not contradicted this result, it has also not shown any significant positive
abnormal return upon announcement, thus potentially indicating only limited value creation for Lundbeck’s
shareholders, if any.

Based on this analysis and its underlying assumptions, Lundbeck has likely not overpaid. However, valuation
is not always a precise science but often “part art, part science” as practitioners commonly confess. This also
holds (somewhat) true in biotech despite the very scientific background of biopharmaceutical product
development and marketing. To obtain accurate, defensible valuation results, extensive sector knowledge on
both medicine as well as healthcare commercialisation in a range of different markets is required. For novel
therapeutics, dedicated on-the-ground market research on physicians and patients becomes basically
mandatory unless one resorts to simple approximations as was done in this work. In the end, also in biotech, it
all comes down to navigating the ensuing M&A negotiation. Special agreements such as contingent value
rights offer extended means for both parties to navigate the inherent uncertainty of novel therapeutics. The
case of Alder shows that they can potentially become crucial in tipping the acquisition’s risk-reward balance
in favour of a particular party.

To conclude, in 1984, Myers declared that “the value of R&D is almost all option value” (p. 135). At the same
time, while addressing the divide between financial management based on DCF analysis and strategic
management, Myers (1984) also expressed that “although complete reconciliation will rarely be possible, the
attempt (N.B. to use real option approaches) should uncover hidden assumptions and bring a generally deeper
understanding of strategic choices. The gap may remain, but with better analysis on either side of it” (p. 136).

As it turns out, the key implication from the findings of this paper with regards to real options is congruent to
Myers’ statement. Within biotech and potentially other sectors, real options are likely of greatest use if used
selectively together with (risk-adjusted) DCF analysis, where appropriate. In particular, the thought process of
constructing the more complex real options analysis will likely yield a much more valuable understanding of
the multi-stage uncertainty and the inherent option flexibility of (early-stage) biopharmaceutical R&D assets.

7.2 Limitations

Several limitations of the analysis presented in this paper exist. Firstly, the income statement and free cash
flow projections are generally always subject to debate. It was focused on deriving defensible assumptions
instead of generic ranges / several operating scenarios. In particular, the approach of valuing Eptinezumab and
ALD1910 separately before combining them, exposes the difficulty of accurately allocating certain overhead
expenses etc. With regards to model inputs, Kellogg and Charnes (2000) conclude that the use of average

70
assumptions works well for early-stage projects while more sophisticated knowledge is needed for later-stage
assets. Both types of projects require in-depth sector knowledge.

As an outsider, it is extremely difficult to accurately allocate costs and other income statement items to
individual projects. Hence, certain overhead expenses might have been over- or underestimated but it was tried
to allocate such mainly to Eptinezumab. On a different note, the benchmarking approach using a (large cap)
peer group is not necessarily appropriate for a biotech but a comparison to other biotechs is rarely feasible
either.

Another assumption is that no R&D takes place beyond the terminal value. While a convenient assumption for
this analysis, it likely does not hold true as Lundbeck guided large retention costs for Alder’s employees
(2019a). Alder is establishing Lundbeck’s migraine competence centre, hence one could argue that their R&D
capabilities will yield new drug candidates in the future.

Regarding the cost of capital, a uniform rate is used. This is hardly realistic as each successful completion of
a development stage will significantly de-risk the asset in question. Thus, Myers and others (Myers & Shyam-
Sunder, 1996; Myers & Howe, 1997) have argued that the cost of capital for R&D should decline over the
development process as a step function. They termed the relationship a “risk-return staircase” which implies
an undervaluation of future stages due to too much discounting.

It is important to note that the acquisition case does not take an explicit view on any potential synergies but
merely acknowledges existing regulation regarding the use of acquired NOLs. Any synergies would likely be
on the revenue side given that Lundbeck is an established neurology player with an extensive sales force across
many markets, thus limiting the human resources investment that Alder would need to make on a standalone
basis. On the other hand, Lundbeck does not have an extensive biologics manufacturing capability either, thus
limited cost synergies apart from consolidating corporate function which will likely be small for Alder.

With ALD1910, the developed epidemiology model is only a very simple approximation. More
access to epidemiology data and more knowledge on different healthcare systems across main markets would
likely yield a more precise and reliable forecast. For example, the gross-to-net discount will certainly be
different in Europe compared to the U.S.

Also, the cost structure of the development phase is difficult to estimate without a bottom-up budgeting plan
and information on how many trials are envisaged at each stage.

The binomial tree was constructed with nine periods. Hull (2018) mentions the use of around 30 in order to
more precisely approximate an equivalent closed-form solution. However, in the case of Alder, the value of
ALD1910 will likely still be small, thus just binding computational resources.

In the event study, it was discovered that Lundbeck’s returns are not particularly well-explained by
both the NBI and the OMXC25 index. Subsequently, the linear market model could perhaps not be the ideal

71
model to predict returns in this event study. In addition, an intraday event study could offer additional insight
by further isolating the information release and mitigating other noise.

7.3 Avenues for Future Research

Empirical surveys have shown repeatedly that real options – despite their promises – have seen rather reluctant
adoption by practitioners due to variety of reasons (Holmberg & Jeppsson, 2010; Block, 2007; Hartmann &
Hassen, 2006; Puran, 2005). There also exists a broad range of different methods with diverse, often little
intuitive, underlying assumptions (Collan, 2011; Barton & Lawryshyn, 2010).

Future research should focus on providing more (parsimonious) industry specific applications of the available
pool of methods so that adoption hurdles for practitioners are reduced. In the public debate, relatively simplistic
but potent models such as the binomial tree suffer from the prominence of the Black-Scholes-Merton closed-
form model, despite it being of very limited use for valuing real options.

Regarding biopharmaceutical R&D, one of the best fields for applying a real options approach, future research
should also focus on the application and impact of stage-dependent cost of capitals, i.e. the “risk-return
stairway”. Puran (2005) finds only very limited evidence of this in practice despite the potentially significant
change in value for early-stage assets. Also, the sector would clearly benefit from more knowledge sources for
evaluating for example licensing or contracting options.

Further areas could include “tactical” real options in a general M&A transaction context, where one could
incorporate game theory.

72
II. References

AbbVie. (2014). Form 10-K 2013. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

AbbVie. (2015). Form 10-K 2014. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

AbbVie. (2016). Form 10-K 2015. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

AbbVie. (2017). Form 10-K 2016. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

AbbVie. (2018). Form 10-K 2017. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

AbbVie. (2019). Form 10-K 2018. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

AbbVie. (2020). AbbVie Reports Full-Year and Fourth-Quarter 2019 Financial Results [Website]. Retrieved
May 15, 2020, from https://investors.abbvie.com/news-releases/news-release-details/abbvie-reports-
full-year-and-fourth-quarter-2019-financial

ACADIA Pharmaceuticals. (2016). Form 10-K 2015. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

ACADIA Pharmaceuticals. (2017). Form 10-K 2016. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

ACADIA Pharmaceuticals. (2018). Form 10-K 2017. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

ACADIA Pharmaceuticals. (2019). Form 10-K 2018. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Acorda Therapeutics. (2014). Form 10-K 2013. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Acorda Therapeutics. (2015). Form 10-K 2014. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Acorda Therapeutics. (2016). Form 10-K 2015. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Acorda Therapeutics. (2017). Form 10-K 2016. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Acorda Therapeutics. (2018). Form 10-K 2017. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

73
Acorda Therapeutics. (2019). Form 10-K 2018. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Alder BioPharmaceuticals. (2014). Form 10-K 2013. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Alder BioPharmaceuticals. (2015). Form 10-K 2014. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Alder BioPharmaceuticals. (2016). Form 10-K 2015. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Alder BioPharmaceuticals. (2017). Form 10-K 2016. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Alder BioPharmaceuticals. (2018a). Form 8-K, January 29. Retrieved from SEC EDGAR at
https://www.sec.gov/Archives/edgar/data/1423824/000119312518029470/d520136d8k.htm

Alder BioPharmaceuticals. (2018b). Form 10-K 2017. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Alder BioPharmaceuticals. (2019a). Form 10-Q For The Quarter Ended June 30, 2019. Retrieved May 22,
2020, from https://www.bamsec.com/filing/156459019029315?cik=1423824

Alder BioPharmaceuticals. (2019b). Q2 2019 Earnings Call [Transcript]. Retrieved May 22, 2020, from
https://www.fool.com/earnings/call-transcripts/2019/08/07/alder-biopharmaceuticals-inc-aldr-q2-
2019-earnings.aspx

Alder BioPharmaceuticals. (2019c). Form 10-K 2018. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Alder BioPharmaceuticals. (n.d.). Our Focus [Company website]. Retrieved May 14, 2020, from
http://web.archive.org/web/20190916091832/https://www.alderbio.com/pipeline/

Alexion Pharmaceuticals. (2014). Form 10-K 2013. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Alexion Pharmaceuticals. (2015). Form 10-K 2014. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Alexion Pharmaceuticals. (2016). Form 10-K 2015. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Alexion Pharmaceuticals. (2017). Form 10-K 2016. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Alexion Pharmaceuticals. (2018). Form 10-K 2017. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Alexion Pharmaceuticals. (2019). Form 10-K 2018. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

74
Alkermes. (2014). Form 10-K 2013. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Alkermes. (2015). Form 10-K 2014. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

Alkermes. (2016). Form 10-K 2015. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

Alkermes. (2017). Form 10-K 2016. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

Alkermes. (2018). Form 10-K 2017. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

Alkermes. (2019). Form 10-K 2018. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

American Headache Society. (2018). The American Headache Society Position Statement On
Integrating New Migraine Treatments Into Clinical Practice [Consensus Statement]. Headache, 59,
1-18. doi:10.1111/head.13456

American Migraine Foundation. (2020). Navigating Life with Migraine [Website]. Retrieved May 17, 2020,
from https://americanmigrainefoundation.org/living-with-migraine/migraine-essentials/

Amgen. (n.d.). Paying for Aimovig [Company website]. Retrieved May 16, 2020, from
https://www.aimovig.com/paying-for-aimovig/#0

Amgen. (2014). Form 10-K 2013. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

Amgen. (2015). Form 10-K 2014. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

Amgen. (2016). Form 10-K 2015. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

Amgen. (2017). Form 10-K 2016. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

Amgen. (2018). Form 10-K 2017. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

Amgen. (2019). Form 10-K 2018. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

Armstrong, M. (2019, September 16). Alder won’t solve Lundbeck’s headaches. Evaluate Vantage. Retrieved
May 15, 2020, from https://www.evaluate.com/vantage/articles/news/deals/alder-wont-solve-
lundbecks-headaches

