Factor Investing

Pure and simple

Quantitative Equity Research

For professional clients

Date: November 2015

 Quantitative Equity Research
HSBC Pure Factor Indices

Smart beta investing has become increasingly popular

in recent years. Many index providers now offer a broad
suite of smart beta strategies. The most common tend
to be based on the well-established risk premia factors:
value, small cap, momentum, low volatility and quality.
However, HSBC’s research shows that many of these
indices exhibit unintended exposures to unrelated factors
because of their simplistic construction method. A more
sophisticated approach eliminates these unwanted risks,
providing a ‘pure’ factor index for smart beta investors.

The aim of this paper is to illustrate how HSBC’s

approach embeds an emphasis on:
u Precision
u Unbiasedness
u Robustness
u Efficiency

We investigate a method of measuring the efficiency

of smart beta indices based on the portion of active
risk driven by the targeted factor. We find that HSBC’s
pure beta indices exhibit higher factor efficiency than
commercially available alternatives.

We also compare HSBC’s indices with conventional

factor implementation to demonstrate the effectiveness
of HSBC’s construction method and the advantages of
factor neutralisation.

Finally, we discuss practical applications to portfolio

management and the value of HSBC’s indices as a tool
to enhance investment performance.

2 | HSBC Global Asset Management

Introduction to Factors Andrew Ang uses the following analogy to describe factors:
‘factors are to assets what nutrients are to food’. According
The appeal of smart beta indices is that they are systematic
to Ang, assets earn risk premia because they are exposed to
and transparent, and thus easy to construct and rebalance.
the underlying factor risks. Over time a growing proportion
They can also be a inexpensive way for investors to obtain
of investment performance has been explained in terms of
exposure to factors they might be lacking within their
factor exposures. Outperformance previously understood
portfolios. Many smart beta indices are constructed with
as ‘alpha’ is increasingly described as ‘beta’. Beta can come
an emphasis on simplicity, often using simple sorting and
from equity exposure, style, exotic factors, etc.
weighting techniques. These are usually based on either a
single factor (e.g. book-to-price) or a composite score (e.g. Historically, factor investing was considered an active
value). Other smart beta indices are put together to maximise strategy. Following the recent rise in investor demand for
investability, with factor tilts combined with market cap factor exposures, new cost efficient and highly accessible
weighting. Although both these approaches result in higher factor indices have been introduced by index providers. This
exposures to the targeted factor, there is little restriction on new dimension in product design has opened up a set of
exposures to other factors. This can lead to unintended factor opportunities designed to maximise convenience for investors.
exposure and taking on undesired risk.

Factor investing has become a topic of interest as it helps

answer a fundamental question: is the concept of diversification
still alive? The financial crisis saw the synchronised movement
of traditionally uncorrelated assets. Supposedly diversified
strategies proved to be less diversified than thought, leading
to dramatic underperformance.

Time

Alpha
Alpha
Alpha
Exotic Beta
Style
Alpha
Beta
Style
Beta
Equity Beta
Equity Beta
Equity Beta

HSBC Global Asset Management | 3

Theory behind factor based-investing
1970 CAPM The Capital Asset Pricing Model (CAPM) was first introduced in the 1960s
by Treynor, Sharpe, Lintner and Mossin. It was the first formal model to capture
Returns from a single
the notion of factors being the driving force behind returns. This one-factor
systematic risk
model implies that asset returns can be explained by just a single factor: the
sensitivity of the asset’s excess return to the excess return of the market. This
sensitivity is referred to as the beta of the asset. The intuition behind CAPM
is that the expected return of an asset, which is required to compensate for
its undiversifiable risk, should be a function of its correlated volatility with the
market (ß).

1976 APT Arbitrage Pricing Theory (APT) was first introduced in 1976 as an alternative
to CAPM and was one of the earliest multi-factor models. Its premise is that
Returns from multiple
expected returns can be decomposed into a linear combination of factors. These
sources
can be chosen either through economic intuition or through factor analysis
to identify the drivers of returns (a common method is principle components
analysis). The appeal of APT is that it imposes fewer assumptions and requires
less economic structure than CAPM.

1993 Fama-French One of the best known multi-factor models was introduced by Fama and French.
Using a 50-year dataset between 1941 and 1990, they found that the link
Value – Size
between market beta and average return had been weak. They proposed adding
two factors (size and book-to-market) to the single factor CAPM model to better
explain the cross-section of security returns.

1997 Cahart Building on Fama-French legacy Cahart extended the three factor model to
include a momentum factor. The addition of the MOM factor, as it is commonly
Value – Size –
known, improved the explanatory power of the model. Until recently was
Momentum
considered to be the reference evaluation framework for active management
and mutual funds.

2014 Frazzini et al. Recently Frazzini et al. introduced a quality factor (QMJ) and a low beta factor
(BAB). This followed the same methodology as Fama-French, extending further
Value – Size –
the range of potential valid factors. In addition Novy-Marx introduced a different
Momentum -
quality factor, claiming that it captures alpha.
BAB - QMJ

4 | HSBC Global Asset Management

Building Factor Indices A transparent and intuitive construction process

A plethora of factor index construction methods have been Objective: We try to give investors maximum exposure
proposed in the academic literature. Some have been to a factor, capturing as much of the premium as possible.
implemented by index providers. In this expanding ecosystem The Challenge: Unfortunately this isn’t enough if we want
of factor based products, there is a common misconception to focus on multiple ‘independent’ sources of risk/return.
that factor investing is very simple, providing superior Moving from the theoretical and impractical long-short
results to traditional funds (e.g. cap-weighted indices, active portfolios of Fama-French to long only investable solutions
management, strategic asset allocation). ‘Raw’ indices requires an understanding of the correlations and exposures
approach factor construction by overweighting stocks that between factors. There are three ways to tackle this:
exhibit a particular characteristic (e.g. Price-to-Book). To
u We could impose risk contribution constraints.
respond to the challenge of transforming academic risk
u We could apply a transformation algorithm such
factors into investable portfolios we focus on Precision,
Unbiasedness, Robustness and Efficiency. as ‘minimum-torsion’1 to approximate the closest
orthogonal (uncorrelated) factors.
Precise: The factors we seek exposure to are precisely
u We could apply neutralisation constraints to
defined, guided by empirical research.
unwanted factor exposures.
Unbiased: Our indices are constructed to remove hidden
bias towards untargeted factors. The first two approaches require parameterisation of the
factor model. This limits transparency when interpreting
Robust: Strong technological infrastructure, proprietary
individual stock factor exposures.
risk models and the conceptual clarity of our mathematical
formulation ensure robust implementation. The Solution: We take the third approach, following our
emphasis on transparency. We also incorporate an active
Efficient: Our indices deliver strong factor efficiency ratios,
weight constraint to improve diversification, a capacity
exhibiting a high proportion of targeted risk per unit of active risk.
constraint to avoid illiquid names and a turnover constraint
We refer to this family of indices as our “pure” indices. to control costs.

