Comments on Richard Zeckhauser's Investing in the Unknown and Unknowable
12 Pages Posted: 4 Feb 2013
Date Written: January 23, 2013
Abstract
I agree entirely with Larry Summers' two main points: (1) Richard Zeckhauser has written an important paper and (2) it's probably no easier to earn excess profits by hiring skilled managers or investing alongside them than it is to pick market-beating investments in the first place. Zeckhauser's paper is important because those of us who invest for a living seem to spend more time grappling with the unknowable than we do with anything that fits into a framework that could be characterized by probability distributions and utility functions. If a mechanism for generating excess returns could be expressed in terms of a process, someone would have arbitraged it away by now. My comment here will focus on Summers' second point: I believe some managers can outperform the indices, but to the extent that the uninformed public can gain access to these managers or identify who they are, it's too late. A superior manager will either charge high fees that extract most of the excess returns, the managed assets will trade at a premium to net asset value (rewarding the early investors who bet on the manager before his ability was generally observable) or the manager will ration access based on investors' characteristics the manager values. This argument leads to a slightly modified definition of market efficiency: excess returns are available, but unless you have some advantages over other investors in terms of experience, private information, ability or scale, they are not available to you.
It is tempting to add a third "U" to the "unknown and unknowable": "uncommunicable." However, this addition would be redundant since unknowable things, by their nature, cannot be expressed. David Ricardo, on the eve of the Battle of Waterloo, might have said, "British Government bonds offer a high reward for the risk." On the surface, this looks like a clear proposition. It has the same grammar as "investors can achieve higher expected returns and less risk with a diversified portfolio of stocks than with a portfolio consisting of a few randomly-selected stocks" or "stocks that pay high dividends do not, on average, generate higher total returns than stocks that pay no dividends at all." These last two statements can be tested by collecting data on stock prices over time. But what would it look like for Ricardo's hypothetical statement to be proven false? Suppose France had won, or Britain had won but the bonds did not appreciate as much as Ricardo expected. Maybe it was bad luck (although it would be a misuse of language to say "bad luck" since it does not make sense in this context to distinguish between bad luck and bad judgment). The conditions that led Ricardo to buy the bonds were full of ambiguity and would never occur again. It would be nonsense for him to claim that buying British government bonds before Waterloo maximized his expected utility. All he could say is that he was willing to buy them.
Perhaps investors such as David Ricardo or Warren Buffet cannot transmit their knowledge of the UU, but why can’t uninformed investors tag along by following their advice or hiring them to manage our money? An investor who is hiring a manager faces his own encounter with the unknown. Sorting out the good ones from the bad cannot be straightforward. If managers' skills were observable at zero cost, the best ones would command the highest fees until the market cleared and all investors ended up with the same returns over time after fees. Similarly, if there were no way at all to discern managers' quality, bad ones would flood the market until the average manager's returns after fees matched the returns an investor could achieve through a passive strategy. Even if an investor finds a skillful manager, it might not last – skill, itself, is a transitory concept in a non-stationary environment subject to intense competition. "Past performance," as they say, "does not imply a guarantee of future performance."
Many academics will reject the debate out of hand and assert that investing to beat the market has the intellectual heft of astrology: sometimes it works, sometimes it doesn't, and as long as we exclude the cases where it doesn’t work, we can present data that is fairly impressive. Believers in the extreme form of market efficiency draw support from the failure of simple rules to beat the market such as "buy stocks with a high book value relative to share price" or "buy stocks that went up last month." We can find examples in nearly any issue of any finance journal. Ultimately, it's an empirical question and Zeckhauser tells three convincing anecdotes from his own experience – Tengion, Gazprom and Davis Oil – as well as a Buffet story to illustrate how a savvy investor who is not afraid of some ambiguity can earn outsized returns. I will begin this comment by adding an example of my own. My trades, while more pedestrian than Zeckhauser's, come in the form of a natural experiment and contribute one more piece of evidence suggesting that the strongest formulations of the market efficiency hypothesis may overstate the case.
This paper is a comment on Investing in the Unknown and Unknowable by Richard J. Zeckhauser which can be found at: http://ssrn.com/abstract=2205821.
Suggested Citation: Suggested Citation