Investor Psychology & Market Rationality Argument

Investor Psychology
& Market Rationality Argument

KIRAN KUMAR K V

Market Rationality
Argument

There
are certain suppositions upon which markets work. It assumes that human beings are rational, in fact, always rational. Their sole objective
is to maximize the return for a given level of risk that they are taking. The
concept of risk-adjusted return, that’s the foundation block of many asset
pricing and valuation models, highlights the risk-aversion behaviour of investors. Risk-aversion relates to the
behaviour of individuals under uncertainty. If an individual is offered two
options: one, where he will get Rs. 50 for sure and two, a gamble with a 50%
chance of getting Rs. 100 and another 50% chance of getting nothing. The
investor may react in three ways: He may choose the latter and gamble; He may
choose the former and be conservative; He may choose to be indifferent, as the
expected value before the bet in both the cases is Rs. 50
When
he chooses to gamble, he is referred to as risk-seeker. A risk-seeking or
risk-loving investor chooses uncertainty over certainty, as he gets extra ‘utility’ from the uncertainty associated
with the gamble. It can be observed in individual’s behaviors like buying a
lottery, or gambling in a casino or even sitting on a giant wheel. Risk-seekers
are also the ones who calculate the risk. Let’s consider in the above example,
if the latter option gave a 40% chance of getting Rs. 50 and 60% of nothing, a
risk-seeker may have avoided the gamble.
When
he chooses to be indifferent, the investor is referred to as risk-neutral.
He is an investor who cares only about the return. Uncertainty is not at all a
parameter of analysis for a risk-neutral investor. A risk-neutral behaviour can
be found when the investment at stake is insignificant part of their wealth.
If an investor chooses the
guaranteed income over gamble, he can be referred to as a risk-averse investor. He
will generally shy away from risky investments, even if the return is lower as
long as it is guaranteed. That does not mean he will not at all take risk. It
just means that he is not comfortable with the return he is getting for the
amount of risk he is assuming. He looks for a risk-return tradeoff that results
in that extra-utility for him. Because individuals are different in their
preferences, all risk-averse individuals may not rank investment alternatives
in the same manner. Take the example of Rs. 50 gamble; all risk-averse
individuals will rank the guaranteed outcome of Rs. 50 higher the option of
gambling the same. What would have happened if the guaranteed outcome was Rs.
40 and not Rs. 50? Some risk-averse investors might consider Rs. 40 inadequate,
others might accept it, and still others might become indifferent. This
suggests that individuals are risk-averse and they prefer more to less. They
are also able to rank different investment alternatives in order of their
preference and such ranking are internally consistent. Such utility function is
represented by the economist’s function:           
 

Where,
U is the utility function, E(r) is the expected return and

 is the variance of the investment and A is the measure of risk-aversion. It
can be computed as the marginal reward that an investor requires to accept a
unit of additional risk. A higher risk-averse investor requires greater
compensation for accepting additional risk (a higher A) and vice versa.

Apparent Irrationality

Below presented is the data of
the number of years when BSE Sensex has given a specific range of return
intervals. For example, two years, viz., 1991 and 2009 were the years when the
annual returns of Sensex were in the range of 75% to 96%.

Figure 1: Sensex Returns 1991-2015 (Source: Author’s
Calculation)
The
non-normal distributions of returns of Indian stocks in the past 25 years or so
show that there is a positive skewness. That also means there is higher
frequency of positive deviations from the mean. If the investors were rational
and always took the right decisions in terms of computing their risk-adjusted
returns, the distribution of returns must have been normal (assuming the
central limit theorem could be applied here, as Sensex in itself a portfolio of
30 companies and we are attempting to present returns for 25 years). The only
conclusion we can draw is that there is an anomaly here that is either
underestimating or overestimating the risk (as measured by the standard
deviation) by the investing community.
Investor Psychology as an Explanation

While the major blocks of investment finance,
like the mean-variance portfolio, CAPM, investor rationality and risk-return
trade-off have been the ingredients in valuation and security analysis models
widely used in real world, there is also an argumentative perspective to the
(above-discussed) basic tenet upon which such models are constructed.
Behavioural Finance questions the very assumption that people are guided by
reason and logic and independent judgment. It is also possible that investors
are not rational, at least always, and emotions and herd instincts might be
playing a role in influencing their investment decisions.  This approach attempts to enumerate
explanations why individuals make the decisions they do, irrespective of
rationality. Few such behavioural inconsistencies are discussed below to give
an idea of the approach:
–         
Representativeness
Bias
: A tendency to form judgments based on stereotypes. An
investor may see the effect of a policy rate change by the central bank based
on historical trend
–         
Anchoring
Bias
: A tendency to be unwilling to change an earlier
decision, despite a new, relevant information arrival
–         
Familiarity
Bias:
A tendency to be comfortable with things that are
familiar to them. Being familiar with employer or the industry they work in
people tend to invest in the known sectors.
–         
Affect
Heuristic Bias:
A tendency to go by the gut feel or intuition.
–         
Innumeracy
Bias:
A tendency to give more focus on big numbers and less
weight to small figures. Ignoring the base rate, wrongly counting the
probability of an event
–         
Loss
Aversion Bias:
A tendency of people to dislike losses more than they
like comparable gains. Many overreactions or panic selling in market we have
witnessed occur due to this bias. Especially, in countries like India, where
generally the retail investors are largely conservative in their risk appetite,
we see an increased overreaction
–         
Narrow
framing Bias:
A tendency of investors to focus on
issues/events/portfolios in isolation and respond based on hos such issues are
posed
–         
Mental
Accounting Bias:
A tendency of investors to keep track of gains and
losses for different investments in separate mental accounts and treat those
accounts differently
–         
Shadow
of the Past Bias:
A tendency of people to consider a past outcome as a
factor in evaluating a current risky decision. A Snake-bite effect refers to the behaviour where individuals being
less inclined to take risk after a incurring a loss.
–         
Herd
Instinct Bias:
A tendency to move with the group, even if one has a contradicting
solution
The
biases discussed above are just a few examples of what behavioural finance
approach assumes to be the reasons for markets being inefficient (i.e., market
prices always uncorrelated with any known variable). Behavioural finance thus
can be seen as a contrasting approach to the traditional investment finance
theory and the market efficiency theory. Behavioural finance seems to have
answers for irrationality of market prices in converging with the intrinsic
value.
Behavioural biases affect all
market participants. Now the question is how do rebuild the asset pricing or
valuation models to account for behavioural biases such that all the factors –
rational & irrational (behavioral) – that influence the price discovery in
security markets are accounted for and fully reflect the causations of given
market behaviour. If there can be a model that can do this, market participants
would be able to recognize, respond and make improved decisions, individually
and collectively.

References: Behaviour Finance, Chandra, 2016; Investopedia; IAPM, Reilly & Brown, 2014
Share this Post: Facebook Twitter Pinterest Google Plus StumbleUpon Reddit RSS Email

Related Posts

Leave a Comment