One of the challenges of factor investing is

determining which factors really drive returns.
Cochrane (2011) referred to a ‘zoo of new factors’ and
Harvey et al. (2014) counted over 300 factors, showing
a dramatic increase in recent years. In this ‘zoo’ it is
essential to focus only on factors that are strongly
supported by empirical evidence with solid economic
justifications. From this perspective the value, size,
momentum, low volatility and quality factors seem a
natural choice.

1
Minimum-torsion refers to a mathematical technique which applies a linear
transformation to the original factors in order to find the closest orthogonal
(uncorrelated) ones. For more information, see Meucci, Santangelo and
Deguest (2013) – Risk Budgeting and Diversification Based on Optimized
Uncorrelated Factors.

HSBC Global Asset Management | 5

 Why are turnover constraints important?
Control Turnover: A turnover constraint helps control costs and enhances portfolio stability. For example, momentum
strategies naturally exhibit high turnover. With no turnover constraint, momentum has ~300% average annual turnover,
imposing significant transaction costs on the portfolio.

Raw Momentum Monthly Turnover Pure Momentum Monthly Turnover

50% 50%

40% 40%

30% 30%

20% 20%

10% 10%

0% 0%
Jun-01
Jun-02
Jun-03
Jun-04
Jun-05
Jun-06
Jun-07
Jun-08
Jun-09
Jun-10
Jun-11
Jun-12
Jun-13
Jun-14
Jun-15
Jun-01
May-02
Apr-03
Mar-04
Feb-05
Jan-06
Dec-06
Nov-07
Oct-08
Sep-08
Aug-10
Jul-11
Jun-12
May-13
Apr-14
Mar-15
Oct-15

Sources: Factset, Thomson Reuters. October 2015.

2 Neutral: A focus on premia purity and approximate

Style independence
Active Factor from other sources of risk/return is essential to building factor
Exposures
efficient
1.5 indices. Factor neutralisation relative to the benchmark ensures low correlation with other styles and better risk adjusted
excess
1 returns (IR). Consider the active factor exposure of our pure momentum index against a simple raw momentum index:
Raw Momentum

0.5
Why
0 do we want pure factor beta?
-0.5 Momentum
Raw
-1
2 Active Factor Exposures
-1.5
1.5
-2
1
Raw Momentum

2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
0.5
0
-0.5
-1
-1.5
-2
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

Pure Momentum Index

Pure Momentum Index Pure Momentum Index

1
0.5
0
-0.5
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

Value Volatility Momentum Trading.activity Growth

1
Profitability Size Leverage Earnings.variability
0.5
Active exposures against MSCI World Index (MXWO). “Raw” style indices refer to the equally weighted first quintile of the desirable style.
0
Sources: Factset, Thomson Reuters, MSCI, IBES, Worldscope.
-0.5
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
6 | HSBC Global Asset Management

Value Volatility Momentum Trading.activity Growth

Raw momentum exhibits significant bias to small caps and
high volatility stocks. This unintended exposure could prove
problematic for performance and risk. HSBC’s pure momentum
index is by construction immunised from such exposure.

Furthermore, style neutrality translates into lower correlations

among factor excess returns:

Raw Factors

Size Volatility Quality Value Momentum

Size 100% -13% 62% 67% 21%

Volatility -13% 100% 9% -30% 41%

Quality 62% 9% 100% 40% 34%

Value 67% -30% 40% 100% -11%

Momentum 21% 41% 34% -11% 100%

Average raw pairwise correlation: 22%

Pure Factor Indices

Size Volatility Quality Value Momentum

Size 100% 29% 12% 41% 10%

Volatility 29% 100% 7% -24% 26%

Quality 12% 7% 100% 18% -3%

Value 41% -24% 18% 100% -2%

Momentum 10% 26% -3% -2% 100%

Average pure pairwise correlation: 11%.

Correlations of factor excess total returns over MSCI World (USD), monthly returns 07-2001 to 10-2015.
Sources: Factset, Thomson Reuters, MSCI, IBES, Worldscope.

Compared to raw factor implementation only size and

volatility seem to have a higher correlation, but this is still
within acceptable levels. As we will discuss later, correlations
follow time varying patterns so a static calculation reveals
little about their structure.

A recent paper from EDHEC (Amenc et al.) argues for the

importance of robustness in smart beta index construction.
‘Factor Fishing’, ‘Model-Mining’, ‘Non-Robust Weighting
Scheme’ and ‘Dependency on Individual Factor Exposures’
are common pitfalls to avoid.

HSBC Global Asset Management | 7

The risk of time-varying correlations
A popular factor blending approach is to combine value and
momentum. This is primarily because these factors exhibit
low correlations. However, in extreme circumstances, these
correlations can break down.

A raw factor implementation of value and momentum

depends on their correlation remaining small and stable.
This is often assumed to be constant and negative. The graph
below demonstrates that this is not the case - the correlation
varies over time, depending on the economic environment:

2 Years Rolling Correlation

1
0.8
0.6
0.4
0.2
0
-0.2
-0.4
-0.6
-0.8
-1
Jun-03 Jun-04 Jun-05 Jun-06 Jun-07 Jun-08 Jun-09 Jun-10 Jun-11 Jun-12 Jun-13 Jun-14 Jun-15

Pure Value-Momentum Index Raw Value-Momentum Index

Correlations calculated using daily excess (against MSCI World Index) total returns (2 years rolling) in USD from 04/06/2003 – 30/10/2015.
Sources: Factset, Thomson Reuters, MSCI, IBES, Worldscope.

