Skip to Content

Jonathan B. Berk


Jonathan's Working Papers

Working Papers

"Assessing Asset Pricing Models using Revealed Preference," with Jules H. van Binsbergen.

Abstract: We propose a new method of testing asset pricing models that does not rely on prices and returns but on quantities (flows) instead. Under the assumption that capital markets are competitive and investors rational, an asset pricing model can only be correct if investors are using it in their capital allocation decisions. Therefore, any investment opportunity that the model identifies as having a nonzero alpha must be accompanied by capital flows of the same sign as the alpha. We use the data on active mutual funds to identify such flows, and find that the recent alternatives to the Capital Asset Pricing Model do not improve upon the original model.
Note on the relation between the chronology of Barber, Huang and Odean and this paper.


"Measuring Managerial Skill in the Mutual Fund Industry," with Jules H. van Binsbergen, National Bureau of Economic Research Working Paper 18184.

Abstract: Using the dollar-value a mutual fund manager adds as the measure of skill, we find that not only does skill exist (the average mutual fund manager adds about $2 million per year), but this skill is persistent, as far out as 10 years. We further document that investors recognize this skill and reward it by investing more capital with skilled managers. Higher skilled managers are paid more and there is a strong positive correlation between current managerial compensation and future performance.
Data Appendix


"Limited Capital Market Participation and Human Capital Risk," with Johan Walden, National Bureau of Economic Research Working Paper 15709.

Abstract: The non-tradability of human capital is often cited for the failure of traditional asset pricing theory to explain agents' portfolio holdings. In this paper we argue that the opposite might be true | traditional models might not be able to explain agent portfolio holdings because they do not explicitly account for the fact that human capital does trade (in the form of labor contracts). We derive wages endogenously as part of a dynamic equilibrium in a production economy. Risk is shared in labor markets because firms write bilateral labor contracts that insure workers, allowing agents to achieve a Pareto optimal allocation even when the span of asset markets is restricted to just stocks and bonds. Capital markets facilitate this risk sharing because it is there that firms offload the labor market risk they assumed from workers. In effect, by investing in capital markets investors provide insurance to wage earners who then optimally choose not to participate in capital markets. The model can produce some of the most important stylized facts in asset pricing: (1) limited asset market participation, (2) the seemingly high equity risk premium, (3) the very large disparity in the volatility of consumption and the volatility of asset prices, and (4) the time dependent correlation between consumption growth and asset returns.


"Return Persistence and Fund Flows of the Worst Performing Mutual Funds," with Ian Tonks, National Bureau of Economic Research Working Paper W13042.

Abstract: We document that the observed persistence amongst the worst performing actively managed mutual funds is attributable to funds that have performed poorly both in the current and prior year. We demonstrate that this persistence results from an unwillingness of investors in these funds to respond to bad performance by withdrawing their capital. In contrast, funds that only performed poorly in the current year have a significantly larger (out) flow of funds/return sensitivity and consequently show no evidence of persistence in their returns.


"Can Boundedly Rational Agents Make Optimal Decisions? A Natural Experiment," with Eric Hughson, Robert Day School of Economics and Finance Research Paper 2008-7.

Abstract: The television game show The Price Is Right is used as a laboratory to test consistency of suboptimal behavior in an environment with substantial economic incentives. On the show, contestants compete sequentially in two closely related games. We document that contestants who use transparently suboptimal strategies in the objectively easier game use the optimal strategy almost all of the time in the game that is much more difficult to solve. Further, there is no consistency in the mistakes that are made in the two games. One cannot predict, conditional on play in one game, whether play in the other game will be optimal. The results have implications for the consistency of behaviorally based economic theory that relies on evidence derived in a laboratory setting.