Our intial discussion with C. Thomas Howard about his ”Behavioral Portfolio Management” has done so well we thought you would like to read more of the discussion about how to deploy behavioral finance practically. Particularly germane to implementing behavioral portfolio management is understanding Howard’s views on emotions and volatility, and how he handles them when managing money. Recall that Howard is professor emeritus at Daniels College of Business, University of Denver, and co-founder of AthenaInvest, whose Athena Pure Valuation is one of the top performing portfolios in recent years.
CFA Institute: So when you are implementing your behavioral portfolio management you are equating emotions with irrationality and with volatility?
C. Thomas Howard: Yes, short-term volatility is largely the result of emotional investor overreaction to unfolding events. But this is not unique to me, it was first proposed by Robert Shiller over 30 years ago, and his assertion of excess volatility remains intact today, even after numerous studies attempted to link volatility to changing fundamentals. So we are left with the inescapable conclusion that volatility is a measure of emotion and not risk, except in the case of building portfolios to meet short-term needs.
Isn’t it possible for two hyperrational people to simply be considering very different sets of data, and hence, coming to two very different opinions about an asset? Wouldn’t this also lead to a lot of volatility, especially if they don’t trade with one another (i.e., there is no cancelling out of each other in the data)?
This is a conceptually difficult issue that Hersh Shefrin does a superb job of addressing in his book
A Behavioral Approach to Asset Pricing. Let me take a stab at responding to your question, with apologies to Hersh if I fumble the answer.
If the market were comprised of only rational investors, we would expect to see some volatility as new information arrives in the market. We would also expect this volatility to be largely driven by changes in underlying fundamentals, since that is all rational investors care about. This is why tests challenging Shiller’s excess volatility claim use various measures of fundamental change to see if volatility can be explained. It cannot.
You mention rational investors with different data sets. In a rational world, there will be only one relevant data set that every rational investor is seeking to uncover. If one consistently misses out on generating this set, then one will eventually go belly up as the better informed investors take advantage of improperly informed trades. So in a rational market, we would expect all investors to drive to the same data set and, in turn, to the same future expectations.
Thus investors with different data sets can probably explain very little of the excess volatility. Thus emotions dominate.
Well, let me give you an example of two radically different data sets and investors and logical outcomes. At time minus two years, one investor is a US-based fundamental investor conducting her own proprietary research on a business. As a part of that research she examines the 10-K of a firm, as well as other relevant public disclosures such as a 10-Q or 8-K. Her examination of the data leads to her buy shares in the firm. Now fast forward two years and that same data set, based on financial statements, leads her to believe it is time to sell. The buyer of those shares is a Hong Kong-based investor who notices that the beta of the US company relative to the Hong Kong dollar and the Hang Seng is much less than one, so he buys shares in the business as a part of an asset allocation strategy. Two very different data sets, but rational investors who have very differing views of the underlying asset. So isn’t it possible that volatility is the result of different, rational views? [Note: Kent Osband in a recent interview essentially argued this same point.] After all, here in the US, where we are politically gridlocked, we have two sides, both hyperrational, with their choices leading to tremendous volatility of outcomes.
In your example, there would be some volatility attributable to the changing fortunes of the company, as determined by the first investor. If we measured the volatility of the stock over that two year time period, how much of it would be explained by the changing outlook (i.e., fundamentals) of the company? Research says very little. What we would find is that the stock price varied wildly over these two years, with almost none of it explained by changes in fundamentals, such as those that are the focus of your first investor. We are left with the conclusion that investor emotions are the root cause of the vast majority of these price changes.
Your Hong Kong investor appears to be an uninformed investor that the first investor is thrilled to have in the market, since this person is willing to take the other side of her informed trade. In the language of BPM, the first investor is a behavioral data investor (BDI) while the second investor is part of the emotional crowd. BDIs take positions opposite emotional crowds in order to earn superior returns.
Isn’t there also a danger in attempting to quantify behavior with irrationality with volatility? What makes you so confident that there is actual causality in the data?
Attributing excess volatility to emotions is a default position. If stock prices are the result of a rational valuation process, then as new company, industry, market, and economic information arrive in the market, it should be the case that prices fluctuate randomly, as demonstrated years ago by Paul Samuelson. But price volatility exceeds underlying fundamental volatility by a wide margin. So the supposition is that investor emotions drive the rest.
Maybe we will uncover an as-yet-unidentified measure of fundamental volatility that will do a better job of capturing more of the market’s volatility (recently, municipal garbage was put forward as just such a measure based on the contention that it is closely related to consumption). Until such a measure is discovered, emotions will retain the title of key driver of market volatility.
