MGFD40 1 Winter 2025

Session Six Index

MGFD40 Winter 2025

Behavioural Finance
MGFD40

Session Six: 10-Feb-2025

UTSC Management Dept

Winter, 2025
 Coming Attractions

• Today
– Housekeeping
• Status of Essay Assignment One
• Quiz Six (10 min)

– AD Ch 6,9:
• Overconfidence – Psych Theory (30 min)
• Overconfidence – Applications in Finance (45 min)

MGFD40 4 Winter 2025

 Housekeeping – Deadlines

– Next week is Reading Week, no class

– Week 7, 24-Feb-24
• Review, Practice Q’s in class on Monday
• Joint Midterm on Wednesday, 5:00 pm
– Week 8, 04-Mar-24
• AD Ch 7: Effect of Emotion on Decisions
• AD Ch 10: Evidence of Emotion in Decisions

– Essay Assignment 2 due Wk 8 (03-Mar)

• Instructions on Quercus
MGFD40 5 Winter 2025
 Essay Assignment Two Topics

Assignment Two Essay Prompts

Discuss pros and cons of an opt-out plan for organ donations to an opt-in
A plan.
TS Chapter 13 offers a good introduction.
Analyze causes for student procrastination, such as time-inconsistent
B preferences, weak self-awareness, biases, or cognitive limitations. Suggest
corrective measures (nudges).
Analyze how biases, emotion or interpersonal differences can lead to
C
teams missing performance goals. Suggest corrective practices (nudges).

Discuss whether the disposition effect is caused by loss aversion or

D
conservatism in beliefs
Compare and contrast mental accounting to the sunk cost fallacy and
E opportunity costs. Use examples from personal finance or corporate
finance to illustrate the comparison.

F Propose a different topic and get approval for it by Week Six.

MGFD40 6 Winter 2025

 Article with nice list of biases

Shana Lebowitz and Allana Akhtar,

“60 cognitive biases that screw up everything we do”,
Business Insider epub, Oct 15, 2019.

https://www.businessinsider.com/cognitive-biases-2015-10

MGFD40 7 Winter 2025

Behavioural Finance
MGFD40

Segment 2: Overconfidence, Theory

UTSC Management Dept

Winter, 2025
 The Phenomenon of Overconfidence

• Overconfidence:
upwardly biased assessment of one’s own:
• precision of information,
• breadth of knowledge, or
• performance ability compared to others

MGFD40 9 Winter 2025

 Example similar to published research

• For each of six questions on the next slide,

estimate 90% confidence intervals (L, U):
• lower bound L:
choose L so that true value < L with prob. 5%

• upper bound U:
choose U so that U < true value with prob. 5%

MGFD40 10 Winter 2025

 The Phenomenon of Overconfidence

• Four Manifestations of Overconfidence

1. Mis-calibration of Precision
2. Better-Than-Average Effect
3. Excessive Optimism
4. Illusion of (unreasonable) Control

MGFD40 11 Winter 2025

 Illustrative Example

1. What is the average weight of the adult blue whale?

2. How fast does the earth spin at the equator (in kms per hour)?

3. How many independent countries were there at the end of 2023?

4. What is the distance, in air miles, from Delhi to Toronto?

5. How many bones are in the human body?

6. How long is the Amazon River (kms)?

MGFD40 12 Winter 2025

 Overprecision Example

• Answers to questions Within

Below bounds Above
1. adult blue whale 113,636 kgs (250,000 lbs) 4 2 26
2. speed of Earth 1,691 kph (1,044 mph) 10 10 12
3. indep countries 195 9 15 8
4. Paris to Sydney 17,080 kms (10,543 miles) 2 6 24
5. Bones in body 206 6 17 9
6. Amazon’s Length 6,480 kms (4,000 miles) 3 3 26

MGFD40 13 Winter 2025

 More on Miscalibration

– Alpert and Raiffa (1982)

