While Kosowski et al. (2006, Journal of Finance 61, 2551–2595) and Fama and French (2010, Journal of Finance 65, 1915–1947) both evaluate whether mutual funds outperform, their conclusions are very different. We reconcile their findings. We show that the Fama‐French method suffers from an undersampling problem that leads to a failure to reject the null hypothesis of zero alpha, even when some funds generate economically large risk‐adjusted returns. In contrast, Kosowski et al. substantially overreject the null hypothesis, even when all funds have a zero alpha. We present a novel bootstrapping approach that should be useful to future researchers choosing between the two approaches.

In almost every area of empirical finance, researchers confront multiple tests. One high-profile example is the identification of outperforming investment managers, many of whom beat their benchmarks purely by luck. Multiple testing methods are designed to control for luck. Factor selection is another glaring case in which multiple tests are performed, but numerous other applications do not receive as much attention. One important example is a simple regression model testing five variables. In this case, because five variables are tried, a t-statistic of 2.0 is not enough to establish significance. Our paper provides a guide to various multiple testing methods and details a number of applications. We provide simulation evidence on the relative performance of different methods across a variety of testing environments. The goal of our paper is to provide a menu that researchers can choose from to improve inference in financial economics.

Identifying the factors that drive the cross-section of expected returns is challenging for at least three reasons. First, the choice of testing approach (time series versus cross-sectional) will deliver different sets of factors. Second, varying test portfolio sorts changes the importance of candidate factors. Finally, given the hundreds of factors that have been proposed, test multiplicity must be dealt with. We propose a new method that makes measured progress in addressing these key challenges. We apply our method in a panel regression setting and shed some light on the puzzling empirical result that the market factor drives the bulk of the variance of stock returns, but is often knocked out in cross-sectional tests. In our setup, the market factor is not eliminated. Further, we bypass arbitrary portfolio sorts and instead execute our tests on individual stocks with no loss in power. Finally, our bootstrap implementation, which allows us to impose the null hypothesis of no cross-sectional explanatory power, naturally controls for the multiple testing problem.

The constant maturity zero-coupon yield curve for the US Treasuries is one of the most studied datasets. We construct a new yield curve using a non-parametric kernel-smoothing method with a novel adaptive bandwidth specifically designed to fit the Treasury yields. Our curve is globally smooth while still capturing important local variation. Economically, we show that applying our data leads to different conclusions from using the leading alternative data of Gürkaynak et al. (2007) (GSW) when we repeat two popular studies of Cochrane and Piazzesi (2005) and Giglio and Kelly (2018). Statistically, we show our dataset preserves information in the raw data and has much smaller pricing errors than GSW. Our new yield curve is maintained and updated online, complemented by bandwidths that summarize information content in the raw data.

Our paper explores the link between cross-sectional fund return dispersion and performance evaluation. The foundation of our model is the simple intuition that in periods of high return dispersion, which is associated with high levels of idiosyncratic risk for zero-alpha funds, unskilled managers can more easily disguise themselves as skilled. Rational investors should be more skeptical and apply larger discounts to reported performance in high dispersion environments. Our empirical results are consistent with this prediction. Using fund flow data, we show that a one standard deviation increase in cross-sectional return dispersion is associated with an 11% to 17% decline in flow-performance sensitivity. The effect is stronger for recent data and among outperforming funds.

Past fund performance does a poor job of predicting future outcomes. The reason is noise. Using a random effects framework, we reduce the noise by pooling information from the cross-sectional alpha distribution to make density forecasts for each individual fund’s alpha. In simulations, we show that our method generates parameter estimates that outperform alternative methods, both at the population and at the individual fund level. An out-of-sample forecasting exercise also shows that our method generates improved alpha forecasts. Received November 23, 2016; editorial decision November 1, 2017 by Editor Andrew Karolyi.

Hundreds of papers and factors attempt to explain the cross-section of expected returns. Given this extensive data mining, it does not make sense to use the usual criteria for establishing significance. Which hurdle should be used for current research? Our paper introduces a new multiple testing framework and provides historical cutoffs from the first empirical tests in 1967 to today. A new factor needs to clear a much higher hurdle, with a t-statistic greater than 3.0. We argue that most claimed research findings in financial economics are likely false.

We develop a model of a Financial Stability Fund (the “Fund” henceforth) for a union of sovereign countries. By design, the contract prevents country defaults, as well as undesired expected losses, which in a union translate into excessive risk mutualizations. A participant country has greater ability to borrow and share risks than using sovereign debt financing. The Fund contract also provides better incentives for the country to reduce endogenous risks. These efficiency gains arise from the ability of the Fund to offer long-term contingent financial contracts, subject to limited enforcement and moral hazard constraints. We develop the theory and quantitatively compare the constrained-efficient Fund economy with an incomplete markets economy with default. We calibrate our economy to the euro area “stressed countries” in the debt crisis (2010–2). Substantial welfare gains are achieved, particularly in times of crisis. The Fund is, in fact, a risk-sharing, crisis prevention and resolution mechanism, which transforms the participant countries’ defaultable sovereign debt into the union’s safe assets. In sum, our theory can help to improve current official lending practices and, for example, to eventually design a European Fiscal Fund.

Multiple testing plagues many important questions in finance such as fund and factor selection. We propose a new way to calibrate both Type I and Type II errors. Next, using a double‐bootstrap method, we establish a t ‐statistic hurdle that is associated with a specific false discovery rate (e.g., 5%). We also establish a hurdle that is associated with a certain acceptable ratio of misses to false discoveries (Type II error scaled by Type I error), which effectively allows for differential costs of the two types of mistakes. Evaluating current methods, we find that they lack power to detect outperforming managers.

We propose a latent-variables approach to recover extrapolative beliefs from asset prices. We estimate a present-value model of the price–dividend ratio of the market that embeds both return extrapolation and cash-flow extrapolation, alongside discount rates and rational expectations of dividend growth. This approach allows us to measure extrapolation bias without having to rely on survey data, and it inherently guarantees that the researcher focuses on a set of beliefs that matter for price formation. We show that extrapolative beliefs extracted from prices are highly correlated with surveys and that survey-based and price-based extrapolative beliefs share similar predictive properties for future returns, with the former improving upon the latter.