Behavioral Finance

Self Deception
Self Deception
 Shows itself mainly as investors tending to
overconfidence
 They overestimate their knowledge and abilities,

underestimate risks, exaggerate ability to control

events and are often over optimistic
Facets of Overconfidence
 Forms of overconfidence
 Causes of overconfidence
 Factors impeding overconfidence
 Other forms of self deceptions
 How much do experts know?
 Untangling skill and luck in business
Forms of Overconfidence
 Miscalibration
 Betterthan average effect
 Optimistic bias
Miscalibration
 Investors tend to overestimate or mis-calibrate
precision of their knowledge
 e.g., if a person is asked a large number of
questions to estimate the 90% confidence interval
for known or knowable magnitudes, e.g., population
of Moscow, etc. then 90% responses should bracket
the right answer if the person is properly calibrated
 Or if same question is asked to a large number of
persons then 90% of responses should bracket the
right answer if they are properly calibrated
 In reality, much less than 90% of the responses do
Better Than Average Effect
 When people are asked to rate themselves on some
positive personal attribute, e.g., driving ability, they
tend to rate themselves above average
 Objectively, only 50% of people can be better than
average
 One reason may be that the definition of excellence
is not exact
 e.g., some may see best as being most adept at
avoiding accidents, some may see it as being most
skilful at speeding, and others may see it as
providing the smoothest ride
Better Than Average Effect
(Cont’d)
 From a cognitive standpoint the performance criteria
that most easily come to one’s mind are those that
one is best at
 However, from a motivational standpoint believing
that you are better than average enhances self-
esteem
Optimistic Bias
 Optimism is the default state and is embedded in
System 1
 Optimism bias arises from nature since it provides
evolutionary advantage in survival
 Optimists play an important role in our lives, e.g.,
inventors, entrepreneurs, political and military
leaders, etc., must be optimists
 Most risk takers tend to be disappointed but they still
seek out challenges and risks due to optimism
Overconfidence and Earnings
Manipulation
 Why do managers manipulate earnings when they
know it is unsustainable and self-destructive?
 Common scenario is that the company’s quarterly
earnings are not good so managers cook the books
by bringing forward some revenues from next
quarter to make the numbers
 They know this front loading will start the next
quarter with deficits but overconfidence persuades
them that revenues next quarter will be higher and
compensate
 This rarely happens and small frauds tend to balloon
over time
Causes of Overconfidence
 Illusion of knowledge
 Illusion of control
 Illusion of understanding
 Illusion of validity
 Illusion of skill
Illusion of Knowledge
 We believe that more information means more
knowledge and better forecasts
 This is not necessarily correct because we may not
be able to interpret information, some information
may be irrelevant or misleading and we tend to
interpret new information as confirming our prior
beliefs
 The investment industry is addicted to information
 But investors would be better off analysing only a
few things that they really need to know
Illusion of Control
 We tend to be more overconfident when we feel we
control outcomes
 Some participants were asked to toss a coin and bet
on the outcome
 Other participants were asked to bet on a coin that
was already tossed with the result concealed
 The first group bet more because they had an
illusion of control
 Factors that cause illusion of control are choice,
information, outcome sequence and task familiarity
Illusion of Control (Cont’d)
 Choice – choice induces sense of control. When
someone chooses his own lottery number he feels
he is more likely to win
 Information – greater information induces a sense
of control. New information causes us to emphasize
its importance but most information is really noise
 Outcome sequence – affects sense of control.
Early positive outcomes induce sense of control
compared to early negative outcomes.
 Task familiarity – creates greater sense of control
Illusion of Understanding
 A narrative fallacy is a flawed story of the past that
shapes future expectations
 They arise from our continuous attempt to make
sense of the world
 Explanatory stories assign a larger role to talent and
intentions rather than luck and focus on the few
striking events that happened
 But they do not focus on countless events that failed
to happen
 We fool ourselves by constructing flimsy stories that
nurture an illusion of understanding
Illusion of Validity
 Operating on the WYSIATI principle System 1 jumps
to conclusions with little evidence
 Amount and quality of evidence are made irrelevant
since poor evidence often makes a good story
Illusion of Skill
 Billions of shares are traded everyday because
buyers think prices will rise and sellers think prices
will fall
 Why does this happen despite both having same
information?
 Because they believe that they know more than the
market – this is the illusion of skill
 Researchers have investigated trading accounts
and found that on average stocks that were sold did
better than stocks bought
Illusion of Skill (Cont’d)
 So majority of investors would be better off following
a passive strategy
 Research has shown that those who trade the most
have the poorest results and those who trade least
have the best results
What Supports Illusions of
Skill and Validity?
 Most investors believe they can outperform
markets contrary to economic theory and
objective evaluations of their own results
 This is because investors pick stocks by doing

