We analyze the reliability of voluntary disclosures of financial information, focusing on widely‐employed publicly‐available hedge fund databases. Tracking changes to statements of historical performance recorded between 2007 and 2011, we find that historical returns are routinely revised. These revisions are not merely random or corrections of earlier mistakes; they are partly forecastable by fund characteristics. Funds that revise their performance histories significantly and predictably underperform those that have never revised, suggesting that unreliable disclosures constitute a valuable source of information for investors. These results speak to current debates about mandatory disclosures by financial institutions to market regulators.

We propose a new explanation for the persistence of gravity in international investment flows based on new facts about large cross-border commercial real estate transactions. Buyers in these transactions preferentially match with counterparties from own or proximate countries; such affinity-based matching helps alleviate financial investment frictions. We set up and structurally estimate a model of capital allocation in a decentralized market with an investment friction, which delivers the price, volume, and counterparty matching patterns in the data. The model shows that if clusters of high-affinity counterparties lie along historical routes, as in the data, preferential matching can perpetuate gravity relationships.

We investigate the role of trade credit links in generating cross-border return predictability between international firms. Using data from 43 countries from 1993 to 2009, we find that firms with high trade credit located in producer countries have stock returns that are strongly predictable based on the returns of their associated customer countries. This behavior is especially prevalent among firms with high levels of foreign sales. To better understand this effect we develop an asset pricing model in which firms in different countries are connected by trade credit links. The model offers further predictions about this phenomenon, including stronger predictability during periods of high credit constraints and low uninformed trading volume. We find supportive empirical evidence for these predictions.

Innovations in statistical technology in functions including credit‐screening have raised concerns about distributional impacts across categories such as race. Theoretically, distributional effects of better statistical technology can come from greater flexibility to uncover structural relationships or from triangulation of otherwise excluded characteristics. Using data on U.S. mortgages, we predict default using traditional and machine learning models. We find that Black and Hispanic borrowers are disproportionately less likely to gain from the introduction of machine learning. In a simple equilibrium credit market model, machine learning increases disparity in rates between and within groups, with these changes attributable primarily to greater flexibility.

In household finance markets, inactive households can implicitly cross-subsidize active households who promptly respond to financial incentives. We assess the magnitude and distribution of cross-subsidies in the mortgage market. To do so, we build a structural model of household mortgage refinancing and estimate it on rich administrative data covering the stock of outstanding mortgages in the UK. We estimate sizeable cross-subsidies that flow from relatively poorer households and those located in less-wealthy areas towards richer households and those located in wealthier areas. Our work highlights how the design of household finance markets can contribute to wealth inequality.

We build an empirical model to attribute delays in mortgage refinancing to psychological costs inhibiting refinancing until incentives are sufficiently strong; and behavior, potentially attributable to information-gathering costs, lowering the probability of household refinancing per unit time at any incentive. We estimate the model on administrative panel data from Denmark, where mortgage refinancing without cash-out is unconstrained. Middle-aged and wealthy households act as if they have high psychological refinancing costs; but older, poorer, and less-educated households refinance with lower probability irrespective of incentives, thereby achieving lower savings. We use the model to understand frictions in the mortgage channel of monetary policy transmission.

We use a comprehensive data set of funds‐of‐funds to investigate performance, risk, and capital formation in the hedge fund industry from 1995 to 2004. While the average fund‐of‐funds delivers alpha only in the period between October 1998 and March 2000, a subset of funds‐of‐funds consistently delivers alpha. The alpha‐producing funds are not as likely to liquidate as those that do not deliver alpha, and experience far greater and steadier capital inflows than their less fortunate counterparts. These capital inflows attenuate the ability of the alpha producers to continue to deliver alpha in the future.

We quantify reference dependence and loss aversion in the housing market using rich Danish administrative data. Our structural model includes loss aversion, reference dependence, financial constraints, and a sale decision, and matches key nonparametric moments, including a “hockey stick” in listing prices with nominal gains, and bunching at zero realized nominal gains. Households derive substantial utility from gains over the original house purchase price; losses affect households roughly 2.5 times more than gains. The model helps explain the positive correlation between aggregate house prices and turnover, but cannot explain visible attenuation in reference dependence when households are more financially constrained.