Efficiency vs. Behavioral Finance

Download as pptx, pdf, or txt
Download as pptx, pdf, or txt
You are on page 1of 60

FINA 363 FALL 2014

Efficient Markets
Behavioral Finance

Efficient Market Hypothesis (EMH)


Capital markets are huge computing devices

Inputs: information

Outputs: security prices

EMH Capital markets compute correct


prices on a timely basis
Why do we care?

Random Walk and The EMH


Observable time series of stock prices
look like a Random Walk (Bachelier,
1900)
Random Walk
DPt = a + et
a = trend
et = random noise

AT&T Daily Chart

Price Changes
Past price changes do not predict future
price changes
Cov(et, et+1) = 0

Why?

Prices change only due to new information


New information by definition is not
predictable by past information

Why Are Capital Markets Efficient


Markets are efficient when two
conditions are met:

Prices react quickly to new information


Prices react correctly to new information

Financial markets

Low barriers to entry, many participants


Most participants have similar preferences
(care about risk and return)

Why Are Capital Markets Efficient


Competition Quick price adjustments

New information usually becomes available to


more than one investor
If an informed investor waits, somebody else will
trade on the information
Competition between investors assures that
prices quickly adjust to the information

Types of Information

Historical
trading info

All public info

All public nontrading info

private info

All info

Forms of EMH
Weak - prices reflect only historical
trading information
Semi-strong - prices reflect all publicly
available information
Strong - prices reflect all information

Relationship between EMH Forms


Semi-strong form includes weak form
as special case
Strong form includes semi-strong and
weak form as special cases

Security Analysis and Market


Efficiency
Technical Analysis

Use historical trading (prices, volume, short


interest) information to predict future prices
Weak form efficiency is not consistent with
successful technical analysis

Fundamental Analysis

Use publicly available economic and accounting


information to predict stock prices
Semi strong form efficiency is not consistent with
successful fundamental analysis

EMH and Portfolio Management


If markets are efficient, then small (price
taker) investors cannot outperform the
market

Compute Optimal Risky Portfolio using portfolio


optimization
Hold mix of Optimal Risky Portfolio and risk-free
asset
Minimize tax burden

What about large (non-price-taker)


investors?

EMH and the Finance Industry


Financial analyst recommendations
Technical analysis data
Actively managed mutual funds
Discretionary brokerage accounts

Empirical Evidence on EMH


Quick price adjustments

Event studies

Correct price adjustments

Event studies
Return from trading strategies using historical
trading info
Returns from trading strategies using publicly
available info
Insider trading returns
Performance of professional money managers

Event Study Methodology


Examine speed and accuracy of the markets
response to new information. Does the
market under-react or over-react to new
information?
Detect abnormal returns around a narrow
event window

Test if cumulative abnormal returns are different


than zero

Make an assumption about normal returns

Use market model

Event Study Chart M&A Example

Return to Trading Strategies


Anomalies
Market bubbles and crashes
Small Firm Effect
Calendar Effects

January Effect
October Effect
Day of the week effects

Post-Earnings Announcement Drift /


Momentum
Long-term Reversals
Value Line Enigma

Speculative bubbles
Historical perspective

The Tulip-Bulb Craze: Holland, 1634-1637


The South Sea Bubble: Early 1700
Florida Real Estate Bubble: 1920s
Wall street: 1928-1929
Internet companies: 1996-2000
Recent turmoil

Speculative bubbles as naturally


occurring Ponzi schemes

January Effect - Gultekin &


Gultekin (1983)

Mutual Fund and Professional


Manager Performance
Overall mutual funds underperform on a riskadjusted basis

Potential errors in the risk-adjustment?