Armstrong, M., Brown, A. & Fagg, J. (2019, August 30). Upcoming events – pivotal tests approach for Lilly
and Biohaven. Evaluate Vantage. Retrieved May 16, 2020, from https://www.evaluate.com/vantage/
articles/events/upcoming-events/upcoming-events-pivotal-tests-approach-lilly-and-biohaven

75
Barton, K. & Lawryshyn, Y. (2010). Reconciling Real Option Models: An Approach to Incorporate Market
and Private Uncertainties. 14th Annual International Conference on Real Options Theory Meets
Practice [Conference paper]. Boston, USA. Retrieved May 17, 2020, from
http://www.realoptions.org/papers2010/106.pdf

Basu, P. K., Suresh, p., Joglekar, G., Rai, S. & Vernon, J. (2008). Analysis of Manufacturing Costs in
Pharmaceutical Companies. Journal of Pharmaceutical Innovation, 3, 30-40. doi:10.1007/s12247-
008-9024-4

Bauer, H. H. & Fischer, M. (2000). Product life cycle patterns for pharmaceuticals and their impact on R&D
profitability of late mover products. International Business Review, 9(6), 703-725. doi: 10.1016/
S0969-5931(00)00028-7

Bendtsen, L., Sacco, S., Ashina, M. et al. (2018). Guideline on the use of onabotulinumtoxinA in chronic
migraine: a consensus statement from the European Headache Federation. Journal of Headache and
Pain, 19(91). doi:10.1186/s10194-018-0921-8

Benninga, S. (2014). Financial Modeling (4th ed.). MA, Cambridge: The MIT Press. ISBN:9780262027281

Berkshire Hathaway. (2009). Annual Letter to the Shareholders, p.5. NE, Omaha. Retrieved May 11, 2020,
from https://www.berkshirehathaway.com/letters/2008ltr.pdf

Bertels, Z. & Pradhan, A. A. (2019). Emerging Treatment Targets for Migraine and Other Headaches.
Headache – The Journal of Head and Face Pain, 59(52), 50-65. doi: 10.1111/head.13585

Bigal, M. E., Walter, S. & Rapoport, A. M. (2015). Therapeutic antibodies against CGRP or its receptor. British
Journal of Clinical Pharmacology, 79(6), 886-895. doi:10.1111/bcp.12591

BIO. (2016). Clinical Development Success Rates 2006-2015 [PowerPoint Report]. Retrieved May 20, 2020,
from http://go.bio.org/rs/490-EHZ-999/images/Clinical%20Development%20Success%20Rates
%202006-2015%20-%20BIO%2C%20Biomedtracker%2C%20Amplion%202016.pdf?_ga=2.5
8159848.371056006.1585326608-941322881.1585326608

Biogen. (2014). Form 10-K 2013. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

Biogen. (2015). Form 10-K 2014. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

Biogen. (2016). Form 10-K 2015. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

Biogen. (2017). Form 10-K 2016. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

Biogen. (2018). Form 10-K 2017. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

Biogen. (2019). Form 10-K 2018. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

Black, F. & Scholes, M. (1973). The Pricing of Options and Corporate Liabilities. Journal of Political

76
 Economy, 81(3), 637-659. Retrieved May 1, 2020, from https://www.jstor.org/stable/1831029

Block, S. (2007). Are “Real Options” Actually Used in the Real World? The Engineering Economist, 52(3),
255-267. doi:10.1080/00137910701503910

Bogdan, B. & Villiger, R. (2005a). Getting real about valuations in biotech. Nature Biotechnology, (23)(4),
423-428. doi:10.1038/nbt0405-423

Bogdan, B. & Villiger, R. (2005b). Valuing Pharma R&D: The Catch‐22 of DCF. Journal of Applied
Corporate Finance, 17(2), 113-116. doi: 10.1111/j.1745-6622.2005.00039.x

Bogdan, B. & Villiger, R. (2006). Pitfalls of valuation in biotech. Journal of Commercial Biotechnology, 12(3),
175-181. Retrieved May 14, 2020, from https://link.springer.com/content/pdf/10.1057/palgrave.jcb
.3040164.pdf

Bogdan, B. & Villiger, R. (2010). Valuation in Life Sciences (3rd ed.). Heidelberg: Springer.

Boyle, P. P. (1977). Options: A Monte Carlo Approach. Journal of Financial Economics, 4(3), 323-338.
doi:10.1016/0304-405X(77)90005-8

Boyle, P. P., Broadie, M. & Glasserman, P. (1997). Monte Carlo Methods for Security Pricing. Journal of
Economic Dynamics and Control, 21(8-9), 1267-1321. doi:10.1016/S0165-1889(97)00028-6

Brandao, L. E., Dyer, J. S. & Hahn, W. J. (2005). Using Binomial Trees of Solve Real Option Valuation
Problems. Decision Analysis, 2(2), 69-88. doi:10.1287/deca.1050.0040

Brealey, R. A., Myers, S. C. & Allen, F. (2014). Principles of Corporate Finance (11th ed.). New York, NY:
McGraw-Hill Education. ISBN-13:9780077151560

Broadie, M., Glasserman, P. & Jain, G. (1997). Enhanced Monte Carlo Estimates for American Option Prices.
The Journal of Derivatives, 5(1), 25-44. doi:10.3905/jod.1997.407983

Brown, A. (2019). Lundbeck leaps into the Tourette’s pipeline. Evaluate Vamtage. Retrieved May 25, 2020,
from https://www.evaluate.com/vantage/articles/news/corporate-strategy/lundbeck-leaps-tourettes-
pipeline

Buse, D. C., Scher, A. I., Dodick, D. W.,Reed, M. L., Fanning, K. M., Manack Adams, A. & Lipton, R. B.
(2016). Impact of migraine on the family: perspectives of people with migraine and their
spouse/domestic partner in the CaMEO study. Mayo Clinic Proceedings, 91(5), 596-611. doi:
10.1016/j.mayocp.2016.02.013

Cairns, E. (2018). Alder on a Promise, but it could be too late. Evaluate Vantage. Retrieved May 16, 2020,
from https://www.evaluate.com/vantage/articles/news/alder-promise-it-could-be-too-late

Calo-Fernández, B., & Martínez-Hurtado, J. L. (2012). Biosimilars: company strategies to capture value from
the biologics market. Pharmaceuticals (Basel, Switzerland), 5(12), 1393-1408.
doi:10.3390/ph5121393

Cammenga, J. (2019). State Corporate Income Tax Rates and Brackets for 2019. Tax Foundation. Retrieved
May 22, 2020, from https://taxfoundation.org/state-corporate-rates-brackets-2019/

Celgene. (2014). Form 10-K 2013. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

77
Celgene. (2015). Form 10-K 2014. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Celgene. (2016). Form 10-K 2015. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

Celgene. (2017). Form 10-K 2016. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

Celgene. (2018). Form 10-K 2017. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

Celgene. (2019). Form 10-K 2018. Retrieved from SEC EDGAR at

https://www.sec.gov/edgar/searchedgar/companysearch.html

Christiansen, A. T. (2018, November 26). Financial Statement Analysis – Valuation, Chapter 9 [PowerPoint
presentation]. Copenhagen Business School. Copenhagen.

Collan, M. (2011). Modeling Choices in the Valuation of Real Options: Reflections on Existing Models and
Some New Ideas. 15th Annual International Conference on Real Options Theory Meets Practice
[Conference paper]. Retrieved May 21, 2020, from http://www.realoptions.org/papers2010/106.pdf

Copeland, T. & Antikarov, V. (2003). Real options: A practitioner's guide (revised 1st ed.). London: TEXERE
Publishing.

Copeland, T. & Tufano, P. (2004). A Real-World Way to Manage Real Options. Harvard Business Review, 3.
Retrieved April 20, 2020, from https://hbr.org/2004/03/a-real-world-way-to-manage-real-options

Corrado, C. J. (1989). A nonparametric test for abnormal security-price performance in event studies. Journal
of Financial Economics, 23(2), 385-395. doi:0.1016/0304-405X(89)90064-0

Cowan, R. P. (2018). What Might the Migraine Landscape Look Like Post-CGRP? [Editorial talk].
Practical Neurology. Retrieved May 17, 2020, from https://practicalneurology.com/articles/2018-
feb/what-might-the-migraine-landscape-look-like-post-cgrp

Cowan, R., Cohen, J.M., Rosenman, E. et al. (2019) Physician and patient preferences for dosing options in
migraine prevention. Journal of Headache and Pain 20(50). doi:10.1186/s10194-019-0998-8

Cox, J. C., Ross, S. A. & Rubinstein, M. (1979). Option Pricing: A Simplified Approach. Journal of Financial
Economics, 7(3), 229-263. doi:10.1016/0304-405X(79)90015-1

Damodaran, Aswath. (n.d.). Estimating Risk Parameters. NYU Stern School of Business. Retrieved May 22,
2020, from http://people.stern.nyu.edu/adamodar/pdfiles/papers/beta.pdf

Damodaran, Aswath. (2005). The Promise and Peril of Real Options (July 2005). NYU Working Paper No. S-
DRP-05-02. Available at SSRN: https://ssrn.com/abstract=1295849

Datar, V. T., Matthews, S. H., & Johnson, B. (2007). A Practical Method for Valuing Real Options. Journal
of Applied Corporate Finance, 19(2), 95-104. doi:10.1111/j.1745-6622.2007.00140.x

Davitian, M., Fitzpatrick, H., Perkins, G., Breton, T., Denzler, R. & Giovanniello, D. (2018). Launch
Excellence: Once In A Life Cycle Opportunity [Whitepaper PowerPoint]. Health Advances.