EIC analysis which convinces them that they

have high level skills and this induces illusions
 These illusions boost self esteem but cause

losses
Other Forms of Self Deception
 Avoidance of cognitive dissonance
 Self attribution bias
 Confirmation bias
 Hindsight bias
 Naïve realism
 Distorted self perceptions
Avoidance of Cognitive
Dissonance
 When newly acquired information contradicts pre-
existing understanding we experience mental
discomfort called cognitive dissonance
 Cognitions are attitudes, emotions, beliefs and
values and dissonance arises when contradictory
cognitions interact
 We dislike cognitive dissonance so we resort to
rationalizations to synchronize our cognitions to
maintain psychological stability
 This shows up in selective perception and selective
decision making
Avoidance of Cognitive
Dissonance (Cont’d)
 To minimize cognitive dissonance we resort to
following 3 strategies
 Change the conflicting belief so that it is congruent
with other beliefs
 Diminish the importance of the conflicting belief
 Emphasise more supportive beliefs that outweigh
the dissonant belief
Self Attribution Bias
 We tend to ascribe our success to our skill and
our failures to bad luck
 Self attribution is any perceptual process that is

distorted to maintain self esteem

 We focus on our strengths and achievements

and overlook our faults and failures

 We deny validity of negative feedback, take

credit for our team’s work and downplay the

contributions of others
 These behaviours protect self esteem
Confirmation Bias
 We tend to overlook information that is contrary
to our views in favour of information that
confirms them
 We are more likely to remember and repeat

confirming information which is why beliefs are

maintained even in the face of opposing
evidence
Hindsight Bias
 We have a tendency to view events as more
predictable than they really are especially in
hindsight
 This is the ‘I-knew-it-all-along’ phenomenon
Naïve Realism
 We think we see the world as it really is
 We also believe that everyone sees facts as we

see them and hence others should agree with us

 We commit selfish and short sighted acts but our

inner lawyer ensures that we do not blame

ourselves
 But we are quick to see bias, greed and duplicity

in others
Distorted Self Perceptions
 It is easier to spot a cheater when we look
outward but harder when we look inward
 In our thinking we generally take a position and

look for confirming evidence and if we do find

some we stop thinking about other evidence
How Much do the Experts
Know?
 Illusion of pundits
 Why forecasting thrives?
 Superiority of Formulae, Models or

Algorithms
 Robustness of Simple Algorithms
 Hostility of algorithms
 When can you trust expert intuition?
Illusion of Pundits
 Researchers asked experts who made a living from
giving advice on political and economic trends
 To assess probabilities that certain events would
occur in the near future
 In each case experts rated the probabilities of 3
alternatives: continuation of status quo, more of
something and less of that thing
 Experts performed worse than they would have if
they had considered each outcome equiprobable,
i.e., randomly picked outcomes
 This is because experts develop illusions of their
skill and become unrealistically overconfident
Why Forecasting Thrives?
 If forecasts are unreliable why produce them?
 Because demand creates supply
 When analysts are asked why they engage in the
futile exercise of publishing target prices they say it
is because clients want them
 Why do people use forecasts?
 Because when we are given a number we tend to
cling to it without even realizing it
 This is known as anchoring
Superiority of Formulae,
Models and Algorithms
 Many studies have shown that algorithms are more
accurate than experts at predictions
 This is because statistical models use only a few
criteria consistently and they are emotionless
 But experts use complex combinations
 This is especially true in domains which have high
uncertainty
Robustness of Simple
Algorithms
 Equal weighted formulas are superior to multiple
regression formulas
 This is because they are not affected by