Some evidence of persistent positive and


negative performance

Incentives for creating superstars


Incubators for young mutual funds

Conclusion
The market for US large-cap liquid stocks is
approximately semi-strong form efficient
No reason to believe that other markets are
It is not easy to outperform the market on a
risk-adjusted basis

Have to be better informed


Larger
Faster

The Explosion of Challenger


(Mahoney & Mulherin, 2002)
11:39 a.m., January 28, 1986 The space shuttle
Challenger crashes
11:47 a.m. The news is announced on the Dow Jones
News Wire
Possible culprits
Rockwell shuttle maker
Lockheed ground support
Martin Marietta shuttle fuel tanks
Morton Thiokol solid fuel booster rocket

The Reason for the Explosion


Special Commission Report June 1986
Reason for the crash lack of resiliency at low
temperature of the seals of the booster rocket
(extremely cold that day in Florida)
Morton Thiokol is the guilty party

NASA and MT aware of the problem a year


ago
This is NOT public information for most stock
market participants

Stock Price Reaction


12 noon

R = -6.12%, L = -5.05%, MM = -2.83%


MT trading halt (sell order imbalance)

Around 1 p.m.

R -4%, L -4%, MM -5%


MT -10%

End of day

R = -2.48%, L = -2.14%, MM = -3.25%


MT = -11.86%

Evidence for Strong form Market


Efficiency
Quick price adjustments

Market needed less than an hour

Correct price adjustments?

200 M market value loss


Compare to
7 M settlement with astronauts families
10 M settlement with NASA
40 M foregone profits in repair work
150 M loss of NASA contract

How Did the Market Figure It Out


Informed trading

Selling in MT (large sell order imbalance)


Buying in other three

NYSE Specialists observe abnormal


trading and adjust prices

Behavioral Finance
Relaxes one of the most important set
of assumptions in classic finance
theory:

Investor Rationality

Investor Rationality
Consistent and homogeneous beliefs

Use all available information


No cognitive biases
No framing effects
Unbiased by emotions, herd effects

Consistent preferences

Tangible and measurable objectives (maximize


wealth, minimize risk, etc.)
Time consistent preferences
No regret

Real Investors
Triunal Brain
Cognitive biases
Time-inconsistent preferences
Influenced by hot states

Triunal Brain
The human brain can be decomposed in
three parts

Neocortex reason
Limbic system emotions
Reptilian brain actions

Most decisions are made by all three


parts of the brain

Cognitive Biases
Biases of Judgment

Overconfidence/Optimism
Hindsight bias
Quick Judgments
Anchoring
Representativeness

Confusing cause and effect

Overconfidence
Misjudge probability of downside risk
Place large weight on personal ability
Consequences

Frequent trading
Large positions
Undiversification

Odean and Barber (2001)

Men trade more and have a smaller number of


larger positions
Inferior performance compared to women

Cognitive Biases
Heuristics for reducing complexity

Humans are cognitive misers


Simplifying the facts
Mental accounting
Ignoring information/Selective perception

Mental Accounts
Example

$150 concert ticket


Scenario 1: Ticket was purchased in
advance. On the way to the theater you
lose it.
Scenario 2: Ticket was reserved and has to
be picked up at the theater. On the way to
the theater you lose $150 dollars.

Selective Perception
Bruner&Postman Experiment

Selective Perception
You can see 6 different cards.
Think on one.
Just think on it.
Are you thinking intently?
I will find the card on your mind.

Selective Perception
Now, look straight into my eyes
and think about your card
I cant see the card you have
chosen
but I know exactly the card
that is on your mind

Selective Perception

Look!
Your card is gone!

Do it again?
You can see 6 different cards.
Think on one.
Just think on it.
Are you thinking intently?
I will find the card on your mind.

Selective Perception
Now, look straight into my eyes
and think about your card
I cant see the card you have
chosen
but I know exactly the card
that is on your mind

Selective Perception

Look!
Your card is gone!

Cognitive Biases
Errors of preferences (Kahneman &
Tversky)
Prospect Theory

Valuing Changes, not States


Asymmetric Value Function

Consequences

Disposition effect
Framing bias

Prospect Theory
Choose one option

100% chance of $1,000


50% chance of $2,500

Most people choose the sure thing,


which can be explained by risk aversion
But

Prospect Theory
Choose one option

100% chance of losing $1,000


50% chance of losing $2,500

Most people choose the latter option,


which is actually risk seeking behavior
People treat gains and losses differently

Disposition Effect
Investors are reluctant to sell losers
Consequences

Sell winners
Keep losers
Documented both in individual accounts (Barber and Odean)
and in some mutual funds (Cici and Gibson)

Suboptimal strategy

Tax losses
Does not take advantage of short-term momentum
Wall St. Rule Cut your losses and let your winners ride

Cognitive Biases
Evaluating Consequences of Decisions

Regrets of commission and omission


Regret and risk-taking

Regret of Commission or
Omission
State your biggest mistake in past
investments

Is it because you did something?