• Ackert & Deaves has good description & summary

• Hard-Easy Effect
– Harder questions elicit strong overconfidence,
with less mass than normal in center
– Easier questions get more mass than normal in center,
but still with heavy (noisy) tails

• Look at our in-class exercise

– See a hard-easy effect?
* - M Alpert & H Raiffa, “A progress report on the training of probability assessors”, ch 21 (294-305) in
Kahneman, Slovic & Tversky (eds), Judgement under Uncertainty, Heuristics and Biases, Cambridge
Univ Press, 1982

MGFD40 14 Winter 2025

 2. The Better-than-Average Effect

• Myopic positive bias of own ability

– Biased memory of good and bad
• Self-attribution for positive outcomes
• Attribute (random) bad luck for negative outcomes

– Inattention blindness to others’ ability

– Recall Wk 1 survey of driving ability

MGFD40 15 Winter 2025

 2. The Better-than-Average Effect

• Sedikides et al, 2014 Br J of Soc Psych *

– 79 convicts in English prison
• Most (72%) convicted of violent crimes
• Anonymous survey administered in cafeteria
• Confidential interpersonal comparisons

* - C Sedikides, R Meek, MD Alicke, and S Taylor. “Behind bars but above the bar: Prisoners
consider themselves more prosocial than non-prisoners”, Br J of Soc Psych 53 (2014), 396-403.

MGFD40 16 Winter 2025

 2. The Better-than-Average Effect

• Sedikides et al, 2014 Br J of Soc Psych

– Assess nine characteristics:
moral kind to others
trustworthy honest
dependable compassionate
generous law-abiding
self-controlled

MGFD40 17 Winter 2025

 Sedikides et al, 2014 Br J of Soc Psych

– Mark on 11-point scale

- 5 = I am much less [trait] than the average prisoner
+ 5 = I am much more [trait] than the average prisoner

– Two forms
• Compare self to other prisoners

• Compare self to society in general

MGFD40 18 Winter 2025

 Sedikides et al, 2014 Br J of Soc Psych

Results

• Signif. better on avg than prisoners, outside comm.

• Only exception: equally law-abiding to community
MGFD40 19 Winter 2025
 3. Illusion of Control

• Langer (1975) * hypothesized that …

when chance situations mimic a skill situation,
• people behave as if they have control over the
uncontrollable event
• even when the fact that success or failure depends on
chance is salient
* - EJ Langer, “The Illusion of Control”, Journal of Personality and Social Psychology
32:2 (1975), 311-28.

MGFD40 20 Winter 2025

 3. Illusion of Control

– Examples in gambling
• prefer to throw dice (or give to lucky person)
• prefer to pick one’s own lottery numbers
• subjects name higher WTA price for lottery ticket
who may chose lottery numbers themselves rather
than randomly assigned

MGFD40 21 Winter 2025

 3. Illusion of Control

• Gino, Sharek & Moore (2011) *

– Circumstances matter:
• If events are very highly random, subjects
overestimate control (horse track, dice)
• If events are significantly controllable, subjects use
feedback rationally but biased to underestimate
• If events are largely controllable, subjects
underestimate control, overestimate randomness
F Gino, Z Sharek, D Moore, “Keeping the illusion of control under control: Ceilings, floors, and
imperfect calibration” Organizational Behavior and Human Decision Processes 114:2 (2011), 104-114.

MGFD40 22 Winter 2025

 4. Excessive Optimism

• People make practice of being positive

– Common Examples
• Overly optimistic planning horizons
• Projects more often over budget than under

– Business decisions benefit from science:

• Cold hard evidence and unbiased estimates,
• Care to statistical learning

MGFD40 23 Winter 2025

 Cautions on Overconfidence Hypotheses

• Alternative Explanations

– Various descriptors are not very distinct

• Miscalibration, BTAE, Control, Optimism
tend to show patterns of correlation
• More work required to label distinct effects apart
from bigger picture

MGFD40 24 Winter 2025

 Cautions on Overconfidence Hypotheses

• Alternative Explanations

– Miscalibration effect often replicated, but …

• lack of knowledge colors range of possibilities,
so “overconfidence” may reflect framing