78
 Retrieved May 21, 2020, from https://healthadvances.com/whitepaper/launch-excellence-life-cycle-
healthadvances.pdf

Deardorff, A., Baylan, J., Trzcinska, M. & Eskow, D. (2018). Approaches to Financial Valuation of
Biopharmaceutical Assets: Select Case Studies [PowerPoint]. Bluestar BioAdvisors. Retrieved May
15, 2020, from https://bluestarbioadvisors.com/resources/docs/bluestar-bioadvisors-valuation-case-
studies-2018-07.pdf

Dikmen, P. Y., Kosak, S., Aydinlar, E. I., & Kocaman, A. S. (2018). A single-center retrospective study of
onabotulinumtoxinA for treatment of 245 chronic migraine patients: survey results of a real-world
experience. Acta neurologica Belgica, 118(3), 475-484. doi:10.1007/s13760-018-0978-9

Dillon, P. & Hobbs, M. (n.d.). Untangling tax reform: business losses and NOLs for corporate and
noncorporate taxpayers [Website]. Baker Tilly. Retrieved May 22, 2020, from
https://www.bakertilly.com/insights/untangling-tax-reform-business-losses-and-nols-for-corporate-
and-noncorpora

DiMasi, J. A, Risks in new drug development: Approval success rates for investigational drugs. Clinical
Pharmacology & Therapeutics, 69(5), 297-307. doi:10.1067/mcp.2001.115446

DiMasi, J. A., Hansen, R. W. & Grabowski, H. G. (2003). The price of innovation: new estimates of drug
development costs. Journal of Health Economics, 22(2), 151-185. doi: 10.1016/S0167-
6296(02)00126-1

DiMasi, J. A., & Grabowski, H. (2007). The Cost of Biopharmaceutical R&D: Is Biotech Different?
Managerial and Decision Economics, 28(4/5), 469-479. Retrieved May 22, 2020, from
www.jstor.org/stable/25151529

DiMasi, J. A., Feldman, L., Seckler, A. & Wilson, A. (2010). Trends in Risks Associated With New Drug
Development: Success Rates for Investigational Drugs. Clinical Pharmacology & Therapeutics,
87(3), 272-277. doi: 10.1038/clpt.2009.295.

DiMasi, J. A., Grabowski, H. G. & Hansen, R. W. (2016). Innovation in the pharmaceutical industry: New
estimates of R&D costs. Journal of Health Economics, 47, 20-33. doi:10.1016/j.jhealeco.2016.01.012

Do, T. P., Guo, S. & Ashina, M. (2019). Therapeutic novelties in migraine: new drugs, new hope? The Journal
of Headache and Pain, 20(37). doi:10.1186/s10194-019-0974-3

Dunn, A. (2019). Why Alder thinks its late-comer migraine drug can succeed [Interview]. BioPharma Dive,
April 11. Retrieved May 21, 2020, from https://www.biopharmadive.com/news/why-alder-thinks-its-
late-comer-migraine-drug-can-succeed/552272/

Eli Lilly & Co. (2014). Form 10-K 2013. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Eli Lilly & Co. (2015). Form 10-K 2014. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Eli Lilly & Co. (2016). Form 10-K 2015. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Eli Lilly & Co. (2017). Form 10-K 2016. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

79
Eli Lilly & Co. (2018). Form 10-K 2017. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Eli Lilly & Co. (2019). Form 10-K 2018. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Eli Lilly & Co. (n.d.). How much should I expect to pay for Emgality? [Company website]. Retrieved May 15,
2020, from https://www.lillypricinginfo.com/emgality

EMA. (2020). Authorisation of medicines [Website]. Retrieved May 14, 2020, from
https://www.ema.europa.eu/en/about-us/what-we-do/authorisation-medicines

Eurostat. (2020, April 28). Population on 1st January by age, sex and type of projection [Data]. Retrieved May
21, 2020, from https://appsso.eurostat.ec.europa.eu/nui/show.do?dataset=proj_19np&lang=en

Fabozzi, F. J., Focardi, S. M. & Jonas, C. (2018). Equity Valuation Science, Art, or Craft? CFA Institute
Research Foundation Publications, 2017(4). Retrieved May 4, 2020, from
http://dx.doi.org/10.2139/ssrn.3254580

Fama, E. F. (1970). Efficient Capital Markets: A Review of Theory and Empirical Work. The Journal of
Finance, 25(2), 383-417. doi:10.2307/2325486

Flynn, T., Richter, S., Bradshaw, C., Park, E., Siddhanti, S. & Tang, K. (2017, April 19). The Hitchhiker’s
Guide to Biotech. Goldman Sachs Global Investment Research.

Fosheim, E. K. & Langerød, N. O. (2017). The use of Contingent Value Rights in M&A - An empirical review
[M.Sc. Thesis]. Norwegian School of Economics. Retrieved May 22, 2020, from
https://pdfs.semanticscholar.org/e3fb/39ddf6f201eb1c6a2baee80eefdf0556bfec.pdf

French, Craig W. (2003). The Treynor Capital Asset Pricing Model. Journal of Investment Management, 1(2),
60-72. Retrieved May 6, 2020, from https://ssrn.com/abstract=447580

Friedman, M. (1975). There’s No Such Thing As a Free Lunch. Open Court Publishing Company. ISBN-
10:087548297X

Gardner, J. (2019, February 7). Lilly hopes volume can make up for late migraine entry. Evaluate Vantage.
Retrieved May 16, 2020, from https://www.evaluate.com/vantage/articles/news/corporate-
strategy/lilly-hopes-volume-can-make-late-migraine-entry

GBD 2016 Disease and Injury Incidence and Prevalence Collaborators. (2017). Global, regional, and national
incidence, prevalence, and years lived with disability for 328 diseases and injuries for 195 countries,
1990-2016: a systemic analysis for the Global Burden of Disease Study 2016. The Lancet,
390(10100), 1211-1259. doi:10.1016/S0140-6736(17)32154-2

Gooch, J., Cordes, F., Bressau, F. & Berk, P. (2017). What Drives Biopharma Cost Performance. The Boston
Consulting Group, 7/17. Retrieved May 16, 2020, from https://image-src.bcg.com/Images/BCG-
What-Does_and-Does-Not_Drive-Biopharma-Cost-Performance-July-2017_tcm9-164210.pdf

Gottschalk, C. (2019, July 14). Christopher Gottschalk, MD: Advantages of Eptinezumab and CGRP Inhibitors
[Interview]. NeurologyLive. Retrieved May 21, 2020, from https://www.neurologylive.com/
conferences/ahs-2019/christopher-gottschalk-advantages-eptinezumab-cgrp-inhibitors

80
Haahtela, T. (2010, June 16-19). Recombining Trinomial Tree for Real Option Valuation with Changing
Volatility. 14th Annual International Conference on Real Options Theory Meets Practice [Conference
paper]. Boston, USA. Retrieved May 1, 2020, from http://www.realoptions.org/papers2010/241.pdf

Haahtela, T. (2012). Differences between financial options and real options. Lecture Notes in Management
Science, 4, 169-178. Retrieved May 5, 2020, from
https://orlabanalytics.ca/lnms/archive/v4/lnmsv4p169.pdf

Harrington, S. E. (2012). Cost of Capital for Pharmaceutical, Biotechnology, and Medical Device Firms. In
Danzon, P.M. & Nicholson, S. (Eds.), The Oxford Handbook of the Economics of the
Biopharmaceutical Industry, (pp. 75-99). NY: Oxford University Press

Hartmann, M. & Hassan, A. (2006). Application of real options analysis for pharmaceutical R&D project
valuation – Empirical results from a survey. Research Policy, (35)3, 343-354. doi:10.1016/j.respol
.2005.12.005

Hay, M., Thomas, D. W., Craighead, J. L., Economides, C. & Rosenthal, J. (2014). Clinical development
success rates for investigational drugs. Nature Biotechnology, 32(1), 40-51.
doi: https://doi.org/10.1038/nbt.2786. Retrieved May 12, 2020, from
http://diyhpl.us/~nmz787/pdf/Clinical_development_success_rates_for_investigational_drugs.pdf

Hay, M., Thomas D. W., Burns, J., Audette, J., Carroll, A. & Dow-Hygelund, C. (2016). Clinical Development
Success Rates 2006-2015. Retrieved from Biotechnology Innovation Organization (BIO) on May 12,
2020, via http://go.bio.org/rs/490-EHZ-999/images/Clinical%20Development%20Success
%20Rates%202006-2015%20-%20BIO%2C%20Biomedtracker%2C%20Amplion%202016.pdf
?_ga=2.132412813.2073857747.1589290020-1579906537.1589290020

Hepp, Z., Bloudek, L. M., & Varon, S. F. (2014). Systematic review of migraine prophylaxis adherence and
persistence. Journal of Managed Care Pharmacy, 20(1), 22-33. doi:10.18553/jmcp.2014.20.1.22

Hepp, Z., Dodick, D. W., Varon, S. F., Gillard, P., Hansen, R. N., & Devine, E. B. (2015). Adherence to oral
migraine-preventive medications among patients with chronic migraine. Cephalalgia : An
International Journal of Headache, 35(6), 478-488. doi:10.1177/0333102414547138

Hoffman, M. (2019, April 23). Eptinezumab Endeavors to Maintain Migraine Prevention Progress.
NeurologyLive. Retrieved May 14, 2020, from https://www.neurologylive.com/clinical-
focus/eptinezumab-endeavors-maintain-migraine-prevention-progress

Holford, N. H. G. & Sheiner, L. B. (1981). Understanding the Dose-Effect Relationship. Clinical

Pharmacokinetics, 6, 429-453. doi:10.2165/00003088-198106060-00002

Holmberg, E. & Jeppsson, H. (2010). The Use of Valuation Models by European Biotechnology Analysts
[Dissertation]. University of Gothenburg. Retrieved on April 30, 2020, from
https://gupea.ub.gu.se/bitstream/2077/28243/1/gupea_2077_28243_1.pdf

Hu, X. H., Markson, L. E., Lipton, R. B., Stewart, W. F., Berger, M. L. (1999). Burden of Migraine in the
United States: Disability and Economic Costs. Arch Intern Med, 159(8), 813–818.
doi:10.1001/archinte.159.8.813

Hull, J. C. (2018). Options, Futures and Other Derivatives (9th global ed.). London: Pearson Education.
ISBN-13:9781292212890

H. Lundbeck A/S. (2014). Annual Report 2013. Retrieved May 22, 2020, from

81
 https://investor.lundbeck.com/static-files/b0ce5802-706e-43a3-a9ff-f1106dfa7b47

H. Lundbeck A/S. (2015). Annual Report 2014. Retrieved May 22, 2020, from
https://investor.lundbeck.com/static-files/1d6c241c-5c62-4fc5-8e04-35a504bd011a

H. Lundbeck A/S. (2016). Annual Report 2015. Retrieved May 22, 2020, from
https://investor.lundbeck.com/static-files/0e582177-5655-47c9-b8db-5d4d3cd5facb

H. Lundbeck A/S. (2017). Annual Report 2016. Retrieved May 22, 2020, from
https://investor.lundbeck.com/static-files/8bfe19a3-e135-4cd4-a548-9a7f60702f00

H. Lundbeck A/S. (2018). Annual Report 2017. Retrieved May 22, 2020, from
https://investor.lundbeck.com/static-files/100bc1d0-679e-4953-8527-01536fa9727c

