Economics Honors Thesis

On October 19, 1987 stock prices fell over 20 percent in one day without any important news.  Rational models have trouble explaining all of the behavior we see in financial markets.

Five themes of research in Behavioral Finance:

1. Limits to Arbitrage - not all inefficiencies (such as the suboptimal choices of "irrational" agents) present an opportunity for other "rational" agents to step in and make a profit (i.e. career choice, marriage, and saving for retirement).  In other cases, arbitrageurs need long time horizons in order to correct slow-moving market prices (i.e. LTCM, Royal Dutch Shell, and 3Com Palm).


2. Stock Returns and the Equity Premium - 1) numerous researchers have found that over a timespan of several years, the stock market is somewhat predictable (not a random walk as long thought previously) and 2) behavioral psychology may provide an explanation for why there is an "equity premium", which is that according to traditional asset pricing models of expected utility maximization, the higher returns from equities over the risk-free rate is too large (explanations include loss aversion, narrow framing, and house money effect).


3. Empirical Studies of Overreaction and Underreaction - the question of whether or not the higher return to value strategies (such as buying stocks with extremely low ratios of price to book value) can be attributed to risk.  Stocks tend to exhibit mean-reversion over the long term, but they also tend to follow the opposite pattern (momentum) in the short term.


4. Theories of Overreaction and Underreaction - there are three models: 1) Barberis, Schleifer, and Vichny (BSV) 2) Daniel, David Hirschleifer, and Avanidhar Subrahmanyam (DHS), and 3) Harrison Hong and Jeremy Stein.  The first two deal with representative agent models, and the third envisions a world in which there are two types of boundedly rational agents: "Newswatchers" or those who make price forecasts based on private information only, and "Momentum traders" or those who trade only on past price movements.


5. Investor Behavior - In a study that examined large datasets of trading information, they found that investors are reluctant to sell stocks that have declined in value and investors trade too much (overconfidence, especially pronounced among men).

From this reading, I have noticed that two important factors that affect long-term investment returns are time horizons and financial constraints (you may remember these were mentioned as limits to arbitrage).  I do not believe that these two factors are taken into much consideration in the Efficient Markets Hypothesis, and I want to look into what the EMH response is.  Also, I plan to read all three of the papers listed under theories of overreaction and underreaction this upcoming week. 

For some background idea of where my theory is coming from, I highly suggest reading this paper summarizing Adaptive Markets Hypothesis.  It's only about 10 pages long double-spaced plus it's written by me.

Lately, I have been trying to find some more AMH articles to read.  I learned a couple of tips on how to research Economics literature from the social sciences librarian here at school.  An overview of what tools are available can be found here: http://www.amherst.edu/library/research/courserelated/fall07/econ77.html

A couple resources that I found to be useful:

1. EconLit
2. WorldCat
3. Wall Street Journal (from 1982 to present)
4. Importing references into End Note
5. Infomine
6. UMass Isenberg School of Business

I will be post specific articles that I looked at when I get a chance.  Also, I have formed a small group with Mee Sun Song and Travis Scholes.  We all have topics somewhat related to finance, and I'm looking forward to working with them as we critique each other's work.

This is a review of the following three essays:

1. "A Survey of Behavioral Finance" by Nicholas Barberis and Richard H. Thaler
2. "Reconciling Efficient Markets with Behavioral Finance: The Adoptive Markets Hypothesis" by Andrew Lo

3. "Risk Management for Hedge Funds: Introduction and Overview" by Andrew Lo