
Behavioral Economics — Behavioral Finance: Limits to Arbitrage and Investor Biases
Behavioral Finance — For Biases to Stay in Prices, Arbitrage Must Be Blocked
Financial markets are the least favourable arena for behavioral economics. Stakes are large, participants are professional, and mispricing is an opportunity for arbitrage profits. The argument of behavioral finance therefore has two steps. First, some investors’ biases are systematic (Chapters 2–6). Second, those biases remain in prices because there are limits to arbitrage. Neither alone can explain price anomalies.
1. The efficient market hypothesis and its defence
The standard defence of the efficient market hypothesis runs as follows. Even if there are irrational investors, their errors cancel out; and if they do not, rational arbitrageurs exploit and correct the mispricing. This is the no-arbitrage principle of Financial Economics Chapter 1.
2. Arbitrage is costly and risky
| Factor | Description |
|---|---|
| Fundamental risk | Without a perfect substitute asset, risk remains even when hedged |
| Noise trader risk | Mispricing can widen before it corrects |
| Funding constraints | When losses occur, investors withdraw funds and positions are forcibly liquidated |
| Short-sale constraints | Shares to borrow are unavailable or expensive |
The clearest case is Royal Dutch and Shell. The two companies were a single group that had agreed to split cash flows 60:40, so the ratio of the two shares’ market values should also have been 60:40. In practice the ratio deviated by tens of per cent for years. If even twin shares with fixed cash-flow shares did not converge through arbitrage, larger errors can persist in ordinary shares with no clear substitute.
Shleifer and Vishny stressed that arbitrageurs manage other people’s money. If prices move further apart and short-term losses occur, investors withdraw funds, and arbitrageurs must close positions just when they are most attractive. It is a paradox: the larger the mispricing, the weaker arbitrage may become.
3. The disposition effect: gains quickly, losses slowly
Investors sell shares that have risen too quickly and hold shares that have fallen too long. Odean (1998) reported from brokerage account data that individual investors realized about 14.8% of their paper gains (PGR) and about 9.8% of their paper losses (PLR). The propensity to realize gains was about 1.5 times as large.
The reflection effect of Chapter 2 explains this. Taking the purchase price as the reference point, winning shares are in the gain domain (risk aversion → lock in a sure gain) and losing shares in the loss domain (risk seeking → gamble on recovery). The mental accounting of Chapter 6 points the same way: if the share is not sold, the loss in its account is not confirmed.
The disposition effect has costs. For tax purposes it is better to realize losses and offset them against gains, and in Odean’s data the winning shares that were sold went on to earn higher returns than the losing shares that were held.
4. Excessive trading
The overconfidence of Chapter 5 shows up in trading frequency. Barber and Odean (2000) analysed US household accounts for 1991–1996 and reported that the households that traded most earned a net return of 11.4% a year, far below the market’s 17.9%. Before trading costs the gap was small. The problem was not poor stock selection but paying costs by buying and selling too often. A follow-up study found that men trade about 45% more than women and lose correspondingly more return, consistent with psychological evidence that overconfidence is stronger among men.
5. The equity premium puzzle and myopic loss aversion
Historically, equities have earned several per cent a year more than government bonds. As Financial Economics Chapter 2 showed, explaining this gap by consumption risk alone requires unrealistically high risk aversion.
Benartzi and Thaler (1995) explained the puzzle with the narrow framing and loss aversion of Chapter 2. The more often investors evaluate their portfolios, the more often they see losses. Equities fall on nearly half of all days but have a small probability of loss over ten years. With short evaluation periods and , equities become much less attractive, and investors hold them only if paid a high premium. They calculated that the observed premium is explained when the evaluation period is about one year.
A policy implication of this explanation concerns how often pension accounts report returns. In experiments, participants who received return information less often invested more in equities.
Check your understanding
Share A, bought at 100,000 won, is now worth 120,000 won, and share B, bought at the same price, 80,000 won. If both have the same expected future return and you need 80,000 won in cash, which is better to sell for tax purposes? If capital gains are taxed, selling B and realizing a 20,000-won loss is better, because the loss can offset other gains. The disposition effect pushes you to sell A instead.
References
- Andrei Shleifer and Robert Vishny, “The Limits of Arbitrage,” Journal of Finance (1997)
- Terrance Odean, “Are Investors Reluctant to Realize Their Losses?,” Journal of Finance (1998)
- Brad Barber and Terrance Odean, “Trading Is Hazardous to Your Wealth,” Journal of Finance (2000)
- Shlomo Benartzi and Richard Thaler, “Myopic Loss Aversion and the Equity Premium Puzzle,” Quarterly Journal of Economics (1995)
- Nicholas Barberis and Richard Thaler, “A Survey of Behavioral Finance,” Handbook of the Economics of Finance (2003)
OIYO Editorial
Editorial DeskThe OIYO editorial desk researches money, law, lifestyle, and self-understanding topics against primary sources and public statistics. Every piece carries source notes and is reviewed on a regular cycle for accuracy and usefulness.