H. Lundbeck A/S. (2019). Annual Report 2018. Retrieved May 22, 2020, from
https://investor.lundbeck.com/static-files/fd57befc-4625-42f3-8003-e2c1082ca2c7

H. Lundbeck A/S, Investor Relations. (2018, October 25). Lundbeck updates on clinical phase III study for Lu
AF35700 in Treatment-Resistant Schizophrenia [Press release]. Retrieved May 26, 2020, from
https://investor.lundbeck.com/news-releases/news-release-details/lundbeck-updates-clinical-phase-
iii-study-lu-af35700-treatment

H. Lundbeck A/S, Investor Relations. (2019a, September 16). Lundbeck to acquire Alder
BioPharmaceuticals – a company committed to transforming migraine treatment and prevention – in
a transaction valued at up to USD 1.95 billion net of cash [Press release]. Retrieved May 11, 2020,
from https://investor.lundbeck.com/news-releases/news-release-details/lundbeck-acquire-alder-
biopharmaceuticals-company-committed

H. Lundbeck A/S, Investor Relations. (2019b, September 16). Lundbeck to acquire Alder BioPharmaceuticals
[Teleconference transcript]. Retrieved May 16, 2020, from https://investor.alderbio.com/static-
files/43181a9c-a49a-476c-a82e-37cdafcf16a1

H. Lundbeck A/S, Investor Relations. (2019c, October 22). Lundbeck completes the acquisition of Alder
BioPharmaceuticals – a company committed to transforming migraine treatment and prevention
[Press release]. Retrieved May 14, 2020, from https://investor.lundbeck.com/news-releases/news-
release-details/lundbeck-completes-acquisition-alder-biopharmaceuticals-company/

H. Lundbeck A/S, Investor Relation. (2019d, September 27). Amendment No.1 to Schedule TO. Retrieved May
26, 2020, from https://investor.alderbio.com/node/10876/html

Imam, S., Barker, R. & Clubb, C. (2008). The Use of Valuation Models by UK Investment Analysts,
European Accounting Review, 17(3), 503-535. doi:10.1080/09638180802016650

IMF – International Monetary Fund. (2020). World Economic Outlook (April 2020) [Data]. Retrieved May 21,
2020, from https://www.imf.org/external/datamapper/PCPIPCH@WEO/OEMDC/USA?year=2020

Institute for Clinical and Economic Review. (2018, June). A Look At CGRP Inhibitors For Migraine
Prevention [Report]. Retrieved May 16, 2020, from https://icer-review.org/wp-content/uploads/
2018/07/Migraine_RAAG_070318.pdf

Internal Revenue Service. (2019). RR-2019-23 – Applicable Federal Rates. Retrieved May 22, 2020, from
https://www.irs.gov/pub/irs-drop/rr-19-23.pdf

82
IQVIA. (2018). Medicine Use and Spending in the U.S. [Website]. Retrieved May 16, 2020, from
https://www.iqvia.com/insights/the-iqvia-institute/reports/medicine-use-and-spending-in-the-us-
review-of-2017-outlook-to-2022

Jaroszek, L. (2019, March 29). CFIVO1005U Empirical Finance: Event Studies [Lecture / PowerPoint].
Copenhagen Business School.

Kalyanaram, G. (2008). The order of entry effect in prescription (Rx) and over-the-counter (OTC)
pharmaceutical drugs. International Journal of Pharmaceutical and Healthcare Marketing, 2(1), 35-
46. doi:10.1108/17506120810865415

Kaplan, S. N. & Ruback, R. S. (1995). The Valuation of Cash Flow Forecasts: An Empirical Analysis. The
Journal of Finance, 50(4), 1059-1093. doi:10.1111/j.1540-6261.1995.tb04050.x

Kashoki, M., Hanaizi, Z., Yordanova, S., Veselý, Bouygues, C, Llinares, J. & Kweder, S. L. (2019). A
Comparison of EMA and FDA Decisions for New Drug Marketing Applications 2014–2016:
Concordance, Discordance, and Why. Clinical Pharmacology & Therapeutics, 107(1), 195-202.
doi:10.1002/cpt.1565

Kellogg, D. & Charnes, J. M. (2000). Real-Options Valuation for a Biotechnology Company. Financial
Analysts Journal, 56(3), 76-84. doi:10.2469/faj.v56.n3.2362

Kirman, I. & Goldfeld, V. (2011). Contingent Value Rights (CVRs). Wachtell, Lipton, Rosen & Katz. Retrieved
May 23, 2020, from http://www.wlrk.com/webdocs/wlrknew/AttorneyPubs/WLRK.18405.11.pdf

Kuick Research. (2014). Global Orphan Drug Market to Reach US$ 120 Billion by 2018 [Website]. Retrieved
May 17, 2020, from https://www.prnewswire.com/news-releases/global-orphan-drug-market-to-
reach-us-120-billion-by-2018-244195511.html

Lintner, J. (1965). The Valuation of Risk Assets and the Selection of Risky Investments in Stock Portfolios
and Capital Budgets. The Review of Economics and Statistics, 47(1), 13-37. doi:10.2307/1924119

Longstaff, F. A. & Schwartz, E. S. (2001). Valuing American Options by Simulation: A Simple Least-Squares
Approach. The Review of Financial Studies, 14(1), 113-147. Retrieved May 6, 2020, from
http://dse.univr.it/safe/Workshops/PhD/2003/Paper3-Schwartz.pdf

MacKinlay, A. (1997). Event Studies in Economics and Finance. Journal of Economic Literature, 35(1), 13-
39. Retrieved May 28, 2020, from www.jstor.org/stable/2729691

Maier, S. (2017). An Approach for Valuing Portfolios of Interdependent Real Options under both
Exogenous and Endogenous Uncertainties. 21th Annual International Conference on Real Options
[Conference paper]. Retrieved May 1, 2020, from http://www.realoptions.org/openconf2017
/data/papers/55.pdf

Maladwala, R (2017). Global migraine market set to be worth $8.7 Billion by 2026 [Report]. GlobalData.
Retrieved May 21, 2020, from https://www.globaldata.com/global-migraine-market-set-to-be-worth-
8-7-billion-by-2026-2/

Mavridis, T., Koniari, C., Fakas, N. & Mitsikostas, D. D. (2019). Anti-calcitonin gene-related peptide
monoclonal antibodies: Adverse effects. What do we really know? EMJ Innovation, 3(1), 64-74.
Retrieved May 21, 2020, from https://www.emjreviews.com/innovations/article/anti-calcitonin-
gene-related-peptide-monoclonal-antibodies-adverse-effects-what-do-we-really-know-a-literature-

83
 review/

McKinsey. (2019). Infographic: European biotech financing is maturing, but the gap with the United States
keeps growing [Website]. Retrieved May 18, 2020, from https://www.mckinsey.com/industries/
pharmaceuticals-and-medical-products/our-insights/infographic-european-biotech-financing-is-
maturing-but-the-gap-with-the-united-states-keeps-growing

Merton, R. C. (1973). Theory of Rational Option Pricing. Bell Journal of Economics and Management Science,
4(1), 141-183. doi:10.2307/3003143

Mossin, J. (1966). Equilibrium in a Capital Asset Market. Econometrica, 34(4), 768-783. doi:10.2307/1910098

Myers, S. C. (1984). Finance Theory and Financial Strategy. Interfaces, 14(1), 126-137. Retrieved May 13,
2020, from www.jstor.org/stable/25060526

Myers, S. and Shyam-Sunder, L. (1996). Measuring pharmaceutical industry risk and the cost of capital, in
Helms, R.B. (ed.) Competitive Strategies in the Pharmaceutical Industry, Washington, DC, American
Enterprise Institute Press

Myers, S. C. & Howe, C. D. (1997). A life-cycle financial model of pharmaceutical R&D. Program on the
Pharmaceutical Industry, Massachusetts Institute of Technology.

Neurocrine Biosciences. (2017). Form 10-K 2016. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Neurocrine Biosciences. (2018). Form 10-K 2017. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Neurocrine Biosciences. (2019). Form 10-K 2018. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Nordberg, M., Duffus, J., & Templeton, D. M. (2004). Glossary of terms used in toxicokinetics (IUPAC
Recommendations 2003), Pure and Applied Chemistry, 76(5), 1033-1082.
doi:10.1351/pac200476051033

Novartis. (2014). Annual Report 2013. Retrieved May 22, 2020, from
https://www.novartis.com/investors/financial-data/annual-results

Novartis. (2015). Annual Report 2014. Retrieved May 22, 2020, from
https://www.novartis.com/investors/financial-data/annual-results

Novartis. (2016). Annual Report 2015. Retrieved May 22, 2020, from
https://www.novartis.com/investors/financial-data/annual-results

Novartis. (2017). Annual Report 2016. Retrieved May 22, 2020, from
https://www.novartis.com/investors/financial-data/annual-results

Novartis. (2018). Annual Report 2017. Retrieved May 22, 2020, from
https://www.novartis.com/investors/financial-data/annual-results

Novartis. (2019). Annual Report 2018. Retrieved May 22, 2020, from
https://www.novartis.com/investors/financial-data/annual-results

84
PhRMA. (2019). Biopharmaceuticals in Perspective – Summer 2019. Retrieved May 19, 2020, from
https://www.phrma.org/-/media/Project/PhRMA/PhRMA-Org/PhRMA-Org/PDF/P-
R/PhRMA_2019_ChartPack_Final.pdf

Plumacher, M. (2018). Modeling of a binomial decision tree for real options [Dissertation]. Université Catholique
de Louvain & École Polytechnique de Luvain. Retrieved on May 3, 2020, from
https://dial.uclouvain.be/memoire/ucl/en/object/thesis%3A17169/datastream/PDF_01/view

Porter, M. E. (1979). How Competitive Forces Shape Strategy. Harvard Business Review, 57(2), 137–145.
Retrieved on May 12, 2020, from https://hbr.org/1979/03/how-competitive-forces-shape-strategy

Prescription Drug User Fee Act of 1992. 21 USC §§102-571.