In fact the 2 year historic correlation between value and

momentum is generally closer to zero, and more stable, for
pure factor indices. The only exception is the period around
the ‘quant crisis’ in 2007 when factor payoffs became
unstable, and even then the correlations were close.

The long only nature of the portfolio construction process

also impacts the combination of value and momentum.
Academic studies that refer to a consistent negative
correlation usually point to Fama-French factors based on
long-short portfolios incorporating illiquid securities.

To illustrate the importance of this effect, we now look at the

impact of this correlation instability in the period just after the
financial crisis – a period when equity markets rose rapidly.
This was a strong value driven rally, sustained for a number of
months - a good time to be exposed to the value factor.

8 | HSBC Global Asset Management

 A comment on manager diversification How can we measure the
A study from RVK showed that manager diversification
purity of a factor index?
(i.e increasing the number of funds in a multi-manager
portfolio basket) could potentially lead to negative effects. The inspiration for this section comes from a recently
As more managers are added to a portfolio: published paper by Hunstad and Dekhayser, where they
introduce a new measure called the factor efficiency
u Portfolio active share declines
ratio. As discussed above, most smart beta indices have
u Cost increases unintended exposures to untargeted factors. This usually
u There is minimal diversification benefit stems from the requirements of transparency, simplicity
or investability. There is evidence that simple minimum
Ultimately returns suffer:
variance optimisation, a common smart beta strategy,
results in time-varying factor exposures. Goldberg et al.
Median Seven-Year Return by Number
suggest that it is important to be aware of these exposures
of Managers in Portfolio
and highlight the benefits of targeting pure exposures when
10.5 building such indices.
Return (Annualised %)

10 The factor efficiency ratio is defined as the ratio of tracking

error from the desired factor(s) to the total tracking error. It
9.5 follows that this can be used to measure the efficiency of an
index, i.e. the ratio of desired to undesired active risk.
9.0
Formally, it is calculated as:
8.5
∑AR D
8 Factor Efficiency Ratio =
1 2 3 4 5 6 7 8 9 10 AR – ∑AR D
Number of Managers
Median Source: RVK, inc (2015) ∑AR D is the sum of active risk contributions of the desired
factors while AR is the total active risk of the portfolio.
Avoiding this problem requires a parsimonious approach
of building thematic blocks and identifying the point of
diminishing returns.

Typically additional managers are added to the roster to bring

complementary, uncorrelated exposures to the overall portfolio.
Our pure factor indices provide a useful set of tools to achieve
this. They are designed to represent independent sources of risk
and return at low cost. This provides the opportunity to control
overall factor exposure without affecting true ‘active’ share or
introducing new unwanted risk exposures.

HSBC Global Asset Management | 9

Decomposition of total active risk in The contributions to total active risk can be estimated using a
a cross-sectional factor risk model factor risk model.

Hunstad and Dekhayser highlight theFactor Efficiency

interesting Ratios as of March 2
disparity
1.60
between active exposure and factor efficiency. An index
Tracking Error
with
1.40 high exposure to a particular factor will not necessarily
have high factor efficiency. For example, it is well known
1.20
that a pure ranking of value stocks often has significant
Systematic Non-Factor 1.00
small-cap exposure. If we were to buy the top quintile of
Active Risk Risk value
0.80 names, we would anticipate a high exposure to both
value and small cap factors. We would prefer a pure beta
0.60
index to have a large proportion of active risk driven by the
Style 0.40 factor of interest and minimal active risk contributions
style
from
0.20 other factors.
In the chart below we show the factor efficiency ratios for
0.00
Industries
HSBC’s developed world pure beta indices
Momentum Quality against thoseSmall
for Cap
MSCI’s developed world style indices. (i.e MSCI Enhanced
Value, MSCI Momentum Tilt, MSCI Quality Tilt, MSCI
Country Volatility Tilt).

Factor efficiency ratios

Currency
Factor Efficiency Ratios as of March 2015
2.00
1.75
1.50
1.25
1.00
0.75
0.50
0.25
0.00
Momentum Quality Value Low Volatility
HSBC's Pure Indices MSCI

MSCI indices used: MSCI World Enhanced Value, MSCI World

Momentum Tilt, MSCI World Quality Tilt, MSCI Wold Volatility Tilt
(as of 31/10/2015)
Sources: Factset, Thomson Reuters, MSCI, IBES, Worldscope.

10 | HSBC Global Asset Management

HSBC’s Pure Indices

Total Active Proportion of Active Factor Active Exposure

Risk (%) Risk Contributed by Efficiency
relevant Factor (%) Ratio

Momentum 3.43 64.32 1.80 0.71

Quality 1.50 23.67 0.31 0.89

Value 2.72 39.21 0.65 1.10

Low Volatility 3.39 61.70 1.61 -0.63

MSCI’s Style Indices

Total Active Proportion of Active Factor Active Exposure

Risk (%) Risk Contributed by Efficiency
relevant Factor (%) Ratio

MSCI Momentum Tilt 1.88 47.43 0.90 0.29

MSCI Quality Tilt 0.84 9.25 0.10 0.27

MSCI Enhanced Value 3.99 18.91 0.23 0.92

MSCI Volatility Tilt 1.52 54.55 0.60 -0.26

Sources: Factset, Thomson Reuters, MSCI, IBES, Worldscope (as of 31/10/2015).

As can be seen, factor exposure does not directly correlate with factor efficiency. For example, HSBC’s momentum index has
a low factor exposure at 0.71 but a high factor efficiency of 1.80. This makes sense as momentum is a more volatile factor with
a higher active risk contribution. Looking at the desired factor’s exposures alone might be misleading. Factor exposures fail to
take into account the risks contributed by other potentially undesirable factors.

Concentrating on active risk contribution also connects back to the general debate on risk premia factors. There is a degree
of risk in investing in factors and their returns are time-varying. Note that indices with the same factor exposures may have
different active risks based on the nature of the factor. An index that is factor efficient has less contribution from undesired
risks. The key point is that we are only taking a risk on the factors that we choose to invest in.

