Reference-dependent Preferences and Sentiment-driven Asset Prices, with Jess Benhabib, Xuewen Liu, and Pengfei Wang
R&R at Journal of Finance
This paper studies asset pricing under expectations-based reference-dependent preferences in a general equilibrium framework. Our model shows that reference-dependent preferences can generate sentiment-driven asset prices that are impossible in a standard expected utility based model of asset pricing. The model helps to explain empirical puzzles in asset pricing such as i) excess volatility, ii) asymmetric volatility, iii) asymmetric sentiment over the business cycle, iv) excess asset price comovement, and v) weak correlation between stock returns and economic fundamentals along with sizable equity premium.
Information Acquisition and the Finance-Uncertainty Trap, with Ding Dong, Allen Hu, and Zheng Liu
Presentation: Stanford SITE 2025; FRBSF; HKU; HKBU; HNU; PHBS; SDU; SJTU; CES China 2025
Using novel measures of information acquisition, we document causal evidence of a feedback loop between firms’ credit access and information acquisition. To examine the macroeconomic implications of this feedback loop, we develop a tractable general equilibrium framework with financial frictions and endogenous information acquisition. In line with the empirical evidence, the model predicts that a rise in information costs raises the level of uncertainty and reduces a firm’s equity value, hampering its credit access. On the other hand, tightened credit constraints restrain activity of high-productivity firms, leading to misallocation that reduces aggregate productivity and firm profits, and discouraging information acquisition. This feedback loop creates a finance-uncertainty trap that substantially amplifies and prolongs business cycle fluctuations.
A Reference-Dependent Fiscal Theory of Self-Fulfilling Price Levels (new draft coming soon) , with Xuewen Liu, Thomas J. Sargent, and Pengfei Wang
Presentation: SIQEF Workshop on the Frontier of Macroeconomic Research 2025; WESEAMS 2025; SDU
Under expectation-based reference-dependent preference, even without shocks to the primary surplus, the investors' concerns about changes in the value of its holdings of government debt can generate self-fulfilling sunspot-driven fluctuations in the discount rates and consequently inflation. Our model accounts for key emprical patterns, including asymmetric volatility of inflation and nominal yields, an upward-sloping yield curve, and the return predictability in the term structure of interest rates. A fiscal policy that pushes down the nominal stock of government bonds when inflation uncertainty increases can preclude the existence of sunspot equilibrium.
Monetary Policy in a Data Economy (new draft coming soon) , with Zhiwei Xu, and Francesco Zanetti
Presentation: 8th HenU/INFER Workshop; SWUFE International Conference on Macro Finance; SYSU
We examine monetary policy in a data economy, where data emerges as a by-product of economic activities and provides valuable information for forecasting product demand. We analytically derive a data-economy version of the New Keynesian Phillips curve that incorporates endogenous factors tied to data market conditions, yielding a novel supply-side channel for monetary policy. We show that endogenous data accumulation dampens the responsiveness of inflation while amplifying output dynamics under monetary easing, and provide supporting empirical evidence. We further derive a welfare criterion tailored to the data economy and solve for the optimal commitment policy in the data economy.