Puran, S. (2005). The valuation of part-developed projects in the pharmaceutical sector [Dissertation].
University of Cambridge and Cambridge Healthcare & Biotech. Retrieved on April 30, 2020, from
http://www.chandb.com/resources/20050616_Report_SangeetaDissertation.pdf

Rendleman, R. J. & Bartter, B.J. (1979). Two-State Option Pricing. The Journal of Finance, 34(5), 1093-1110.
doi: 10.1111/j.1540-6261.1979.tb00058.x

Ressler, K. J., Mercer, K. B., Bradley, B., Jovanovic, T., Mahan, A., et al. (2011). Post-traumatic stress disorder
is associated with PACAP and the PAC1 receptor. Nature, 470(7335), 492-497.
doi:10.1038/nature09856

Roche Group. (2014). Annual Report 2013. Retrieved May 22, 2020, from
https://www.roche.com/investors/historic-quarterly-reporting.htm

Roche Group. (2015). Annual Report 2014. Retrieved May 22, 2020, from
https://www.roche.com/investors/historic-quarterly-reporting.htm

Roche Group. (2016). Annual Report 2015. Retrieved May 22, 2020, from
https://www.roche.com/investors/historic-quarterly-reporting.htm

Roche Group. (2017). Annual Report 2016. Retrieved May 22, 2020, from
https://www.roche.com/investors/historic-quarterly-reporting.htm

Roche Group. (2018). Annual Report 2017. Retrieved May 22, 2020, from
https://www.roche.com/investors/historic-quarterly-reporting.htm

Roche Group. (2019). Annual Report 2018. Retrieved May 22, 2020, from
https://www.roche.com/dam/jcr:af865dfd-50fb-458b-9cac-34097db9d3ec/en/ar18e.pdf

Rogers, J. (2002). Strategy, Value and Risk – The Real Options Approach. NY, New York: Palgrave
MacMillan. ISBN: 0333973461

Rubinstein, M. & Reiner, E. (1995). “Exotic Options” [Working paper]. University of California at Berkeley.

Sacco, S., Bendtsen, L., Ashina, M. et al. (2019). European headache federation guideline on the use of
monoclonal antibodies acting on the calcitonin gene related peptide or its receptor for migraine
prevention. The Journal of Headache and Pain, 20(6), doi:10.1186/s10194-018-0955-y

Saunders, M., Lewis, P. & Thornhill, A. (2016). Research Methods for Business Students (7th ed.). London:
Pearson Education. ISBN:978-1-292-01662-7

85
Schneider, M., Tejeda, M., Gabriel, D., Herzog, F., Keel, S. & Geering, H. (2008). Making real options work
for practitioners: a generic model for valuing R&D projects. R&D Management, 38(1), 85-106.
doi:10.1111/j.1467-9310.2007.00500.x

Scolnik, P. A. (2009). mAbs: a business perspective. MAbs, 1(2), 179‐184. doi:10.4161/mabs.1.2.7736

Scuteri, D., Adornetto, A., Rombolà, L., Naturale, M. D., Morrone, L. A., Bagetta, G., Tonin, P., & Corasaniti,
M. T. (2019). New Trends in Migraine Pharmacology: Targeting Calcitonin Gene-Related Peptide
(CGRP) With Monoclonal Antibodies. Frontiers in Pharmacology, 10(363).
doi:10.3389/fphar.2019.00363

Sharpe, W. (1964). Capital Asset Prices: A Theory of Market Equilibrium under Conditions of Risk. The
Journal of Finance, 19(3), 425-442. doi:10.2307/2977928

Sharpe, W. (1978). Investments (1st ed.). Upper Saddle River, NJ: Prentice Hall.

Smith, J. E. (2005). Alternative Approaches for Solving Real-Options Problems: (Comment on Brandão et al.
2005). Decision Analysis, (2)2, 89-102. doi:10.1287/deca.1050.0041

Stapleton, R., & Subrahmanyam, M. (1983). The Market Model and Capital Asset Pricing Theory: A Note.
The Journal of Finance, 38(5), 1637-1642. doi:10.2307/2327592

Teva Pharmaceuticals. (2014). Form 10-K 2013. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Teva Pharmaceuticals. (2015). Form 10-K 2014. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Teva Pharmaceuticals. (2016). Form 10-K 2015. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Teva Pharmaceuticals. (2017). Form 10-K 2016. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Teva Pharmaceuticals. (2018). Form 10-K 2017. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Teva Pharmaceuticals. (2019). Form 10-K 2018. Retrieved from SEC EDGAR at
https://www.sec.gov/edgar/searchedgar/companysearch.html

Teva Pharmaceuticals. (2020, April 27). Teva Announces U.S. Launch of Autoinjector for AJOVY®
(fremanezumab-vfrm) Injection [Press release]. Retrieved May 16, 2020, from
https://www.businesswire.com/news/home/20200427005083/en/Teva-Announces-U.S.-Launch-
Autoinjector-AJOVY%C2%AE-fremanezumab-vfrm

Titel 35 of the United States Code. 35 USC §365.

Trigeorgis, L. (1993). Real Options and Interactions with Financial Flexibility. Financial Management, 22(3),
202-224. Retrieved April 30, 2020, from https://econpapers.repec.org/RePEc:fma:fmanag:trigeorgis

Trigeorgis, L. (2006). Making use of real options simple: An overview and applications in flexible / modular
decision making. The Engineering Economist, 50(1), 25-53. doi:10.1080/00137910590917026

86
Urquhart, L. (2019, March 15). Lundbeck needs deals, but what can it afford? Evaluate Vantage. Retrieved
May 24, 2020, from https://www.evaluate.com/vantage/articles/interviews/lundbeck-needs-deals-
what-can-it-afford

U.S. Census Bureau. (2017). 2017 National Population Projections Tables: Main Series [Data]. Retrieved
May 21, 2020, from https://www.census.gov/data/tables/2017/demo/popproj/2017-summary-
tables.html

U.S. Food & Drug Administration. (2017, October 23). Biosimilar and Interchangeable Products [Website].
Retrieved May 17, 2020, from https://www.fda.gov/drugs/biosimilars/biosimilar-and-
interchangeable-products#biosimilar

U.S. Food & Drug Administration. (2018a, April 1). Step 3: Clinical Research [Website]. Retrieved May 18,
2020, from https://www.fda.gov/patients/drug-development-process/step-3-clinical-research

U.S. Food & Drug Administration. (2018b, April 1). Step 4: FDA Drug Review [Website]. Retrieved May 18,
2020, from https://www.fda.gov/patients/drug-development-process/step-4-fda-drug-review

U.S. Food & Drug Administration. (2018c, April 1). Step 5: FDA Post-Market Drug Safety Monitoring
[Website]. Retrieved May 14, 2020, from https://www.fda.gov/patients/drug-development-
process/step-5-fda-post-market-drug-safety-monitoring

U.S. Food & Drug Administration. (2018d, June 2). What Are "Biologics" Questions and Answers [Website].
Retrieved May 18, 2020, from https://www.fda.gov/about-fda/center-biologics-evaluation-and-
research-cber/what-are-biologics-questions-and-answers

Van Putten, A. B. & MacMillan, I. (2004). Making Real Options Really Work. Harvard Business Review, 12.
Retrieved April 30, 2020, from https://hbr.org/2004/12/making-real-options-really-work

Vogler, S. & Martikainen, J. (2014). Pharmaceutical Pricing in Europe. Springer. doi:10.1007/978-3-319-

12169-7_19.

Wessels, D., Goedhart, M. H., & Koller, T. (2015). Valuation : Measuring and Managing the Value of
Companies, University Edition: Vol. 6th ed. Wiley.

Young, A., Merle, E. & Yuan, D. (2017). U.S. Biotechnology – Sector Forecast – Amgen. Credit Suisse Equity
Research. Retrieved May 16, 2020, from https://plus.credit-suisse.com/rpc4/ravDocV
iew?docid=V7arzc2AF-YsVe

87
Appendix A-1: Unlevered Free Cash Flow Derivation

Unlevered Free Cash Flow Derivation

(Net) revenues

‒ Variable costs (Cost of goods sold, others)

= Gross profit

‒ Fixed costs (Personnel, selling, general & administrative, R&D, others)

= EBITDA

‒ Depreciation & amortisation

= Operating profit / EBIT

‒ Tax (on EBIT)

= NOPAT

+ Depreciation & amortisation

‒ Capital expenditures

+/‒ Change in net working capital

+/‒ Change in other non-cash items

+/‒ Decrease / increase in provisions

= Unlevered free cash flow

Source: Wessels, Goedhart and Koller (2015).

88
Appendix A-2: Exemplary Decision Tree of R&D Projects

Pharmaceutical Decision Tree (Illustration)

Source: Rogers (2002, p. 124, figure 12.8). All copyrights belong to the author.
Note: Post-approval stage assumes that a candidate achieves one of five cases.

89
Appendix B-1: Porter’s Five Forces Framework

How Competitive Forces Shape Strategy (Illustration)

Source: Porter (1979).

90
Appendix B-2: CGRP & PACAP Mechanism of Action

CGRP & PACAP Mechanism of Action (Illustration)

Source: Figure 3 from Do, Guo & Ashina (2019). All copyrights belong to the authors.

91
Appendix B-3: Standard Drug Sales Curve

Standard Sales Curve used for Extrapolation (Illustration)

Source: Figure 3.12 from Bogdan & Villiger (2010). All copyrights belong to the authors.
Note: X-axis denotes time since start of marketing in years. Y-axis denotes penetration as percentage of peak sales.

92
Appendix B-4: Quarterly Sales and Market Share Development of
CGRPs since Launch

Global Sales (Q1 2018 – Q2 2019)

200
Total Product Sales (USD millions)

Aimovig (Amgen / Novartis)

164.3
Emgality (Eli Lilly & Co.)
160
Ajovy (Teva Pharmaceuticals) 23.0

120 110.9 111.3 34.3

3.0
4.9 20.0
80 14.3

103.0 107.0
40 22.0 77.0
0.0 2.0 22.0
Q1 '18 Q2 '18 Q3 '18 Q4 '18 Q1 '19 Q2 '19

Market Share (Q1 ‘18 – Q2 ‘19) Geographic Split (Q1 ‘18 – Q2 ‘19)

100% 100%

75% 75%

50% 50%

25% 25%

0% 0%
Q2 '18 Q3 '18 Q4 '18 Q1 '19 Q2 '19 Q2 '18 Q3 '18 Q4 '18 Q1 '19 Q2 '19
Ajovy Emgality Aimovig United States Europe

Source: Quarterly and annual reports of Amgen / Novartis, Eli Lilly & Co. and Teva Pharmaceuticals. Retrieved via SEC EDGAR
and company websites.

