Publications and working papers

Information Aversion
Journal of Political Economy, 128.5 (2020): 1901-1939.
(with Valentin Haddad)  

Information aversion, a preference-based fear of news flows, has rich implications for decisions involving information and risk-taking. It can explain key empirical patterns on how households pay attention to savings, namely that investors observe their portfolios infrequently, particularly when stock prices are low or volatile. Receiving state-dependent alerts following sharp market downturns such as during the financial crisis of 2008 improves welfare. Information averse investors display an ostrich behavior: overhearing negative news prompts more inattention. Their fear of frequent news encourages them to hold undiversified portfolios.

Horizon-Dependent Risk Aversion and the Timing and Pricing of Uncertainty
R&R The Review of Financial Studies                                                                                                                                                               (with Thomas M. Eisenbach, and Martin C. Schmalz)  

Inspired by experimental evidence, we amend the recursive utility model to let risk aversion decrease with the temporal horizon. Our pseudo-recursive preferences remain tractable and retain appealing features of the long-run risk framework, notably its success at explaining asset pricing moments. Calibrating the agents’ preferences to explain the market returns observed in the data no longer implies an extreme preference for early resolutions of uncertainty; and captures key puzzles in finance on the valuation and demand for risk at long maturities

Risk pricing under gain-loss asymmetry
(Previously circulated as Consumption-based Asset Pricing with Loss Aversion)                                                         
Online Appendix 

When agents evaluate risk with gain-loss asymmetry -- losses relative to a reference point incur discontinuously more disutility than comparable gains -- a formal closed-form analysis reveals two separate asset pricing implications. First, a level effect: risk prices are made higher by the kink in the preferences. Second, a cross-sectional effect: the pricing of risk is higher (lower) for safer (riskier) assets, so expected returns increase non-linearly with the risk-exposures. This second effect, a crucial departure from standard smooth utility models, can rationalize key puzzles in empirical finance.


Work in progress

Ambiguous Trade-Offs: An Application to Climate Change
(with Nina Boyarchenko)

We study optimal long-term investment choices in settings where agents face ambiguity about both the future benefit and the current cost, as is likely to be the case for large scale social programs, such as healthcare choices and climate change policies. Faced with this kind of ambiguity, rational economic agents optimally choose investment paths that are observationally equivalent to choices made under hyperbolic discounting. Using calibrated paths of potential output losses under different global warming scenarios, we evaluate the relative attractiveness of small-scale, large-scale and R&D projects for mitigating climate change. 


Forecasts, Extrapolation and Portfolio Allocation when Returns are Predictable
(with Milo Bianchi, Khanh Huynh and Sebastien Pouget)

We design an experiment to study how investors form their forecasts and risk al- locations under different market conditions. We let our subjects observe not only the past realizations of a risky asset but also a signal a that, in some rounds, helps predict its future returns. Subjects correctly identify the predictable rounds 80.7% of the time. When they do not perceive a as useful information, their forecast "model" is extrapolative: subjects irrationally use the most recent returns realizations to predict future i.i.d outcomes. However, when they perceive returns as predictable by the signal a, they switch to a fully rational forecast model. In all cases, subjects rely on their own expectations when choosing their risky investments, particularly so when they view a as predictive; but the elasticity of investments to forecasts is low, indicative of high implicit risk aversions – a puzzle with respect to the relatively large average risk allocations subjects opt for.

The Term Structure of the Price of Volatility Risk
(with Thomas M. Eisenbach, Martin C. Schmalz, and Yichuan Wang)

We estimate the term structure of the price of variance risk (PVR), which helps fill a gap in the empirical literature and distinguish between competing asset-pricing theories. First, we measure variance premia using the Sharpe ratios of short-term holding returns of at-the-money index straddles, and find they are negative and upward sloping in the term-structure. Second, to distinguish price versus quantity in the variance premia, we estimate the PVR in a Heston (1993) model separately for different maturities. We find the PVR contributes to the shape of the term- structure of variance premia: it is negative and decreases in absolute value with maturity, the more so when volatility is high. These findings are inconsistent with the calibrations of established asset-pricing models that assume constant risk aversion across maturities, but confirm a key prediction of the model with horizon-dependent risk aversion of Andries et al. (2014). 

Changes in the Risk-free Rate: Evaluating Asset Pricing Models
(with Jean-Guillaume Sahuc)

We evaluate how well the most commonly used asset pricing models in macro-finance can accommodate the observed changes in the risk-free rate over the last three decades, i.e. from a rate of 4% in 1990 to essentially 0% today (10-year real rates for both the US and Europe). We study the benchmark models of habit (Campbell and Cochrane, 1999), long-run risk (Bansal and Yaron, 2004; Bansal et al., 2009), and rare disasters (Barro, 2006; Gabaix, 2012). Our analysis not only informs how well such models perform “out-of-sample”, i.e. in the near-zero interest rates current era. It can also help in identifying whether macro-economic conditions experienced a (permanent) regime switch in the past few years, or if the observed changes constituted “business as usual” time variations

Social Responsibility and Asset Prices: Is There a Relation?