Comparing this to the MSCI World Enhanced Value Index, we can see how different HSBC’s value index is in terms of factor
efficiency. The factor efficiency ratio of the MSCI index at the same point in time is 0.23, compared to HSBC’s of 0.65. This
implies that HSBC’s index takes on approximately 3 times as much value-related active risk per 1% of non-value active risk.

HSBC Global Asset Management | 11

The charts below are decompositions of active risk for the HSBC Pure Value Index and MSCI Enhanced Value Index as at end
of 10/2015. This is a common risk attribution output from portfolio attribution packages.

Decomposition of total active risk

HSBC Pure Value MSCI Enhanced Value

100% 100%
90% 90%
80% 80%
70% 70%
60% 60%
50% 50%
40% 40%
30% 30%
20% 20%
10% 10%
0% 0%
Style Countries Industries Currencies Non-Factor Style Countries Industries Currencies Non-Factor

Contribution Positive Contribution Negative Total

Decomposition of style active risk
HSBC Pure Value MSCI Enhanced Value

50% 50%

40% 40%

30% 30%

20% 20%

10% 10%

0% 0%
Volatility

Value

Trading Activity

Size

Profitability

Momentum

Leverage

Growth

Earnings Variability

Dividend Yield

Volatility

Value

Trading Activity

Size

Profitability

Momentum

Leverage

Growth

Earnings Variability

Dividend Yield

Contribution Positive Contribution Negative

Sources: Factset, Thomson Reuters, MSCI, IBES, Worldscope (as of 31/10/2015).

Looking at the decomposition of style active risk for the MSCI index above, we see that a significant portion of its active risk
comes from country active risk. Furthermore we can see that even though the biggest component of active risk is indeed
value, there are still significant contributions from size, growth and volatility. A closer look at the breakdown shows that the
majority of this comes from active exposure to Japan. This does not appear to be consistent with a simple ‘index’ strategy –
the ex-ante performance becomes too dependent on a single risk which is not immediately associated with the strategy.

12 | HSBC Global Asset Management

 Analysis: Value Index
Typically investors are most concerned when an alternative weighting scheme underperforms the capitalisation weighted
index. In the example below we show a 13m period where the value factor underperformed on a relative basis. One might
expect that main source of return will be the targeted factor but this is not always the case. For example, consider the two
portfolios shown below: the Pure Value Index and MSCI Enhanced Value Index. We conduct return attribution analysis which
reveals a clear connection between factor efficiency and realised style exposure. In the case of MSCI’s index, unintended style
4.0% (especially volatility and size) drove the negative style contribution
4.0%
exposures 4.0%to overall performance. However, our pure factor
4.0%
index avoided contamination from other styles which could have led to further
3.0% 3.0% underperformance.
3.0% 3.0%
2.0%
2.0% 2.0%
2.0%
Pure Value -1.51% MSCI Enhanced Value -2.54%
CTR to TR

CTR to TR
CTR to TR

CTR to TR
1.0%
1.0% 1.0%
1.0%
4.0%4.0% 4.0%4.0%
0%0% 0%0%
3.0%3.0% 3.0%3.0%
-1.0%
-1.0% -1.0%
-1.0%
2.0%2.0% 2.0%2.0%
-2.0% -2.0%
CTR to TR

CTR to TR
-2.0% -2.0%
CTR to TR

CTR to TR

1.0% 1.0% 1.0% 1.0%

-3.0%
-3.0% -3.0%
-3.0%
Market

Style

Countries

Industries

Currencies

Residual

Market

Style

Countries

Industries

Currencies

Residual
Market

Style

Countries

Industries

Currencies

Residual

Market

Style

Countries

Industries

Currencies

Residual
0% 0% 0% 0%
-1.0%
-1.0% -1.0%
-1.0%
-2.0%
-2.0% -2.0%
-2.0%
-3.0%
-3.0% -3.0%
-3.0%
Market

Style

Countries

Industries

Currencies

Residual

Market

Style

Countries

Industries

Currencies

Residual
Market

Style

Countries

Industries

Currencies

Residual

Market

Style

Countries

Industries

Currencies

Residual
Return
Return Sources
Sources ? ?
0.6%
0.6% 0.6%0.6%
0.4%
0.4% 0.4%0.4%
0.2%
0.2% 0.2%0.2%
0.0%
0.0% 0.0%0.0%
CTR to Style

CTR to Style
CTR to Style

CTR to Style

-0.2%
-0.2% -0.2%
-0.2%
-0.4%
-0.4% -0.4%
-0.4% Return Sources ?
Return Sources ?
-0.6%
-0.6% -0.6%
-0.6%
0.6% 0.6% 0.6% 0.6%
-0.8%
-0.8% -0.8%
-0.8%
0.4% 0.4% 0.4% 0.4%
-1.0%
-1.0% -1.0%
-1.0%
0.2% 0.2% 0.2% 0.2%
-1.2%
-1.2% -1.2%
-1.2%
Value

Value
Profitability

Trading.Activity

Size

Earnings.Variability

Volatility

Momentum
Leverage

Profitability

Trading.Activity

Size

Earnings.Variability

Volatility

Momentum
Leverage
Growth

Growth
Value

Value
Profitability

Trading.Activity

Size

Earnings.Variability

Volatility

Momentum
Leverage

Profitability

Trading.Activity

Size

Earnings.Variability

Volatility

Momentum
Leverage
Growth

Growth

0.0%0.0% 0.0%0.0%
CTR to Style

CTR to Style
CTR to Style

CTR to Style

-0.2% -0.2% -0.2% -0.2%

-0.4% -0.4% -0.4% -0.4%
-0.6% -0.6% -0.6% -0.6%
-0.8% -0.8% -0.8% -0.8%
-1.0% -1.0% -1.0% -1.0%
-1.2% -1.2% -1.2% -1.2%
Value

Profitability

Trading.Activity

Size

Earnings.Variability

Volatility

Momentum

Value

Earnings.Variability

Volatility

Momentum
Leverage

Profitability

Trading.Activity

Size

Leverage
Growth

Growth
Value

Profitability

Trading.Activity

Size

Earnings.Variability

Volatility

Momentum

Value

Volatility

Momentum
Leverage

Profitability

Trading.Activity

Size

Earnings.Variability

Leverage
Growth

Growth

These graphs analyse illustrative portfolios using our internal risk model and monthly weights. The charts show contribution (CTR) to total active return (TR)
against MSCI World, for portfolio/benchmark weights for 09/2014 to 10/2015.
Sources: Factset, Thomson Reuters, MSCI, IBES, Worldscope.