93
Appendix B-5: Overview of Preventive CGRP Forecast (I/II)
Year 2018A 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E 2030E
Aimovig 119.0 482.0 818.0 1,194.7 1,540.0 1,833.6 1,999.0 2,165.6 2,284.6 2,356.0 2,379.8 2,379.8 2,356.0
% market share 93.8% 59.5% 50.2% 45.2% 40.0% 36.1% 33.7% 32.1% 30.9% 30.0% 29.2% 28.6% 28.1%
% growth 305.0% 69.7% 46.0% 28.9% 19.1% 9.0% 8.3% 5.5% 3.1% 1.0% — (1.0%)
% of peak sales 5.0% 20.3% 34.4% 50.2% 64.7% 77.1% 84.0% 91.0% 96.0% 99.0% 100.0% 100.0% 99.0%

Emgality 4.9 169.0 489.0 707.6 926.0 1,114.3 1,153.0 1,249.1 1,317.7 1,358.9 1,372.6 1,372.6 1,358.9
% market share 3.9% 20.9% 30.0% 26.8% 24.0% 21.9% 19.4% 18.5% 17.8% 17.3% 16.8% 16.5% 16.2%
% growth 3,349.0% 189.3% 44.7% 30.9% 20.3% 3.5% 8.3% 5.5% 3.1% 1.0% — (1.0%)
% of peak sales 0.4% 12.3% 35.6% 51.6% 67.5% 81.2% 84.0% 91.0% 96.0% 99.0% 100.0% 100.0% 99.0%

Ajovy 3.0 159.0 271.0 463.8 637.0 772.2 847.0 917.6 968.0 998.3 1,008.3 1,008.3 998.3
% market share 2.4% 19.6% 16.6% 17.6% 16.5% 15.2% 14.3% 13.6% 13.1% 12.7% 12.4% 12.1% 11.9%
% growth 5,200.0% 70.4% 71.1% 37.3% 21.2% 9.7% 8.3% 5.5% 3.1% 1.0% — (1.0%)
% of peak sales 0.3% 15.8% 26.9% 46.0% 63.2% 76.6% 84.0% 91.0% 96.0% 99.0% 100.0% 100.0% 99.0%

Eptinezumab — — 50.0 189.8 382.0 588.9 773.0 905.9 1,014.6 1,099.1 1,159.5 1,195.7 1,207.8
% market share 3.1% 7.2% 9.9% 11.6% 13.0% 13.4% 13.7% 14.0% 14.2% 14.4% 14.4%
% growth 279.7% 101.2% 54.2% 31.3% 17.2% 12.0% 8.3% 5.5% 3.1% 1.0%
% of peak sales 4.1% 15.7% 31.6% 48.8% 64.0% 75.0% 84.0% 91.0% 96.0% 99.0% 100.0%

Rimegepant — — — 85.8 326.0 617.6 875.0 1,098.0 1,286.8 1,441.2 1,561.3 1,647.1 1,698.5
% market share 3.2% 8.5% 12.2% 14.7% 16.3% 17.4% 18.4% 19.1% 19.8% 20.3%
% growth 280.0% 89.5% 41.7% 25.5% 17.2% 12.0% 8.3% 5.5% 3.1%
% of peak sales 5.0% 19.0% 36.0% 51.0% 64.0% 75.0% 84.0% 91.0% 96.0% 99.0%

Atogepant — — — — 40.0 152.0 288.0 408.0 512.0 600.0 672.0 728.0 768.0
% market share 1.0% 3.0% 4.9% 6.0% 6.9% 7.6% 8.2% 8.7% 9.2%
% growth 280.0% 89.5% 41.7% 25.5% 17.2% 12.0% 8.3% 5.5%
% of peak sales 5.0% 19.0% 36.0% 51.0% 64.0% 75.0% 84.0% 91.0% 96.0%
Total 126.9 810.0 1,628.0 2,641.7 3,851.0 5,078.6 5,935.0 6,744.1 7,383.6 7,853.4 8,153.5 8,331.5 8,387.4
% growth 538.3% 101.0% 62.3% 45.8% 31.9% 16.9% 13.6% 9.5% 6.4% 3.8% 2.2% 0.7%

Year 2031E 2032E 2033E 2034E 2035E 2036E 2037E 2038E 2039E 2040E 2041E 2042E 2043E
Aimovig 2,308.4 2,260.8 1,713.4 1,261.3 880.5 571.1 309.4 142.8 23.8 — — — —
% market share 27.7% 27.5% 24.1% 20.6% 17.5% 15.1% 12.1% 8.7% 2.6%
% growth (2.0%) (2.1%) (24.2%) (26.4%) (30.2%) (35.1%) (45.8%) (53.8%) (83.3%) (100.0%)
% of peak sales 97.0% 95.0% 72.0% 53.0% 37.0% 24.0% 13.0% 6.0% 1.0%

Emgality 1,331.4 1,304.0 988.3 727.5 507.9 329.4 178.4 82.4 13.7 — — — —
% market share 16.0% 15.8% 13.9% 11.9% 10.1% 8.7% 7.0% 5.0% 1.5%
% growth (2.0%) (2.1%) (24.2%) (26.4%) (30.2%) (35.1%) (45.8%) (53.8%) (83.3%) (100.0%)
% of peak sales 97.0% 95.0% 72.0% 53.0% 37.0% 24.0% 13.0% 6.0% 1.0%

Ajovy 978.1 957.9 726.0 534.4 373.1 242.0 131.1 60.5 10.1 — — — —
% market share 11.7% 11.6% 10.2% 8.7% 7.4% 6.4% 5.1% 3.7% 1.1%
% growth (2.0%) (2.1%) (24.2%) (26.4%) (30.2%) (35.1%) (45.8%) (53.8%) (83.3%) (100.0%)
% of peak sales 97.0% 95.0% 72.0% 53.0% 37.0% 24.0% 13.0% 6.0% 1.0%

Eptinezumab 1,207.8 1,195.7 1,171.6 1,147.4 869.6 640.1 446.9 289.9 157.0 72.5 12.1 — —
% market share 14.5% 14.5% 16.5% 18.7% 17.3% 16.9% 17.5% 17.7% 17.2% 14.9% 5.5%
% growth — (1.0%) (2.0%) (2.1%) (24.2%) (26.4%) (30.2%) (35.1%) (45.8%) (53.8%) (83.3%) (100.0%)
% of peak sales 100.0% 99.0% 97.0% 95.0% 72.0% 53.0% 37.0% 24.0% 13.0% 6.0% 1.0%

Rimegepant 1,715.7 1,715.7 1,698.5 1,664.2 1,629.9 1,235.3 909.3 634.8 411.8 223.0 102.9 17.2 —
% market share 20.6% 20.8% 23.9% 27.2% 32.4% 32.7% 35.6% 38.8% 45.1% 45.8% 47.0% 26.3%
% growth 1.0% — (1.0%) (2.0%) (2.1%) (24.2%) (26.4%) (30.2%) (35.1%) (45.8%) (53.8%) (83.3%) (100.0%)
% of peak sales 100.0% 100.0% 99.0% 97.0% 95.0% 72.0% 53.0% 37.0% 24.0% 13.0% 6.0% 1.0%

Atogepant 792.0 800.0 800.0 792.0 776.0 760.0 576.0 424.0 296.0 192.0 104.0 48.0 8.0
% market share 9.5% 9.7% 11.3% 12.9% 15.4% 20.1% 22.6% 25.9% 32.4% 39.4% 47.5% 73.7% 100.0%
% growth 3.1% 1.0% — (1.0%) (2.0%) (2.1%) (24.2%) (26.4%) (30.2%) (35.1%) (45.8%) (53.8%) (83.3%)
% of peak sales 99.0% 100.0% 100.0% 99.0% 97.0% 95.0% 72.0% 53.0% 37.0% 24.0% 13.0% 6.0% 1.0%
Total 8,333.4 8,234.1 7,097.8 6,126.8 5,037.0 3,778.0 2,551.1 1,634.3 912.4 487.5 219.0 65.2 8.0
% growth (0.6%) (1.2%) (13.8%) (13.7%) (17.8%) (25.0%) (32.5%) (35.9%) (44.2%) (46.6%) (55.1%) (70.3%) (87.7%)

Source: Quarterly and annual reports of Amgen / Novartis, Eli Lilly & Co. and Teva Pharmaceuticals. Retrieved via SEC EDGAR
and company websites. 2019, 2020, 2022 and 2024 revenue analyst consensus estimates as per EvaluatePharma (Gardner, 2019;
Armstrong, Brown & Fagg, 2019).
Note: Bold blue coloured numbers refers to analyst consensus input as per EvaluatePharma while non-bold blue numbers refer to
polynomial regression interpolation. Black numbers refer to extrapolation using standard sales curve. Light grey shading denotes last
available consensus estimate while dark grey shading indicates peak sales.

94
Appendix B-6: Overview of Preventive CGRP Forecast (II/II)

Aimovig (USD mn) Emgality (USD mn)

2,500 2,500

2,000 2,000

1,500 1,500

1,000 1,000

500 500

'18A '21E '24E '27E '30E '33E '36E '39E '42E '18A '21E '24E '27E '30E '33E '36E '39E '42E

Ajovy (USD mn) Eptinezumab (USD mn)

2,500 2,500

2,000 2,000

1,500 1,500

1,000 1,000

500 500

'18A '21E '24E '27E '30E '33E '36E '39E '42E '18A '21E '24E '27E '30E '33E '36E '39E '42E

Rimegepant (USD mn) Atogepant (USD mn)

2,500 2,500

2,000 2,000

1,500 1,500

1,000 1,000

500 500

'18A '21E '24E '27E '30E '33E '36E '39E '42E '18A '21E '24E '27E '30E '33E '36E '39E '42E
Quarterly and annual reports of Amgen / Novartis, Eli Lilly & Co. and Teva Pharmaceuticals. Retrieved via SEC EDGAR and
company websites. 2019, 2020, 2022 and 2024 revenue analyst consensus estimates as per EvaluatePharma (Gardner, 2019;
Armstrong, Brown & Fagg, 2019).
Note: Total global sales forecasts. All y-axes normed to USD 2,500 mn maximum.