• Hard-Easy Effect is like anchoring when

we have no prior beliefs

MGFD40 25 Winter 2025

 Cautions on Overconfidence Hypotheses

• Alternative Explanations
– Better-Than Average Effect
• Mistaken belief in ability or miscalculation?
– Not always clear how to distinguish BTAE from
excessive optimism or miscalibration

• Self-selected subjects may seek thrills, enjoy

games of skill, seek to test limits in competition
– Interesting to see if situations with large, real payoffs
yield same effects as “toy” lab games

MGFD40 26 Winter 2025

 Consistent Patterns in Memory

• Attribution Theory creates Overconfidence

– Hindsight bias:
perception of forecast accuracy improves with time
– Self-attribution:
see success as own effort, failure as random chance
– Confirmation bias:
welcome evidence consistent with prior beliefs,
ignore contrary evidence

MGFD40 27 Winter 2025

 Professional Learning

– Not only naïve subjects exhibit overconfidence

• Barber & Odean 1999 FAJ
– Evidence of overconfidence as halo of professional training
– Investment bankers, managers, doctors, lawyers

• Barber & Odean 2000 JoF

– Investors who trade the most perform the worst after fees

• Barber & Odean 2001 QJE

– Men, more prone to overconfidence than women
– Trade more and perform worse than women

MGFD40 28 Winter 2025

 One-minute Summary

– Consistent Patterns of Overconfidence

• Several observed patterns

• People make decisions with what looks like

overconfidence bias, but …
– Circumstance matter
– Descriptions overlap
– Comprehensive theory still to be developed

• The plural of anecdote is not evidence

MGFD40 29 Winter 2025

 One-minute Summary

– Consistent Patterns of Overconfidence

• How can patterns improve financial decisions?

– Professional development (culture of checking for bias)

– Personnel recruitment and assignment

– Contract negotiations

– Bidding behavior

MGFD40 30 Winter 2025

Behavioural Finance
MGFD40

Segment 3: Evidence on Overconfidence

UTSC Management Dept

Winter, 2025
 Research Examples of Overconfidence

– Camerer and Lovallo, AER (1999)

• Bus students play simulation
• Strong BTAE evidence

– Benartzi and Thaler, AER (Mar, 2001)

• Participants in DC benefit plans choose poorly
• Several aspects match heruistics

– Malmeinder and Tate, JFE (2008)

• CEO overconfidence linked to M&A outcomes

MGFD40 32 Winter 2025

 Ex 1: Better-than-Average Effect

• Camerer and Lovallo, 1999 AER

– Most start-up businesses fail
• 61.5 % of entrants exit within 5 yrs
• 79.6 % exit within 10 yrs
(1988 survey of US manufacturing)

– Why so many new entrants, given odds?

• Might be overconfidence

MGFD40 33 Winter 2025

 Camerer and Lovallo, 1999 AER

– MBA students play entrance game

• Depends on skill, can make money
– Some subjects assigned, some self-select

• N players choose simultaneously, and without

communicating, whether to enter a market or not
• Entrants split $50 by descending ranked skill
• If more enter than market capacity, excess low-skill
entrants lose $10 (the ante given to each at start)

MGFD40 34 Winter 2025

 Camerer and Lovallo, 1999 AER

– Payoffs by skill and capacity

c = Market Capacity

E = No. of Entrants

– If c < E, c entrants get paid, rest lose $10

MGFD40 35 Winter 2025

 Camerer and Lovallo, 1999 AER

– Skill ranking chosen two different ways

•R Random number drawing
•S Score on 10 logic puzzles

– Skill assessment occurs after games over

– Key comparison is number of entrants,

R condition versus S condition

MGFD40 36 Winter 2025

 Camerer and Lovallo, 1999 AER

– Procedures for each round

1. Announce whether R or S
2. Announce market capacity c
3. Subjects privately write estimate of E
(correct estimates each earn $0.25 later)
4. Subjects simultaneously choose to enter or not
5. Number of entrants tallied and announced

