Behavioral finance in the investment process

Various articles claim that behaviorally managed portfolios have not achieved higher returns than those managed in traditional styles, such as growth and value. I believe that the lack of success is mostly due to the incorrect application of behavioral finance to the management of portfolios and that behavioral tools can help investors meet and exceed investment objectives.

Among the more puzzling phenomena in finance is the apparent negative empirical relationship between standard measures of risk and reward in equities. Established financial theory dictates that investors expect to be compensated for taking risk, yet the negative relationship between volatility and future stock returns is well documented (Fama and French, 1992, and Frazzini and Pedersen, 2010), standing as an empirical critique of the Capital Asset Pricing Model (CAPM).

Recent market volatility and the growing interest in minimum volatility strategies have increased scrutiny on this puzzling phenomenon. Research by Barberis and Huang (1999) proposes a behavioral rationale for the apparent volatility anomaly. Based on Kahneman and Tversky’s cumulative prospect theory, they postulate that investors have an explicit preference for stocks with lottery-like payoffs, or a low probability of a very high return.

Stocks with a history of infrequent extreme positive returns without a commensurate frequency of similarly strong below-average returns are said to demonstrate positive “skewness.” A behavioral preference for stocks with lottery-like payoffs leads positively skewed stocks to be “overpriced” under the mean-variance assumptions of CAPM, leading to subsequent negative average excess returns.

Recent research has demonstrated that the volatility anomaly may be explained by the negative relationship between skewness and future returns. Investors with a behavioral bias for stocks with lottery-like payoffs appear willing to pay a premium for stocks with high skewness.

For portfolios where skewness is low, expected returns increase with higher levels of risk consistent with CAPM. This relationship is reversed for high levels of skewness where the assumption of normally distributed returns does not hold. On average, skewness increases with higher levels of volatility. As a result, stocks with both high levels of volatility and high historical return skewness exhibit extremely low future returns.

Behavioral tools can improve stock picking

Stock selection, I believe, can also be improved by the use of behavioral tools. Empirical academic research has highlighted two types of pervasive investor behavior which impact share prices, causing them to diverge from those predicted by the Efficient Markets Hypothesis (EMH); underreaction of stock prices to news and overreaction or anchoring to a stream of good or bad news.

As an example, consider a company that has just reported a positive earnings surprise due to some fundamental improvement in its business. If EMH held, then the stock price should immediately change to the new equilibrium price where all the new information would be reflected.

However, in practice, analysts and investors will take a series of steps to fully reflect the new available information and, therefore, an investor who buys on the announcement of the news can generate a positive excess return. When a company’s fortunes change from negative to positive and vice versa, investors tend to overreact and anchor on the past. As a result, it takes time for the change to be reflected in the share price. Similar to underreaction, an investor can generate excess returns by focusing on those companies where investors anchor.

Portfolio managers are, like the rest of us, flawed individuals. They tend to make behavioral mistakes despite awareness that investors are not always rational. Allianz Global Investors Capital has been working closely for the past year with Cabot Research to identify and point out inefficient behaviors to our portfolio managers. We then work with them to avoid similar behavior in the future.

Through proprietary analysis, Cabot Research evaluates every buy and sell in a portfolio and separates each into winners and losers. We can then create a profile of successful and unsuccessful buys and sells. In addition, we can measure the impact of decisions about holding period and position sizing. Another key area for study is how the manager behaves with respect to significant underperformers. The goal is to end up with a list of manager behaviors that can be observed and corrected.

Regret aversion limits returns

For example, we have found that portfolios often exhibit “regret aversion” – the tendency not to reach full position weight in a timely manner. Portfolio managers tend not to buy a complete position in a stock which rises in price after purchase. This can cost the portfolio relative return because of the tendency of investors to underreact. A second significant effect is the “endowment effect” – the tendency to hold older winners too long. In this case, a portfolio manager will overvalue a stock in the portfolio and not sell it when fortunes turn. This is the anchoring effect mentioned above.

Once again, timely sales could enhance portfolio returns. Other potential behaviors include risk aversion, the tendency to sell young winners well before they have contributed their full alpha to the portfolio; avoidance of pain, a manager’s tendency to sell stocks as they are experiencing a substantial drop in price; and loss aversion, the reluctance to sell losers in hope of their possible recovery.

Behavioral finance has been shown to positively impact investors’ financial decision-making. By applying the behavioral toolbox to strategy, design, stock selection and portfolio evaluation, we can use these same tools to better meet and exceed our clients’ expectations.

This article originally appeared at ProjectM Online