95
Appendix B-7: Benchmarking Comparables

Median (L5Y)
Large-Cap (CNS) Peers % COGS of sales % S,G&A of sales % D&A of sales % NWC of sales % Capex of sales
I. AMGN - Amgen 18.1% 22.4% 9.2% 3.3% 3.1%
I. NOVN - Novartis 35.2% 29.3% 12.7% (1.4%) 6.5%
I. LLY - Eli Lilly & Co. 26.2% 30.8% 7.0% 0.1% 8.1%
I. TEVA - Teva Pharmaceutical Industries 47.6% 22.7% 7.0% (0.4%) 3.5%

I. BIIB - Biogen 12.9% 17.0% 7.1% 6.3% 6.4%

I. CELG - Celgene 7.6% 23.7% 4.3% 8.6% 8.6%
I. ABBV - AbbVie 22.7% 22.9% 4.6% 1.5% 4.2%
I. RO - Roche 32.0% 24.0% 7.0% 2.9% 2.9%
I. ALXN - Alexion Pharmaceuticals 8.9% 30.8% 9.8% 18.7% 10.1%
Median 22.7% 23.7% 7.0% 2.9% 6.4%

H. Lundbeck 26.1% 40.3% 8.6% (10.5%) 4.2%

Small- / Mid-Cap CNS Peers

II. NBIX - Neurocrine Biosciences NM NM 1.2% (4.9%) 4.9%
II. ACAD - Acadia Pharmaceuticals NM NM 2.2% (13.5%) 1.0%
II. ACOR - Acorda Therapeutics 20.7% 41.7% 3.0% 1.9% 1.9%
II. ALKS - Alkermes 17.7% 48.1% 12.6% 9.8% 5.9%
Median 19.2% 44.9% 2.6% (1.5%) 3.4%

Source: Quarterly and annual reports. Retrieved via SEC EDGAR and company websites.
Note: Last five years median values (where available). All selected companies are (mainly) active within mAbs / neurology.

96
Appendix B-8: Beta Regression Analysis

1 Year Before Announcement – Daily (N=252)

ALDR vs. NBI ALDR vs. SPX

y = 1.3921x - 0.0017 y = 1.828x - 0.0026

R² = 0.3233 R² = 0.2529

1 Year Before Announcement – Weekly (N=52)

ALDR vs. NBI ALDR vs. SPX

y = 1.2379x - 0.0103 y = 0.7448x - 0.0143

R² = 0.3921 R² = 0.0561

Since IPO in 2014 – Daily (N=1347)

ALDR vs. NBI ALDR vs. SPX

y = 1.4062x - 0.0004 y = 1.6377x - 0.0006

R² = 0.2518 R² = 0.1023

97
 Since IPO in 2014 – Weekly (N=279)

ALDR vs. NBI ALDR vs. SPX

y = 1.473x - 0.002 y = 1.0453x - 0.0024

R² = 0.2951 R² = 0.0389

Since IPO in 2014 – Monthly (N=63)

ALDR vs. NBI ALDR vs. SPX

y = 1.6357x - 0.0118 y = 0.254x - 0.0075

R² = 0.2912 R² = 0.002

Source: Raw data from Bloomberg.

Note: ALDR denotes the ticker of Alder. SPX denotes the ticker of the S&P500 index. NBI denotes the ticker of the NASDAQ
Biotechnology Index.

98
Appendix C-1: ALD1910 Development Project Timeline

FYE - 31st Dec 2018A 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E
Phase I
Phase II
Phase III
BLA
U.S. PDUFA decision
Patented marketing period
Post-patent period

Penetration Curve 5.0%

Transition Probabilities 100.0% 70.1% 100.0% 100.0% 38.1% 100.0% 100.0% 60.7% 86.8% 100.0%
Cash Flow Probabilities 100.0% 100.0% 70.1% 70.1% 70.1% 26.7% 26.7% 26.7% 16.2% 14.1%

FYE - 31st Dec 2029E 2030E 2031E 2032E 2033E 2034E 2035E 2036E 2037E 2038E 2039E
Phase I
Phase II
Phase III
BLA
U.S. PDUFA decision
Patented marketing period
Post-patent period

Penetration Curve 19.0% 36.0% 51.0% 64.0% 75.0% 84.0% 91.0% 96.0% 99.0% 100.0% 100.0%
Transition Probabilities 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
Cash Flow Probabilities 14.1% 14.1% 14.1% 14.1% 14.1% 14.1% 14.1% 14.1% 14.1% 14.1% 14.1%

FYE - 31st Dec 2040E 2041E 2042E 2043E 2044E 2045E 2046E 2047E 2048E 2049E
Phase I
Phase II
Phase III
BLA
U.S. PDUFA decision
Patented marketing period
Post-patent period

Penetration Curve 99.0% 97.0% 95.0% 72.0% 53.0% 37.0% 24.0% 13.0% 6.0% 1.0%
Transition Probabilities 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
Cash Flow Probabilities 14.1% 14.1% 14.1% 14.1% 14.1% 14.1% 14.1% 14.1% 14.1% 14.1%

Source: Author with reference to Bogdan & Villiger (2010).

Note: Transition probabilities as per table 4.

99
Appendix C-2: Epidemiology Model of ALD1910

A. Epidemiology Model Assumptions

U.S. CGRP Patient Populatation as per ICER Assessment (2018): U.S. Assumptions Europe Assumptions Other Assumptions

Migraine cases (m) 28.4 Diagnosis rate 40.00% Diagnosis rate 40.00% USD / EUR 1.1074
of which: episodic 26.0 Prescription start rate 65.33% Prescription start rate 65.33%
of which: chronic 2.4 Adherence / compliance rate 66.21% Adherence rate 66.21% Dosing schedule Monthly
Dosings per year 12.00
Eligible cases for preventive therapy (m) 12.5 CGRP responder rate 62.00% CGRP responder rate 62.00%
of which: episodic 10.1 CGRP non-responder rate 38.00% CGRP non-responder rate 38.00% Pricing Assumptions
of which: chronic 2.4 PACAP responder rate 47.70% PACAP responder rate 47.70%
US (USD) 575
Eligible cases for CGRPs (m) 5.6 U.S. inflation rate 1.76% Euro Area inflation rate 1.43% GER (EUR) 494
of which: episodic 4.5 List price per dosing (USD) 575.0 List price per dosing (EUR) 493.9
of which: chronic 1.1 Gross-to-net discount 27.00% Gross-to-net discount 27.00%
% of population 2018 1.71% Peak market share 42.00% Peak market share 42.00%

B.I. United States

Year 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E 2030E 2031E 2032E 2033E 2034E 2035E
U.S. population 328.2 329.8 331.3 332.9 334.4 336.0 337.5 339.0 340.6 342.1 343.6 345.2 346.7 348.2 349.7 351.2 352.7
% growth 0.5% 0.47% 0.47% 0.47% 0.46% 0.46% 0.46% 0.45% 0.45% 0.45% 0.45% 0.44% 0.44% 0.44% 0.43% 0.43% 0.43%

Eligible U.S. cases for CGRPs (m) 5.63 5.65 5.68 5.71 5.73 5.76 5.79 5.81 5.84 5.86 5.89 5.92 5.94 5.97 5.99 6.02 6.05

U.S. CGRP patient base 0.60 0.61 0.61 0.61 0.61 0.62 0.62 0.62 0.63 0.63 0.63 0.63 0.64 0.64 0.64 0.65 0.65
U.S. PACAP patient base 0.18 0.18 0.18 0.18 0.18 0.18 0.18 0.18 0.18 0.18 0.18 0.19 0.19 0.19 0.19 0.19 0.19

U.S. inflation rate 1.8% 1.8% 1.8% 1.8% 1.8% 1.8% 1.8% 1.8% 1.8% 1.8% 1.8% 1.8% 1.8% 1.8% 1.8% 1.8% 1.8%
List price per dosing 575.0 585.1 595.4 605.9 616.6 627.4 638.5 649.7 661.1 672.8 684.6 696.7 708.9 721.4 734.1 747.0

Net price per dosing 419.8 427.1 434.7 442.3 450.1 458.0 466.1 474.3 482.6 491.1 499.8 508.6 517.5 526.6 535.9 545.3

U.S. CGRP market size 3,055 3,123 3,193 3,264 3,337 3,411 3,487 3,564 3,643 3,724 3,806 3,891 3,976 4,064 4,153 4,245

U.S. PACAP market size 893 913 933 954 976 997 1,019 1,042 1,065 1,089 1,113 1,137 1,162 1,188 1,214 1,241

Penetration curve 5% 19% 36% 51% 64% 75% 84% 91%

U.S. revenue forecast 28 108 205 290 364 426 477 517

U.S. peak sales

B.I. United States

Year 2036E 2037E 2038E 2039E 2040E 2041E 2042E 2043E 2044E 2045E 2046E 2047E 2048E 2049E
U.S. population 354.2 355.7 357.2 358.7 360.2 361.7 363.2 364.6 366.1 367.6 369.0 370.5 371.9 373.4
% growth 0.43% 0.42% 0.42% 0.42% 0.41% 0.41% 0.41% 0.41% 0.40% 0.40% 0.40% 0.39% 0.39% 0.39%

Eligible U.S. cases for CGRPs (m) 6.07 6.10 6.12 6.15 6.17 6.20 6.23 6.25 6.28 6.30 6.33 6.35 6.38 6.40

U.S. CGRP patient base 0.65 0.65 0.66 0.66 0.66 0.67 0.67 0.67 0.67 0.68 0.68 0.68 0.68 0.69
U.S. PACAP patient base 0.19 0.19 0.19 0.19 0.19 0.19 0.20 0.20 0.20 0.20 0.20 0.20 0.20 0.20

U.S. inflation rate 1.8% 1.8% 1.8% 1.8% 1.8% 1.8% 1.8% 1.8% 1.8% 1.8% 1.8% 1.8% 1.8% 1.8%
List price per dosing 760.2 773.5 787.1 801.0 815.1 829.4 844.0 858.9 874.0 889.4 905.0 921.0 937.2 953.7

Net price per dosing 554.9 564.7 574.6 584.7 595.0 605.5 616.2 627.0 638.0 649.3 660.7 672.3 684.1 696.2

U.S. CGRP market size 4,338 4,433 4,530 4,629 4,730 4,833 4,938 5,045 5,155 5,266 5,380 5,497 5,615 5,736

U.S. PACAP market size 1,268 1,296 1,324 1,353 1,383 1,413 1,444 1,475 1,507 1,540 1,573 1,607 1,642 1,677

Penetration curve 96% 99% 100% 100% 99% 97% 95% 72% 53% 37% 24% 13% 6% 1%

U.S. revenue forecast 546 563 568 568 563 551 540 409 301 210 136 74 34 6

U.S. peak sales 568.3

100
B.II. Euro Area

Year 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2026E 2027E 2028E 2029E 2030E 2031E 2032E 2033E 2034E 2035E
European Union population 446.8 447.7 447.9 448.2 448.4 448.6 449.3 448.9 449.0 449.1 449.1 449.1 449.0 448.9 448.8 448.7 448.2
% growth 0.19% 0.05% 0.06% 0.05% 0.05% 0.15% -0.09% 0.02% 0.01% 0.01% 0.01% -0.02% -0.02% -0.03% -0.04% -0.10%