– Repeated for 12 R rounds, 12 S rounds

MGFD40 37 Winter 2025

 Camerer and Lovallo, 1999 AER

– Each round, players know:

• Number and character of players
• Capacity (c) for the round
• Whether round will be awarded by R or S

– Hypothesize that if overconfident,

• E higher in S than R
• Sub-optimal overall (industry) profit

MGFD40 38 Winter 2025

 Camerer and Lovallo, 1999 AER

• Market Capacities announced

MGFD40 39 Winter 2025

 Camerer and Lovallo, 1999 AER

• Subjects in eight experiments

MGFD40 40 Winter 2025

 Camerer and Lovallo, 1999 AER

• Shared profits is $50 per round

E < c entrants get windfall

c<E<c+5 both winners and losers

E=c+5 “industry” breaks even

c+5<E “industry” loses

MGFD40 41 Winter 2025

 Camerer and Lovallo, 1999 AER

• How to play the game optimally ?

– Entering dilutes earnings

– Least “skilled” entrants at risk of a loss

• Entrants ranked below top c entrants all lose $10
• If c < E, one or more entrant loses

– If highly skilled, enter S rounds

• If average skill, stick to R rounds

MGFD40 42 Winter 2025

 Camerer and Lovallo, 1999 AER

• Results
Compare profits:
R vs S
1-4 vs. 5-8

MGFD40 43 Winter 2025

 Camerer and Lovallo, 1999 AER

• Interpretation
For self-selected (sessions 5-8),
• almost all subjects write higher E estimates
• expect lower earnings in skill (S) than random (R)

• 85% of players cite neg. expected profit in S rounds

• large majority of subjects in self-selection sessions

(5-8) seem to think, “I expect the average entrant to
lose money, but not me!”

MGFD40 44 Winter 2025

 Camerer and Lovallo, 1999 AER

• Interpretation
– When payoffs depend on own abilities …
• individuals overestimate relative success,
• enter more frequently

– Overconfidence even stronger when subjects

self-select into experiment,
• knowing success will depend partly on skill
• seeming to ignore skill of reference group

MGFD40 45 Winter 2025

 Camerer and Lovallo, 1999 AER

• Applications
– Hostile takeover battles
• Hubris may explain why multiple firms enter

– Big-budget movies
• Many released on same weekends
(Memorial Day, 4th July)
• Movie studios might overlook reference group

MGFD40 46 Winter 2025

 Camerer and Lovallo, 1999 AER

• Testable Predictions
Should see more effect of overconfidence:
– when business outcomes more variable
(volatility and noise cloaks ineptitude)
– when success criteria more vague
(ambiguity permits excessive optimism)
– when pay is for performance

MGFD40 47 Winter 2025

 Camerer and Lovallo, 1999 AER

• Take-Home Message

– Try to check tendency to be overconfident

– Why should we be better than average?

– Beware reference group neglect

• Care to whether competition is self-selected
• Consider the effect in playoff tournaments

MGFD40 48 Winter 2025

 Example 2

Shlomo Benartzi and Richard Thaler,

“Naive Diversification Strategies in Defined Contribution

Saving Plans” Amer Econ Review 91,1 (2001), 79-98

MGFD40 49 Winter 2025

 Diversification Bias

– Benartzi and Thaler, AER (Mar, 2001)

• Noted trend to defined contribution saving plans
and privatized retirement requirements

• in USA: Defined Contribution vs. Defined Benefit

• In Canada we have SDRSP, TFSA plans

• Individuals must choose how to allocate savings

among fund alternatives

MGFD40 50 Winter 2025

 Benartzi and Thaler, AER (Mar, 2001)

• In DC plans, individuals choose how to

allocate savings among fund alternatives
– Main Question: how do people choose
retirement allocations?
– Especially focus on menu of funds
• Bond funds
• Stock funds
• Blended funds

– Are choices consistent with portfolio theory?