HSBC Global Asset Management | 13

 Analysis: Value + Momentum Investing
We have already illustrated the time varying relationship Consider the following two portfolios:
between value and momentum. The combination of value u Portfolio 1: 50% pure value index
and momentum as an investment strategy has been
+ 50% pure momentum index
well studied in academic literature and is popular among
u Portfolio 2: 50% raw value index
investment practitioners.
Tracking
Tracking
Error
Error + 50% raw momentum index
Decomposition
Decomposition
Tracking Error Portfolio
Portfolio
1: TE
Decomposition1: TE
1.88%
1.88%
Tracking
Tracking
Error
Error
Decomposition
Decomposition
Portfolio
Portfolio
2: TE
2: TE
1.89
1.89
%%
80%
80% 1: TE 2.08%
Portfolio 80%80%
Portfolio 2: TE 4.48%
Portfolio
Portfolio
1: TE
1: TE
1.88%
1.88% Portfolio
Portfolio
2: TE
2: TE
1.89
1.89
%%
80%
80%
60%
60% 80%
80%
60%
60%
TE to TE
TE to TE

TE to TE
TE to TE
60%
60%
40%
40% 60%
60%
40%
40%
PCR toPCR
PCR toPCR

PCR toPCR
PCR toPCR

40%
40%
20%
20% 40%
40%
20%
20%

20%
20%
0% 0% 20%
20%
0% 0%
MarketMarket
MarketMarket

Style Style
Style Style

Countries
Countries

Industries
Industries

Currencies
Currencies

NonFactor
NonFactor

MarketMarket
MarketMarket

Style Style
Style Style

Countries
Countries

Industries
Industries

Currencies
Currencies

NonFactor
NonFactor
0%0% 0%0%
Countries
Countries

Industries
Industries

Currencies
Currencies

NonFactor
NonFactor

Countries
Countries

Industries
Industries

Currencies
Currencies

NonFactor
NonFactor
Style components of Tracking ErrorStyle
Style
Component
Component
of Tracking
of Tracking
Error
Error
35%
35% 35%
35%
Style
Style
Component
Component
of Tracking
of Tracking
Error
Error
30%
30% 30%
30%
35%
35%
25%
25% 35%
35%
25%
25%
to Style
to Style

to Style
to Style

30%
30%
20%
20% 30%
30%
20%
20%
Unintended
Unintended
25%
25% 25%
25%
Style
Style

Style
Style

15%
15% 15%
15% exposure
exposure
PCR toPCR
PCR toPCR

PCR toPCR
PCR toPCR

20%
20%
10%
10% 20%
20%
10%
10% Unintended
Unintended
15%
15%
5%5% 15%
15%
5%5% exposure
exposure
10%
0%10%
0% 10%
0%10%
0%
5% 5%
-5%-5% 5% 5%
-5%-5%
Momentum
Momentum
Value Value
Value Value
Trading.Activity
Trading.Activity
Profitability
Profitability

Earnings.Variability
Earnings.Variability
Size Size
Size Size
Volatility
Volatility

DivYIdDivYId
DivYIdDivYId

Momentum
Momentum
Value Value
Value Value
Leverage
Leverage

Trading.Activity
Trading.Activity
Profitability
Profitability

Earnings.Variability
Earnings.Variability
Size Size
Size Size
Volatility
Volatility

DivYIdDivYId
DivYIdDivYId
Leverage
Leverage
GrowthGrowth
GrowthGrowth

GrowthGrowth
GrowthGrowth

0%0% 0%0%
-5%-5% -5%-5%
Momentum
Momentum

Trading.Activity
Trading.Activity
Profitability
Profitability

Earnings.Variability
Earnings.Variability

Volatility
Volatility

Momentum
Momentum

Earnings.Variability
Earnings.Variability

Volatility
Volatility
Leverage
Leverage

Trading.Activity
Trading.Activity
Profitability
Profitability

Leverage
Leverage

These graphs analyse illustrative portfolios using our internal risk model. The charts show percentage
contribution (PCR) to tracking error (TE) for the cross-section of portfolio/benchmark weights for 10/2015.
Sources: Factset, Thomson Reuters, MSCI, IBES, Worldscope.

From the charts above 37.25% of the tracking error for Portfolio 1 can be attributed to the targeted styles (value-momentum).
However, Portfolio 2 shows that the volatility factor contributed significantly to tracking error, much more than the targeted
styles. This is to be expected: as discussed before, raw momentum exhibits high exposure to volatility and size. Portfolio 1
appears to have a more balanced risk profile.

14 | HSBC Global Asset Management

Applications to Portfolio Management
Diversifying a Passive Holding
Factor tilts can be incorporated into portfolio management as an overlay to reduce risk, improve performance and enhance risk
adjusted returns.

In order to demonstrate this we consider three scenarios where different objectives require different factor tilts.

With the MSCI World Index as a base portfolio:

Case 1: Add 10% size tilt to improve performance

Case 2: Add 20% low vol index tilt to reduce risk
Case 3: Combine 10% size index and 20% low vol index tilt for better risk adjusted returns

Case 1: MXWO + 10% size tilt Case 2: MXWO + 10%low vol tilt Case 3: MXWO + 10%
size + 20% low vol tilt

Size Size Size

Profitability
Profitability Profitability

Momentum
Momentum Momentum

Volatility
Volatility Volatility

Value Value Value

-0.2 -0.15
-0.2 -0.15
-0.1 -0.05
-0.1
-0.2 -0.05
-0.15
0 0.05
-0.1
0 -0.2
0.05
-0.05-0.15
-0.2
0 -0.15
-0.1
0.05-0.05
-0.1
-0.2 -0.05
0 0.05
-0.15 0 -0.2
-0.1 0.05
-0.05 0 -0.1
-0.2
-0.15 0.05
-0.15 -0.1
-0.2-0.05
-0.05 0-0.15 0
-0.10.05
0.05 -0.05 0 0.05

Average active exposures (vs MXWO) for 07/2001 to 10/2015.