Eligible EU cases for CGRPs (m) 7.66 7.67 7.68 7.68 7.69 7.69 7.70 7.70 7.70 7.70 7.70 7.70 7.70 7.70 7.69 7.69 7.68

EU CGRP patient base 0.82 0.82 0.82 0.82 0.82 0.82 0.83 0.83 0.83 0.83 0.83 0.83 0.83 0.83 0.83 0.83 0.82
EU PACAP patient base 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24

EU inflation rate 1.4% 1.4% 1.4% 1.4% 1.4% 1.4% 1.4% 1.4% 1.4% 1.4% 1.4% 1.4% 1.4% 1.4% 1.4% 1.4% 1.4%
List price per dosing 493.9 500.953 508.116 515.382 522.752 530.228 537.81 545.501 553.301 561.213 569.239 577.379 585.635 594.01 602.504 611.12

Net price per dosing 360.5 365.7 370.9 376.2 381.6 387.1 392.6 398.2 403.9 409.7 415.5 421.5 427.5 433.6 439.8 446.1

EU CGRP market size (EURm) 3,562 3,614 3,668 3,723 3,778 3,838 3,889 3,946 4,003 4,060 4,118 4,176 4,235 4,295 4,355 4,413

EU PACAP market size (EURm) 1,041 1,057 1,072 1,088 1,104 1,122 1,137 1,154 1,170 1,187 1,204 1,221 1,238 1,256 1,273 1,290

Penetration curve 5% 19% 36% 51% 64% 75% 84% 91%

EU revenue forecast (EURm) 29 109 206 292 366 429 481 521

EU peak sales (m)

USD / EUR forecast 1.1074 1.1074 1.1074 1.1074 1.1074 1.1074 1.1074 1.1074 1.1074 1.1074 1.1074 1.1074 1.1074 1.1074 1.1074 1.1074

EU revenue forecast (USDm) 32 120 228 323 406 475 532 577

U.S. + EU revenue forecast (USDm) 60 228 433 613 769 902 1010 1094

B.II. Euro Area

Year 2036E 2037E 2038E 2039E 2040E 2041E 2042E 2043E 2044E 2045E 2046E 2047E 2048E 2049E
European Union population 448.2 448.0 447.7 447.3 446.8 446.5 446.1 445.6 445.1 444.5 443.9 443.3 442.6 441.9
% growth 0.00% -0.06% -0.07% -0.08% -0.13% -0.05% -0.10% -0.11% -0.12% -0.13% -0.13% -0.14% -0.15% -0.16%

Eligible EU cases for CGRPs (m) 7.68 7.68 7.67 7.67 7.66 7.65 7.65 7.64 7.63 7.62 7.61 7.60 7.59 7.57

EU CGRP patient base 0.82 0.82 0.82 0.82 0.82 0.82 0.82 0.82 0.82 0.82 0.82 0.82 0.81 0.81
EU PACAP patient base 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24 0.24

EU inflation rate 1.4% 1.4% 1.4% 1.4% 1.4% 1.4% 1.4% 1.4% 1.4% 1.4% 1.4% 1.4% 1.4% 1.4%
List price per dosing 619.859 628.723 637.714 646.833 656.083 665.465 674.981 684.633 694.424 704.354 714.426 724.642 735.005 745.515

Net price per dosing 452.5 459.0 465.5 472.2 478.9 485.8 492.7 499.8 506.9 514.2 521.5 529.0 536.6 544.2

EU CGRP market size (EURm) 4,476 4,537 4,599 4,661 4,722 4,787 4,850 4,914 4,979 5,043 5,109 5,174 5,240 5,307

EU PACAP market size (EURm) 1,309 1,326 1,344 1,363 1,380 1,399 1,418 1,437 1,456 1,474 1,494 1,513 1,532 1,551

Penetration curve 96% 99% 100% 100% 99% 97% 95% 72% 53% 37% 24% 13% 6% 1%

EU revenue forecast (EURm) 549 567 572 572 567 555 544 412 303 212 137 74 34 6

EU peak sales (m) 572.3

USD / EUR forecast 1.1074 1.1074 1.1074 1.1074 1.1074 1.1074 1.1074 1.1074 1.1074 1.1074 1.1074 1.1074 1.1074 1.1074

EU revenue forecast (USDm) 608 627 634 634 627 615 602 456 336 235 152 82 38 6

U.S. + EU revenue forecast (USDm) 1154 1190 1202 1202 1190 1166 1142 866 637 445 289 156 72 12

Peak revenue (U.S. + EU) 1,202.1

Source: Author with reference to various sources (see main text for references).

101
Appendix D-1: EV-to-Equity Bridge: Treatment of Select Items

Relevant
Adjustments: Reasoning:
for Alder:
(Largely) a non-operating asset; redundant to pay extra
when acquiring a company as we immediately get the
Subtracting cash & cash-
liquid funds; assumption that cash is normally used to Yes
equivalents
repay outstanding debt in an acquisition as covenants
require this in most cases
Same reasoning as for cash and cash-equivalents;
Subtracting short-term investments
usually excluding restricted cash (Koller, Goedhart & Yes
and other cash-like items
Wessels, ch. 12)
Do not reflect ongoing future business as they are to be
Subtracting non-core assets held
sold or closed down; represent (somewhat) tangible No
for sale
liquidity (Koller, Goedhart & Wessels, ch. 12)
Represents a financial debt claim on the firm, i.e. to the
unlevered FCF used for computing the EV; if converted
Adding financial debt No
into shares, then exclusion but inclusion of shares
(Koller, Goedhart & Wessels, ch. 12)
Debt-like but off-balance sheet item; if possible, add-
back of implicit lease interest but often neglected as not
Adding operating lease liabilities Yes
enough information available (Koller, Goedhart &
Wessels, ch. 27)
Represents a financial debt claim on the firm, i.e. to the
unlevered FCFF used for computing the EV; if
Adding convertible senior notes Yes
converted into shares, then exclusion but inclusion of
shares (Koller, Goedhart & Wessels, ch. 12)
Represents a debt-like claim on the firm via (superior)
Adding convertible preferred stock dividend rights while lacking equity-like voting rights Yes
(Koller, Goedhart & Wessels, ch. 12)
To be added as unfunded pensions represent future
liabilities and thus a claim on the company; to be
Adding unfunded pension
reduced by the respective tax rate as pension Noa
obligations
contributions are tax deductible for firms; acquirer has
to pay for these (Koller, Goedhart & Wessels, ch. 27)
Source: Author with reference to Wessels, Goedhart and Koller (2015).
Note: Not exhaustive. Other potential adjustments exist but were not deemed relevant for Alder as of June (Q2) 2019.
a Alder offers its employees a defined contribution plan.

102
Appendix D-2: Dilution

Standalone @ Share Price of USD 10.06

I. Employee Stock Option Plans (ESOPs)

Tranche No. (m) Price (USD) In the money? Options exercised (m) Proceeds (USD mn) No. of shares (m)
2014 EIP & 2018 IAP - Stock Options 10.1 18.34 N — — —
2014 EIP - Performance-based Options 0.2 16.20 N — — —
2018 LTIP - RSUs 0.8 0.8
Total — — 0.8

II. Convertible Preferred Stock

Tranche No. (m) Price (USD) Face value (USD mn) Conversion rate Shares (if converted) In the money? Proceeds (USD mn) No. of shares (m)
Class A-1 0.73 137.88 100.0 10.0 7.3 N — —
Dividend 1 - June 2019 0.02 137.88 2.3 10.0 0.2 N — —
Dividend 2 - December 2019 0.02 137.88 2.6 10.0 0.2 N — —
Warrant - Shares 0.075 137.88 10.3 10.0 0.8 N — —
Total 0.84 115.2 8.4 — —

III. 2025 Convertible Senior Notes, net

Tranche Principal (USD mn) Converison rate Shares (if converted) Conversion price (USD) Effective rate (%) In the money? Proceeds (USD mn) No. of shares (m)
2025 250.0 49.3827 12.35 20.25 10.6% N — —
Exercised over-allotment 37.5 49.3827 1.85 20.25 10.6% N — —
Total 287.5 98.8 14.2 — —

Acquisition Case
I. Employee Stock Option Plans (ESOPs)

Tranche No. (m) Strike price (USD) In the money? Options exercised (m) Proceeds (USD mn) No. of new shares (m)
2014 EIP & 2018 IAP - Stock Options 10.1 18.34 Y 10.1 184.57 10.1
2014 EIP - Performance-based Options 0.2 16.20 Y 0.2 2.43 0.2
2018 LTIP - RSUs 0.8 0.8
Total 10.2 187.0 11.0

II. Convertible Preferred Stock

Tranche No. (m) Price (USD) Face value (USD mn) Conversion rate Shares (if converted) In the money? Proceeds (USD mn) No. of shares (m)
Class A-1 0.73 137.88 100.0 10.0 7.3 Y — 7.3
Dividend 1 - June 2019 0.02 137.88 2.3 10.0 0.2 Y — 0.2
Dividend 2 - December 2019 0.02 137.88 2.6 10.0 0.2 Y — 0.2
Warrant - Shares 0.075 137.88 10.3 10.0 0.8 Y 10.3 0.8
Total 0.84 115.2 8.4 10.3 8.4

III. 2025 Convertible Senior Notes, net

Tranche Principal (USD mn) Converison rate Shares (if converted) Conversion price (USD) Effective rate (%) In the money? Proceeds (USD mn) No. of shares (m)
2025 250.0 49.3827 12.3 20.25 10.6% Y — 12.3
Exercised over-allotment 37.5 49.3827 1.9 20.25 10.6% Y — 1.9
Total 287.5 98.8 14.2 — 14.2

Source: Author with reference to Wessels, Goedhart and Koller (2015) based on Alder (2018) and Alder (2019c). Acquisition case
assumes fully diluted method.

103
Appendix E-1: Market Model Regressions

1 Year Before Announcement – Daily (N=252)

LUN vs. NBI LUN vs. OMXC25

y = 0.1422x - 9E-05 y = 0.7668x - 0.0002

R² = 0.015 R² = 0.1634

Source: Raw data from Bloomberg.

Note: LUN denotes the ticker of Lundbeck. OMXC25 denotes the ticker of the OMX Copenhagen 25 index. NBI denotes the
ticker of the NASDAQ Biotechnology Index. One outsized negative return of (26.6%) following a trial results release on the 25th
of October 2018 was deleted for Lundbeck to improve the parameter regressions.

104