Or because you did not do something?

Time-Inconsistent Preferences
Large weight placed on current benefits
or costs compared to all future ones
Consequences

Correct decisions about future actions,


incorrect decisions about the current
period
Procrastination
Addiction

Procrastination
Nave investors

Decision to sell now or postpone for tomorrow


Ignore the fact that tomorrow they will face the
same decision
Result postpone infinitely

More sophisticated investors

Recognize the procrastination effect


Usually behave almost optimally

Hot States and Projection Bias


Current emotional state affects forecasts of
behavior in the future

Food shopping when hungry


Use of anesthesia when in pain
Bullish after a current trading profit

Projection bias

Probability of being in a hot state in the future


Duration of hot state

Behavioral Trading
Professional investors are also subject
to biases like individuals
Biases may lead to suboptimal
performance

Disposition bias
Paralysis
Explosive risk taking

Currency trader story (Goldberg


& von Nitzsch)
Stage 1. Lucky Success

Several days of large profits


Become complacent
Viewed as expert by peers
Start ignoring/filtering information
Take large exposure based on overconfidence

Stage 2. Spiraling Loss

Position starts losing money


Reaction market is temporarily crazy, confirmed by peers
Filter all negative information
A series of hope states
Doubling down (more on strategy)
Exposure becomes huge
Position is finally liquidated at a large loss

Behavioral Trading Controls


Most investment companies manage
the behavioral biases of traders
Disposition bias/Sellers Remorse

Maximum Drawdown Rules

Procrastination/Paralysis

Team decision making


Short decision times
Daily balancing of inventory

Investor Behavior and Market


Efficiency
Many of the behavioral biases do not affect market
efficiency

For example, overconfidence leads to more trading and


collection of information and may actually improve efficiency
Others, may cancel out or be alleviated by a small number
of unbiased traders

Still, some of the biases may lead to systematic and


predictable mispricing or inflated volatility in the
market
The aggregate outcome of such biases is measured
by Sentiment Indicators

Market Sentiment
Sentiment Indicators

Technical
Consumer Sentiment
Market Participant Sentiment

Technical Indicators
Odd-lot trading
Put/Call Ratio

Amount of Put options Compared to Call


options

Short interest
VIX/VXN

Average implied volatility of a set of CBOEtraded options

Consumer Sentiment Indicators


Consumer Sentiment Index (CSI)

Since 1952
University of Michigan
Phone survey of 500-800 households
Preliminary results, the second Friday of each
month

Consuming Confidence Index (CCI)

Since 1967
5000 households
Last Thursday of the month

Market Participants Sentiment


Investor Intelligence Indicator

Based on text from 140 business newspapers


Weekly since 1950

Market Vane

Daily sentiment
Polls of brokerage houses, tip hotlines
Used by commodity traders

American Association of Individual Investors


(AAII)

Weekly sentiment generated by electronic voting


by its members

Evidence on the Value of


Sentiment Indicators
Some predictive power

CSI predicts the future equity premium


Charoenrook (2002)
Fisher and Statman (2002)
Professional sentiment
Semi-professional sentiment
Individuals
A mix of the three predicts SP500

References
Dorsey, W., 2004, Behavioral Trading,
Thomson/Texere
Goldberg, J., and R. von Nitzsch, 2001, Behavioral
Finance, Wiley & Sons
Kahneman, D., and M. Riepe, 1998, Aspects of
Investor Psychology, JPM, Summer, 52-65
Kiev, A., 2002, Trading in the Zone, Wiley & Sons
Mahoney, P., and H. Mulherin, 2002, The Stock Price
Reaction to the Challenger Crash: Information
Disclosure in an Efficient Market, Journal of
Corporate Finance,

You might also like