Annualised VolatilityVolatility
Annualized Annualised Return Return
Annualized Sharpe RatioRatio
Sharpe
15.90% 6.30% 0.40
6.20% 0.39
15.70%
6.10% 0.38
15.50% 6.00% 0.37

15.30% 5.90% 0.36

5.80% 0.35
15.10%
5.70% 0.34
14.90% 5.60% 0.33
MXWO

Size Tilt
MXWO + 10 %
Low Vol Tilt
MXWO + 20%
Size + 20% Vol Tilt
MXWO + 10%

MXWO

Size Tilt
MXWO + 10 %
Low Vol Tilt
MXWO + 20%
Size + 20% Vol Tilt
MXWO + 10%

MXWO

Size Tilt
MXWO + 10 %
Low Vol Tilt
MXWO + 20%
Size + 20% Vol Tilt
MXWO + 10%

Annualised performance numbers from internal backtests covering 07/2001 to 10/2015.

Sources: Factset, Thomson Reuters, MSCI, IBES, Worldscope.

HSBC Global Asset Management | 15

 Completion of an Active Portfolio
Building portfolios with a bottom-up approach can sometimes result in a collection of securities that exhibit unwanted factor
exposures. It is important to manage these biases in order to improve a portfolio’s risk profile.

A value oriented portfolio is used as the base case in this section. Such a portfolio will exhibit a natural bias to the value
factor. After performing factor analysis we can identify any significant unintended negative exposure to momentum. Using the
relevant factor index we are able to mitigate the unwanted exposure to momentum and keep the factor profile of the portfolio
close to benchmark. The wealth curve below demonstrates how the momentum bias correction improves performance after
the financial crisis.

Portfolio 1 with unwanted Momentum Portfolio 2 with tilt correction – adding 10%
tilt - Active factor exposures (vs MXEF) Momentum tilt - Active factor exposures (vs MXEF)

Size

Profitability

Momentum

Volatility

Value

.2 -0.1 0 0.1 -0.3

0.2 -0.2
0.3 -0.1
0.4 00.5 0.1 0.2 0.3 0.4 0.5 -0.3 -0.2 -0.1 0 0.1 -0.3
0.2 -0.2
0.3 -0.1
0.4 00.5 0.1 0.2 0.3 0.4 0.5
2.0

1.5
Cumulative Returns

1.0

0.5
Aug-07

Dec-07

Apr-08

Aug-08

Dec-08

Apr-09

Aug-09

Dec-09

Apr-10

Aug-10

Dec-10

Apr-11

Aug-11

Dec-11

Apr-12

Aug-12

Dec-12

Apr-13

Aug-13

Dec-13

Apr-14

Aug-14

Dec-14

Apr-15

Portfolio 1 Portfolio 1 + Momentum Tilt

The charts above show active average exposures and the cumulative wealth curve of hypothetical strategies for illustrative purposes only.
Data period: 31/07/2007 – 30/04/2015.
Source: Factset, Thomson Reuters, MSCI, IBES, Worldscope, cumulative total returns net of trading cost.

16 | HSBC Global Asset Management

Fulfilling an Equity Exposure
Each factor has its own characteristics, responding differently to economic cycles and environments. Time varying volatility
and performance can cause discrepancies from expectations over short term horizons. According to Andrew Ang (2014)
factors are collections of ‘bad times’. By combining factor indices to create a multi-equity factor product it is possible to
overcome the cyclical behaviour of individual factors and improve risk adjusted returns. As explained before this approach is
potentially more robust using pure factor indices.

Defensive Scenario: 33% value,

33% quality and 33% low vol Tracking Error Annualised Volatility
Size
2.5%
Annualised Volatility
Size
Profitability
2.0%
2.5%
Profitability
Momentum 1.5%
2.0%

Momentum
Volatility 1.0%
1.5%

0.5%
1.0%
Volatility
Value
0%
0.5%
-0.6
Value -0.4 -0.2 0 0.2 0.4 Defensive Balanced Dynamic
0%
-0.6 -0.4 -0.2 0 0.2 0.4 Defensive Balanced Dynamic
Information Ratio
Information Ratio: MXWO
Size
Balanced Scenario: 20% value, 20% momentum, 1.5
20% quality, 20% size and 20% low vol
Profitability

1.0 Information Ratio: MXWO

Momentum
Size
1.5
Volatility
Profitability 0.5

Value 1.0
Momentum
0
-0.6 -0.4 -0.2 0 0.2 0.4 Defensive Balanced Dynamic
Volatility 0.5

Value Active Premium Annualised Return

Size
0
2.5%
-0.6 -0.4 -0.2 0 0.2 0.4 Defensive Balanced Dynamic
Profitability
2.0%
Dynamic Scenario: 33% value,
Momentum 1.5%
33% momentum and 33% size
Volatility 1.0%
Annualised Return
Size 0.5%
Value 2.5%
Profitability 0%
-0.6 -0.4 -0.2 0 0.2 0.4 2.0% Defensive Balanced Dynamic

Momentum 1.5%

Volatility 1.0%
Data period: 31/07/2001 – 31/10/2015. Graphs to the left show average
factor exposures and graphs above show annualised performance
0.5% for developed world strategies against MSCI World (MXWO).
information
Value Sources: Factset, Thomson Reuters, MSCI, IBES, Worldscope.
0%
-0.6 -0.4 -0.2 0 0.2 0.4 Defensive Balanced Dynamic

HSBC Global Asset Management | 17

Multi-Factor Performance (Backtest)
In this section we provide evidence that HSBC’s indices exhibit attractive performance characteristics in comparison to well-
known alternatives from a range of well-known providers.

In the following chart we compare HSBC’s multi-factor blend with FTSE Rafi and MSCI Min Vol.

