The Flaws of Efficient Market Hypothesis in Behavioral Finance

What’s the efficient market hypothesis in behavioral finance? What are the flaws of the efficient market hypothesis?

While famous economist Richard H. Thaler agrees with the EMH’s claim that it’s impossible to consistently beat the market, he disagrees with the claim that securities’ prices always reflect their intrinsic value. He considers three arguments against this thesis.

Take a look at the three arguments against the efficient market hypothesis below.

Argument #1: Closed-End Funds Violate the Law of One Price

According to Thaler, one argument against the efficient market hypothesis in behavioral finance is the violation of the law of one price—the thesis from traditional economics that a security should never sell at two different prices at the same time. Thaler explains that the law of one price follows directly from the EMH’s assumption that securities are always accurately priced. Under this assumption, it’s logically impossible for the same security to have two different prices, since that means at least one of the prices is different from the security’s intrinsic value. 

However, Thaler argues that closed-end funds violate the law of one price and therefore are inconsistent with the EMH. Closed-end funds, he explains, are investing funds—that is, pools of money from investors that money managers invest on their behalf—that raise money via an initial public offering (IPO) in which investors can purchase shares of the closed-end fund. The fund managers then use the money raised through this IPO to purchase stocks. For example, if the IPO raised $10 million, then the fund managers would invest this $10 million in the stock market as they see fit. Crucially, investors can then also buy and sell shares of the closed-end fund itself, although they cannot invest money directly into the fund after the IPO.

Thaler notes that, if the law of one price were true, the market price of closed-end funds should be equivalent to their net asset value—that is, the value of the assets that the fund owns. For example, imagine that after the IPO raised $10 million, the fund’s managers invested $5 million in Netflix which rose 20% (yielding $6 million in value) and $5 million in Google which rose 10% (yielding $5.5 million in value). Then, the IPO’s net asset value would be $11.5 million, and because investing in the closed-end fund is functionally equivalent to investing in its underlying assets, the law of one price states that the market price of the closed-end fund (its share price multiplied by its total number of shares) should also equal $11.5 million.

In practice, however, closed-end funds’ market prices are rarely equivalent to their net asset value. On the contrary, Thaler points out that closed-end funds typically trade at a 10-20% discount of their net asset value (though, occasionally, they also trade above net asset value), meaning that investors are essentially purchasing shares of the closed-end funds’ underlying assets for less (or occasionally more) than their market price. Either way, closed-end funds represent a violation of the law of one price. 

Argument #2: Overreaction in the Stock Market

Thaler argues that, much like closed-end funds demonstrate inefficiencies by violating the law of one price, the stock market also demonstrates inefficiencies by overreacting. In other words, he argues that stocks that recently outperformed the market tend to underperform it in the future. His research demonstrates that significant past overperformance is predictive of significant future underperformance, and vice versa. This contradicts the EMH’s prediction that past market performance cannot predict future performance.

Thaler explains that if stocks were always accurately priced like the EMH claims, there would be no way to determine whether a stock will increase or decrease in value. After all, if you knew that a stock would increase in value, that would illustrate that it’s currently undervalued and therefore not accurately priced. By contrast, if you knew that a stock would decrease in value, that would illustrate that it’s currently overvalued and therefore not accurately priced. 

However, the occurrence of overreaction suggests that some stocks are undervalued or overvalued, contrary to the EMH’s prediction. To show as much, Thaler conducted a study comparing the most successful and unsuccessful stocks on the New York Stock Exchange (NYSE) over various three- to five-year periods—that is, those that outperformed the market by the largest margin and those that underperformed the market by the largest margin. Across various portfolios, his results were consistent: In the subsequent three- to five-year periods, the previously least successful stocks outperformed the market by an average of 30%, and the previously most successful stocks underperformed the market by an average of 10%. 

Argument #3: Robert Shiller’s Case Against Rational Pricing

While Thaler uses violations of the law of one price and overreaction as arguments against EMH, he also cites the ideas of Nobel laureate Robert Shiller. Shiller argued that the US stock market is more volatile than it should be if it were perfectly efficient.

Shiller’s argument is based on the assumption that a stock’s dividends are a proxy for its intrinsic value. To see why, Thaler asks us to imagine we own a stock that we’ll never sell—then, the value of this stock would be equal to the present value of its future dividends (that is, the total dividend payments over the stock’s life discounted at an appropriate rate). However, he points out that we don’t know exactly what a stock’s future dividend payments will be. Thus, the current stock price is effectively a prediction of net future dividend payments, at the appropriate discount rate.

(Shortform note: Experts note that, although dividends can function as a proxy for intrinsic value, not all companies issue dividend payments in the first place. For instance, while established companies typically issue consistent dividends, the same is often untrue of start-up companies, which normally lack the capital to issue dividends to shareholders. Consequently, it can be more difficult to estimate the intrinsic value of these fledgling companies.) 

Thaler then relates Shiller’s argument that, because stock prices are a prediction of future dividends, we should expect the volatility of stock prices to match those of dividends. To take an analogous example, imagine that you’re predicting how a set of 10 counties will vote in the upcoming election. If these counties have historically had similar voting patterns (say, voting on an average 60% Republican plus or minus five percent), then your prediction should reflect this similarity—you shouldn’t predict that one county will vote 20% Republican while another will vote 90% Republican, since this volatility doesn’t reflect the historical data. 

However, when it comes to stock prices, these prices exhibit far more volatility than future dividend payments—whereas future dividends are remarkably consistent over time, with slight-but-steady increases, stock prices exhibit erratic swings in prices. Thus, assuming that dividends are a proxy for intrinsic value, Shiller concludes that stock prices must often depart from intrinsic value, contrary to the EMH.