MGFD40 51 Winter 2025
 Diversification bias

• Simonson (1990)

– Asked undergrad students to pick snacks

• One treatment: pick three snacks all at once to be
consumed over three weekly class sessions.

• Alternate treatment: pick one snack each week, for

each of three class sessions

MGFD40 52 Winter 2025

 Benartzi and Thaler, AER (Mar, 2001)

– Read & Lowenstein:* children at Hallowe’en

• At single house, children may choose two items

from two piles of candy bars (Three Musketeers
and Milky Way)

• Two sequential houses, children may choose one

candy bar at each house from either pile

* - (J of Exper Psych: Applied (Mar 1995)

MGFD40 53 Winter 2025

 Benartzi and Thaler, AER (Mar, 2001)

• Similar results: diversification bias

Test Group Simultaneous Choice Sequential Choice

Children on 100% diversify 48% diversify
Hallowe’en
Undergrad 64% diversify 9% diversify
Students

MGFD40 54 Winter 2025

 Benartzi and Thaler, AER (Mar, 2001)

– Each case is puzzling

• Consumption depends on what is in the bag

• Sequential choice and simultaneous choice give

the same collection

• Why do people diversify when faced with menu of

choices?

MGFD40 55 Winter 2025

 Benartzi and Thaler, AER (Mar, 2001)

– Retirement accounts, same pattern?

• University employees to allocate between funds:
Treatment
Condition Alternative A Alternative B
1. Diversified Equity fund Diversified Bond Fund
2. Equity fund Balanced fund (50/50)
3. Balanced fund (50/50) Bond Fund

• Does menu of funds influence asset allocation?

MGFD40 56 Winter 2025

 Results

• Treatment 1, allocation and stock holding

– Modal choice: 50/50

MGFD40 57 Winter 2025

 Results, 2

• Treatment 2, equity vs. balanced fund

– Modal choice: STILL 50/50

– Causes shift to more equity on average
MGFD40 58 Winter 2025
 Results, 3

• Treatment 3, balanced fund vs. bond fund

– Modal choice: now more in bond fund

– Final asset allocation depends on menu

MGFD40 59 Winter 2025

 Benartzi and Thaler, AER (Mar, 2001)
Money Market Directories
proprietary database
• Look at equity in employer stock

– Note increase in equity allocation!

MGFD40 60 Winter 2025

 Benartzi and Thaler, AER (Mar, 2001)

• Conclusions
– Diversification heuristic seems apt
– People use simple rule of thumb to help make
complex decision: spread contributions evenly
across all investment options
• Do not seem to consider risk-reward tradeoff
• Selections are influenced by menu framing

MGFD40 61 Winter 2025

 Benartzi and Thaler, AER (Mar, 2001)

• Design of retirement saving plans

– Fund sponsors affect retirement choices
• More bond funds to stock funds, participants may
be too conservative
• Introduce more stock funds and participants shift
funds to be less conservative
• Neither is necessarily welfare-maximizing

MGFD40 62 Winter 2025

 Benartzi and Thaler, AER (Mar, 2001)

• Design of retirement saving plans

– Welfare costs may be substantial
– Matters to individuals
– Matters to fund sponsors, governments
• Beware participants making grave mismatch
between fund value and retirement needs

MGFD40 63 Winter 2025

 Overconfidence, Example 4

Ulrike Malmendier and Geoffrey Tate, JFE, 2008.

“Who Makes Acquisitions? CEO Overconfidence and the Market’s
Reaction.” Journal of Financial Economics 89 (1): 20–43.
https://doi.org/10.1016/j.jfineco.2007.07.002.

– Mergers should increase shareholder wealth but several studies

report opposite. What leads to value shrinking mergers?

MGFD40 64 Winter 2025

 CEO Hubris

• Malmendier & Tate, 2008 JFE

– Does CEO overconfidence cause hostile
takeovers?
Many managements apparently were overexposed in impressionable
childhood years to the story in which the imprisoned handsome prince
is released from a toad’s body by a kiss from a beautiful princess.