Developed World FTSE

31/07/2001 to HSBC Pure Factor MSCI WORLD
RAFI MSCI Min Vol
31/10/2015 Balanced Index INDEX (MXWO)

Annual Returns 7.52% 5.76% 6.86% 7.69%

Annual Volatility 15.33% 15.71% 17.29% 11.23%

Sharpe Ratio 0.49 0.37 0.40 0.68

Maximum
52% 54% 57% 43%
Drawdown

Annual Relative
1.76% 1.10% 1.93% 0.93%
Returns

Tracking Error 1.70% 4.02% 7.08% 7.07%

Information Ratio 1.03 0.27 0.27 0.13

The above analysis uses total returns from monthly Bloomberg data for MSCI World Index, FTSE RAFI and MSCI Min Vol
(Bloomberg tickers used: GDDUWI Index, TFRDU Index, M00IWO$T Index).

The HSBC Pure Factor Balanced Index data comes from our internal backtest, which uses monthly data to calculate total
returns with a trading cost of 20 bps on average each way at every rebalance

The following charts compare our individual developed world pure factor indices with the equivalent MSCI tilt products.
Our approach achieves competitive performance characteristics but also avoids unintended exposures by design.

Performance characteristics October 2015 active factor exposures against MXWO

MSCI
31/07/2001 to HSBC Pure
Enhanced
31/10/2015 Value Index
Value Size

Annual Returns 7.54% 8.49% Profitability Profita

Annual Volatility 17.23% 18.50%

Momentum Momen
Sharpe Ratio 0.44 0.46
Volatility Vola
Maximum
56.57% 58.24% Value V
Drawdown

Source: Factset, Thomson Reuters, MSCI, IBES, Worldscope, -1.0 -0.5 0 0.5 1 1.5
Bloomberg Finance LP.
MSCI Value Tilt HSBC Pure Value Index M

18 | HSBC Global Asset Management

 Volatility Volatility

Value Value

-1.0 -0.5 0 0.5 1 1.5 -1.0 -0.5 0 0.5 1 1.5

MSCI Value Tilt HSBC Pure Value Index MSCI Momentum Tilt HSBC Pure Momentum Index

Performance characteristics October 2015 active factor exposures against MXWO

HSBC MSCI
31/07/2001 to
Pure Low Volatility
31/10/2015
Size Size
Vol Index Tilt

Annual Returns
Profitability 10.60% 5.91% Profitability

Annual Volatility 11.22% 13.48%

Momentum Momentum

Sharpe Ratio 0.94 0.44

Volatility Volatility
Maximum Size
39.03% 49.75%
Value
Drawdown Value
Profitability
-1.0 -0.5 0 0.5 1 1.5 -1.0 -0.5 0 0.5 1 1.5

MSCI Quality Tilt HSBC Pure Quality Index MSCI Volatility Tilt
Momentum HSBC Pure Volatility Index
HSBC Pure MSCI
31/07/2001 to
Momentum Momentum Volatility
31/10/2015
Size Index Tilt Size

Annual Returns 6.84% 6.02% Value

Profitability Profitability

Annual Volatility 15.17% 14.83% -1.0 -0.5 0 0.5 1 1.5

Momentum Momentum
MSCI Value Tilt HSBC Pure Value Index
Sharpe Ratio 0.45 0.41
Volatility Volatility
Maximum
53.42% 52.48%
Value
Drawdown Value

-1.0 -0.5 0 0.5 1 1.5 -1.0 -0.5 0 0.5 1 1.5

MSCI Value Tilt HSBC Pure Value Index MSCI Momentum Tilt HSBC Pure Momentum Index
HSBC Pure MSCI
31/07/2001 to
Quality Quality
31/10/2015 Size
Index Tilt

Annual Returns 6.41% 5.99% Profitability

Annual Volatility 15.91% 14.84%

Momentum

Sharpe Ratio 0.40 0.40

Volatility
Size Size
Maximum
53.81% 50.52%
Drawdown Value
Profitability Profitability
Source: Factset, Thomson Reuters, MSCI, IBES, Worldscope, -1.0 -0.5 0 0.5 1 1.5
Momentum
Bloomberg Finance LP. Momentum
MSCI Quality Tilt HSBC Pure Quality Index

Volatility Volatility

The analysis above and to the left is based on monthly net total returns in USD for the MSCI indices, using Bloomberg data for
Value Value
M1WOEV Index, M1WOWMT Index, M1WOWQT Index and M1WOWVT Index.
-1.0HSBC Pure
The -0.5 Indices0data comes
0.5 from our
1 internal 1.5 -1.0
backtest, which uses monthly -0.5to calculate
data 0 total gross
0.5 dividend
1 1.5
returnsMSCI
with aQuality
tradingTilt
cost of 20 bps on
HSBC average
Pure Qualityeach
Indexway at every rebalance. We Volatility
MSCI have used Tiltnet total returns for other
HSBC Pure Volatility Index
providers instead of gross due to the unavailability of historical data before 2015 in most cases.

HSBC Global Asset Management | 19

Conclusion
Factor indices represent a highly accessible and efficient
way of investing in factor premia through passive vehicles.
HSBC’s approach to building factor indices focuses on
the integration of four distinct characteristics: Precision,
Unbiasedness, Robustness and Efficiency. Our methodology
incorporates factor neutralisation in order to improve premia
purity and turnover control for stability and cost reduction.
We compared the risk/return profile of HSBC’s suggested
methodology with the conventional (unconstrained) raw
factor implementations. Not controlling for exposures
to unwanted styles leads to the ‘contamination’ of a
signal’s purity and unintended risk exposure. Finally, we
demonstrated some practical applications of using factor
indices in portfolio management.

HSBC’s approach achieves competitive performance

characteristics avoiding, by design, unintended exposures.

20 | HSBC Global Asset Management

References
Amenc, N., Goltz, F., Lodh, A. and Sivasubramanian, S.
(October 2014): Robustness of Smart Beta Strategies, ERI
Scientific Beta Publication.

Ang, A., 2014, Asset Management: A Systematic Approach to

Factor Investing, Oxford University Press.

Fama, E.F. and French, K.R. (2013), A Four-Factor Model for

the Size, Value and Profitability Patterns in Stock Returns,
working paper.

Frazzini, A. and Pedersen, L.H. (2013) – Betting Against Beta,

Journal of Financial Economics, 111, 1-25.

Goldberg, Leshem and Geddes (2013) – Restoring Value to

Minimum Variance, forthcoming in Journal of Investment
Management.

