Subjective beliefs and risk attitudes
My recent work on Learning from Inflation Experiences with Ulrike Malmendier (UC Berkeley) studies individual's formation of beliefs about macroeconomic variables. We show that individuals learn from experience when they form their beliefs: As cohorts accumulate different
experiences of past inflation, their beliefs about future inflation differ. In the paper Depression Babies: Do Macroeconomic Experiences Affect Risk-Taking? we show that accumulated experiences of past stock and bond returns affect household portfolio allocation decisions and risk-taking. Learning from experience implies that individuals extrapolate from their quite limited historical experience and that learning dynamics are perpetual. In ongoing work, we currently examine subjective beliefs about other macroeconomic variables and asset-return expectations.
Liquidity in financial markets tends to evaporate during times of crises. In the paper Evaporating Liquidity I show that this disappearance of liquidity is associated with an enormous rise in the expected returns from liquidity provision, reflecting reluctance to absorb inventory on the part of market makers. In times of high equity market volatility, as measured by the VIX index,
the profitability of short-term reversal strategies, a proxy for the returns from liquidity provision, rises dramatically. In recent work on Sizing Up Repo with Arvind Krishnamurthy (Northwestern) and Dmitry Orlov (Stanford), we provide the first assessment of funding quantities in the tri-party repo market -- an important funding market in which funding flows from money market funds and other cash lenders to broker-dealers -- prior to and during the financial crisis. In contrast to the common notion that a "run on repo" was central to the financial crisis, we find that most repos prior to the financial crisis were backed by safe collateral and not subject to a run. In other ongoing work, I examine the causes of time-variation in the liquidity premium that financial market participants are willing to pay for highly liquid "money-like" assets such as Treasury Bills. In current work with Arvind Krinshamurthy and Annette Vissing-Jorgensen, we examine the financial market effects of the European Central Bank's interventions in the recent European debt crisis.
Building on my earlier work on the conditional CAPM with Jonathan Lewellen and on the pitfalls in assessing the cross-sectional fit in asset pricing tests with Jonathan Lewellen and Jay Shanken, the paper Estimation and Evaluation of Conditional Asset Pricing Models with Ken Singleton shows that a number conditional asset pricing models that do well in fitting unconditional average returns on size and book-to-market sorted portfolios do so by producing strongly counterfactual predictions about the time-series of conditional moments. For example, the model parameter estimates obtained from fitting to cross-sectional data imply extreme time-variation in risk premia, much more than is actually found in the data. In ongoing work with Arthur Korteweg (Stanford), I focus on applying stochastic discount factor valuation methods and GMM estimation to performance evaluation of venture capital payoffs.