sampling errors
 i.e., it is possible to develop useful algorithms

without statistical research

Hostility to Algorithms
 Forecasters say statistical evidence of their
inferiority does not correspond with experience
 The problem is that the experience of

forecasters may relate to short term predictions

and not long term long term predictions
When can you Trust Expert
Intuition?
 To understand when judgments reflect true expertise
and when they are illusions of validity we should
know the conditions for acquiring a skill
 Skill can be acquired when the domain is sufficiently
regular to be predictable and these regularities can
be learned through sustained practice
 When both these conditions are fulfilled intuitions are
likely to be skilled
 Chess is an example of a high validity domain which
can be learned through practice
 But stock pickers operate in zero validity
environments
What is Luck?
 Chance occurrence that affects a person favorably
or unfavorably and is beyond one’s control
 While randomness operates at system level luck
operates at the level of an individual
 If you benefit from good luck be happy and prepare
for luck to run out
 Provided you have approached the activity in the
correct fashion and you suffered from bad luck you
should continue in the same fashion in future
What is Skill?
 Ability to use knowledge effectively and readily in
execution
 It is not possible to discuss skill in a particular
activity without considering luck
 Some activities such as chess or running a race
depend almost wholly on skill
 Other activities such as roulette depend almost
wholly on luck
 In between are most other activities that depend on
skill and luck in varying proportions
What is Skill? (Cont’d)
 When skill exerts more influence there is intimate
connection between cause and effect
 When luck exerts more influence there is loose
connection between cause and effect in the short
run
 An activity involves skill if you can lose on purpose
Skill and Luck
 Different levels of skill and luck shape lives but we
are not good at distinguishing between them
because of psychological factors
 Intuitive judgments are unreliable because we base
predictions on how well events seem to fit a story
rather than consider how reliable the story is or what
happened before in similar situations
 Once something happens we are inclined to come
up with a cause to explain the event
 We commonly distort or ignore the role of luck in our
success or failure
Skill and Luck (Cont’d)
 History is written about winners because we like to
see cause and effect – luck is boring as the driving
force in a story
 As skill increases performance becomes more
consistent and luck becomes more important
 Paradox of skill is that if everyone in an activity is
uniformly more skillful then luck matters more
 Highly skilled investment managers have entered
investing in large numbers and it may no longer be
feasible for any of them to profit from the errors of
others to outperform market averages
Mean Reversion
 Mean reversion means that an outcome that is far
from average will be followed by an outcome that is
closer to average
 Position of an activity on the luck-skill continuum
determines the rate of mean reversion
 If activity depends only on luck there is complete
mean reversion but if it depends only on skill there is
no mean reversion
 The luck-skill continuum helps us understand when
luck makes the level of skill irrelevant
Luck-Skill Continuum
 Can you easily assign cause to effect? – If yes
the activity is likely on the skill side of the continuum
 What is the rate of mean reversion? – if slow then
skill is dominant otherwise luck is dominant
 How useful are expert predictions? – If expert
predictions are similar and accurate then skill is
dominant, otherwise luck is dominant
 Experts are usually inaccurate in predicting
economic, social and political systems because they
are complex adaptive systems whose outcomes
depend on interaction of lots of individual agents
that hides cause and effect
Untangling Skill and Luck in
Business
 Usual way to understand business success is to find
successful businesses and identify their common
practices
 This approach works if causality is clear
 But business performance depends on skill and luck
implying that a given strategy succeeds only
sometimes
 Attributing success to a given strategy may be naïve
since you are only looking at the sample of winners,
i.e., you are under-sampling failure
Untangling Skill and Luck in
Business (Cont’d)
 The relevant question is how many businesses that
tried a given strategy succeeded or failed
 i.e., we assume that the favourable outcome was
the result of skill and overlook luck
 As the industry matures all competitors become
efficient and excess returns dissipate
 Studies show there is clear evidence of mean
reversion in businesses which means luck
dominates
The End