Harvey, Campbell R., Yan Liu, and Heqing Zhu (2013): … and
the Cross-Section of Expected Returns, unpublished working
paper, Duke University.

Hunstad and Dekhayser (2014) – Evaluating the Efficiency of

‘Smart Beta’ Indexes, working paper.

Meucci, Santangelo and Deguest (2013) – Risk Budgeting

and Diversification Based on Optimized Uncorrelated Factors,
working paper.

John Cochrane of the University of Chicago coined the

term ‘zoo of factors’ in his 2011 presidential address to the
American Finance Association.

http://faculty.chicagobooth.edu/john.cochrane/research/
papers/discount_rates_jf.pdf

HSBC Global Asset Management | 21

Appendix
Our factor indices are constructed from the active universe for the relevant market cap weighted benchmark. The indices are
rebalanced monthly with an 8% turnover allowance (~100% per year).

Factor Construction
Each factor composite is constructed from several individual factors, which we describe in detail below.

In order to combine these components into one factor we first need to normalise them by subtracting the global mean and
dividing by the global standard deviation.

Xi - CapWeightedMean(X)
Zi =
StandardDeviation(X)
This procedure ensures that all the individual components are in the same scale and their combination results in the formation
of meaningful factors.

In addition, extreme normalised values that are outside the range [-3, 3] are set to -3 / 3.

The individual components of each factor are combined dynamically (ie the weights are not static) through a specialised algorithm
u At every point in time (cross-section) we calculate the Spearman rank correlation matrix of components and run a principal
components analysis (PCA)
u We extract the first principal component and normalise to sum to 100%. Unlike the equal weight approach, this captures
more of the information in individual components.

Factor Definitions

Value Size Momentum Volatility Quality

Book/Price Log (Market Total return over Rolling volatility - Income

Cap) 12 months, while Return volatility over ROE =
skipping the most the past 252 trading Book Value
recent two weeks days
to avoid the price
reversal effects

Earnings/Price Log (Sales) Momentum(T) Rolling CAPM beta Income

-2 - Rolling window ROCE =
Capital Employed
= ∑ log(1+rn,T-t )
regression of stock
returns on home
index returns
t=-54
((0.6*Earnings Log (Total Where rn,t total Historical sigma - Income
FY1) Assets) weekly return of residual volatility ROA =
+ security n at the from rolling window Total Assets
(0.4*Earnings week t regression of stock
FY2))/Price returns on home
index returns

EBITDA = EBITDA
Cash Flows from Cumulative range -
Operation/Price the ratio of maximum
and minimum stock Margin Sales
price over the
previous year (26%)

Log (Sales/EV)

EBITDA/EV

Source: GLOBAL EQUITY FUNDAMENTAL FACTOR MODEL, Nick Baturin, Sandhya Persad and Ercument Cahan September 2012, Version 1.1, Bloomberg

22 | HSBC Global Asset Management

Portfolio Optimisation

Target Function

The target function of the portfolio optimisation model drives the optimisation process.
Our target function is to maximise the Rank.
i=N
max w’Rank = max ∑
wiRi ,i ϵ U
i=1
Where Ri and wi is the Rank and weight of the stock i respectively and U the stocks universe.

Relative Factor Exposure Constraints

-0.01 ≤ ∑ Fi,zActive Weighti ≤ 0.01

iϵU
Where z represents each of the Risk factors (Value, Size, Volatility etc), Fi,z is the value of Factor z for stock i.

Relative Sector / Country Exposure Constraints

-5% ≤ ∑ Active Weighti ≤ 5%

iϵ J
Where J is each of the 10 GICS sector classifications (Financials, IT, Industrials etc.), and:

-5% ≤ ∑ Active Weighti ≤ 5%

iϵQ

Where Q is each countries within the universe.

Turnover Constraint

1
2
∑ w t - w t -1
i i
≤ 8%
iϵU
t
Where w i represents the weight of the stock i at time t and 8% the one way turnover.

Active Weight Constraint

Active Weighti ≤ 25bp

HSBC Global Asset Management | 23

For Professional Clients only and should not be distributed to or relied upon by Retail Clients.

The contents of this document are confidential and may not be reproduced or further distributed to any person or entity,
whether in whole or in part, for any purpose. The material contained herein is for information only and does not constitute
investment advice or a recommendation to any reader of this material to buy or sell investments. This document is not
intended for distribution to or use by any person or entity in any jurisdiction or country where such distribution or use would be
contrary to law or regulation. This document is not and should not be construed as an offer to sell or the solicitation of an offer
to purchase or subscribe to any investment.

HSBC Global Asset Management (UK) Limited has based this document on information obtained from sources it believes to be
reliable but which it has not independently verified. HSBC Global Asset Management (UK) Limited and HSBC Group accept no
responsibility as to its accuracy or completeness.

The views expressed above were held at the time of preparation and are subject to change without notice. Any forecast,
projection or target where provided is indicative only and is not guaranteed in any way. HSBC Global Asset Management (UK)
Limited accepts no liability for any failure to meet such forecast, projection or target.

The value of any investments and any income from them can go down as well as up and your client may not get back the
amount originally invested. Where overseas investments are held the arte of currency exchange may also cause the value of
such investments to fluctuate. Investments in emerging markets are by their nature higher risk and potentially more volatile
than those inherent in some established markets.

Stock market investments should be viewed as a medium to long term investment and should be held for at least five years.

Any performance information shown refers to the past should not be seen as an indication of future returns. It is important
to remember that these alternative indices do not outperform all the time. In particular in a momentum driven bubble (such
as with technology stocks in the late 90s) share prices can diverge from fair value for an extended period. In such cases
alternative index strategies will underperform capitalisation weighted indices as rebalancing does not improve returns.
However when the bubble bursts and share prices drop back towards fair value then alternative index strategies are more likely
to outperform.

Approved for issue in the UK by HSBC Global Asset Management (UK) Limited, who are authorised and regulated by the
Financial Conduct Authority.

Copyright © HSBC Global Asset Management (UK) Limited 2015. All rights reserved.

27506CP FP15-2018 EXP 31/07/2016.

24 | HSBC Global Asset Management