Working Papers

"The Risk of Risk-Sharing: Diversification and Boom-Bust Cycles" [Draft]

In this paper, I model a shock whereby financial intermediaries can better diversify borrowers’ idiosyncratic risks. A sector-specific diversification improvement induces intermediaries to reallocate funds toward the shocked sector. As lending spreads fall, intermediaries build up leverage over time. The result is a fragile sectoral boom that can end in an economy-wide bust. This cycle is amplified if the diversification-shocked sector is higher-risk or more external-finance dependent. I apply the model quantitatively to the recent housing cycle. Feeding in a novel mortgage diversification index, the model generates the measured increase in household credit coincident with declining mortgage spreads that spike during an economy-wide bust.

"Entry and Slow-Moving Capital: Using asset markets to infer the costs of risk concentration" [Draft] (R&R at Journal of Financial Economics)

Risk concentration is a major outstanding explanation for crisis dynamics of asset prices and macroeconomic quantities. Apparently, capital flows are slow to correct these crises. By considering costly entry in a canonical limited participation model, I illustrate how asset prices encode costs of risk concentration. These costs must be enormous to match risk premia levels and variability. This finding is robust: auxiliary features that increase risk premia levels mitigate their dynamics, through endogenous entry. In short, either entry costs are large, or limited risk-sharing arises for other reasons. One appealing possibility is extrapolative expectations, which complements entry well.

"Financial Frictions and Aggregate Fluctuations" [Draft]

Concentrated idiosyncratic risk positions may generate aggregate fluctuations. I study a canonical macroeconomic model with a standard moral hazard friction but with a single innovation: fundamental shocks are correlated (but still aggregate to zero). Experts hold concentrated asset positions, while less productive households hold diversified positions in experts' equity. I prove that aggregate output and the wealth distribution have aggregate volatility if and only if observability and contractibility are imperfect. This failure of the law of large numbers holds generically and does not require any assumptions about fat-tailed size distributions. Even though aggregate volatility disappears with perfect contractibility, it can increase with partial contractibility improvements, due to market segmentation between experts and households. These results are immune to allowing agents to frictionlessly hedge the endogenously-arising aggregate shocks.

"Comparative Valuation Dynamics in Models with Financing Restrictions" with Lars Hansen and Fabrice Tourre [Draft coming soon] [Slides] [Model Comparisons Toolbox] [Numerical Method Draft also coming soon]

This paper develops a theoretical framework to nest many recent dynamic stochastic general equilibrium economies with financial frictions into one common generic model. Our goal is to study the macroeconomic and asset pricing properties of this class of models, identify common features and highlight areas where these models depart from each other. In order to characterize the asset pricing implications of this family of models, we study their term structure of risk prices and risk exposures, the natural extension of impulse response functions in economic environments exhibiting non-linear behaviors.

"Commonality in Credit Spread Changes: Dealer Inventory and Intermediary Distress" with Zhiguo He and Zhaogang Song [Draft]

Two intermediary-based factors - an intermediary financial distress measure and a dealer corporate bond inventory measure - explain about 50% of the puzzling common variation of credit spread changes beyond canonical structural factors. A simple model, in which intermediaries facing margin constraints absorb supply of assets from customers, accounts for the documented explanatory power and delivers further implications with strong empirical support. First, the effect of intermediary factors remains the same across bonds with characteristics not tied with margin requirement. Second, dealer inventory affects prices of assets within the corporate credit market only, whereas intermediary distress affects even the non-corporate-credit market. Third, an important component of dealers' inventory change is tied with supply shock of (severely) downgraded bonds by institutional investors, e.g., insurance companies. Instrumenting the effect of dealer inventory on bond prices using institutional sales of "fallen angels", as well as insured losses due to natural disasters, supports this interpretation.

Works in Progress

"Market Power as Skin-in-the-Game" with Jung Sakong

We show that, when households have mistaken beliefs or less information, intermediary market power can limit over-investment, improve resource allocation, and reduce asset price volatility. If intermediaries are compensated based on their returns, market power increases their compensation, leading to better incentives to invest according to the social good. The results are analogous if intermediaries are compensated based on assets under management in a dynamic world: market power creates skin-in-the-game incentives, even if they are absent in the contract. We devise a series of empirical tests, in both finance and non-finance arenas, as to the interacting effects of household irrationality and intermediary competition.