Consequently, they are certain their managerial kiss will do wonders for
the profitability of Company T[arget] . . . We’ve observed many kisses
but very few miracles. Nevertheless, many managerial princesses
remain serenely confident about the future potency of their kisses-even
after their corporate backyards are knee-deep in unresponsive toads.

Warren Buffet, 1981 annual report

MGFD40 65 Winter 2025

 Malmendier & Tate, 2008 JFE

• If CEOs act in interest of shareholders,

– mergers should increase shareholder wealth
– but several studies report opposite

– Moeller et al, 2005:

• of 4,136 acquisitions from 1998 to 2001,
• 87 lost market value of $1b or more
• returns distribution remarkably (negatively) skewed

MGFD40 66 Winter 2025

 Malmendier & Tate, 2008 JFE

FMV Gain or Loss,

Acquiring Firms (m’s $2001)

MGFD40 67 Winter 2025

 Malmendier & Tate, 2008 JFE

• Look at large US acquisitions

– studied 394 deals, 1980 to 1994

– Measure CEO confidence two ways:

1. Press clippings with key words
a. ‘‘confident’’ and ‘‘optimistic’’ vs.
b. ‘‘reliable’’ ‘‘cautious’’ ‘‘conservative’’‘‘practical’’ ‘‘frugal’’
or ‘‘steady’’

2. CEO stock option exercise behavior

MGFD40 68 Winter 2025

 Malmendier & Tate, 2008 JFE

• Hypotheses
– Overconfident CEOs more likely to:
• overestimate own ability to create business value
• overestimate ability to boost Target value

– Importance of capital
• If need to issue equity, CEO thwarted by
perceived “low” stock value
• Study picks firms with “abundant internal
resources”

MGFD40 69 Winter 2025

 Malmendier & Tate, 2008 JFE

• Prediction 1
– overconfident CEO & abundant internal
resources 
– more likely to conduct acquisitions than non-
overconfident CEOs

MGFD40 70 Winter 2025

 Malmendier & Tate, 2008 JFE

• Prediction 2
– If overconfident CEOs do more mergers than
rational CEOs, then …
– average value created in mergers is lower for
overconfident CEOs than for rational CEOs

MGFD40 71 Winter 2025

 Malmendier & Tate, 2008 JFE

• Results: Likelihood to complete a deal

MGFD40 72 Winter 2025

 Malmendier & Tate, 2008 JFE
• Results: Avge number of mergers

MGFD40 73 Winter 2025

 Malmendier & Tate, 2008 JFE
• Market Value Responses

– Overconfident CEOs do more deals …

• of lower average quality

– Market punishes “Longholder CEO” firms

MGFD40 74 Winter 2025

 Malmendier & Tate, 2008 JFE
• Market Response on Announcement

CAR’s
Firms [-1, +1]
Longholder CEO, Cash Bid  32 bp **
Longholder CEO, Stock Bid  135 bp ***
Other CEO, Cash Bid + 70 bp *
Other CEO, Stock Bid  75 bp ***
Overall Effect  29 bp *
*** 1% sig ** 5% sig * 10% sig

MGFD40 75 Winter 2025

 Malmendier & Tate, 2008 JFE
• Market Value destroyed by Longholders
– Longholder measure identifies 10.8% of
CEOs as overconfident …
– who cause 44% of value destruction around
merger bids
– and destroy $7.7 million more value than
other CEOs on avg
– lost total of $2.15b for acquiring shareholders

MGFD40 76 Winter 2025

Behavioural Finance
MGFD40

Segment 4: Session Six Wrap-Up

UTSC Management Dept

Winter, 2025
 Two-Min Summary

• Behavioral ideas now widely accepted

– Overconfidence has substantial impact

– Still, we have several persistent anecdotes

• Measures are not standardized
• Anecdotes suggest explanations
• Not yet a complete theory for choice

MGFD40 78 Winter 2025

 BREAK: End of Week 6

MGFD40 79 Winter 2025