SYNOPSIS ON RESEARCH PAPERS

Ruojun (Rosalin) Wu, UCSD

____________________________________________________________________________________

Stock Return Variability, Forecast Revisions, and Investors’ Learning (Job market paper)

Understanding the source of movements in stock prices is a key question in finance. Present value models by Campbell and Shiller（1988）and others assume investors have full information and suggest that roughly 2/3 of the variation in stock returns comes from changes to the risk premium, and 1/3 from cash flow news.

This paper relaxes the full information assumption and studies the problem from a new and more realistic perspective: investors’ learning under parameter uncertainty and its implication for decomposition of variance of stock returns. Two learning schemes are examined: a naïve learning scheme where investors accept OLS estimation results in real time; and a sophisticated learning scheme where they incorporate their prior beliefs with observed data by Bayes rule. Accordingly I propose to use Bayesian VAR to simulate investors' forecast revisions from period to period and back out the variance components for this type of learning.

This new approach addresses two questions: which priors would investors have in order to make the observed return series to be consistent with the VAR process for cash flow and risk premium? Moreover, given these priors, what do they say about the relative importance of cash flow and risk premium news? Statistically, this approach also has the advantage of providing more robust finite sample inferences by imposing reasonable restrictions on return process and thus alleviates the impact from parameter instability.

In the empirical part, results under full information and under different learning schemes are presented and compared. As expected, learning increases the variability of investors’ forecasts. Moreover, as opposed to the conventional wisdom, the cash flow news contributes a significant amount to the total return volatility and it moves in the same direction as the discount rate news. Potential explanations are provided.

Out-of-sample Return Predictability under Constraints on the Equity Premium (with Pettenuzzo, Timmermann, and Valkanov, 2007, working paper)

This paper proposes a new approach for incorporating theoretical constraints on return forecasting models such as non-negativity of the conditional equity premium and sign restrictions on the coefficients linking state variables to the equity premium. Our approach makes use of Bayesian methods that update the estimated parameters at each point in time in a way that optimally exploits information in these constraints. Using a variety of predictor variables from the literature on predictability of stock returns, we find that theoretical constraints have an important effect on the coefficient estimates and can significantly reduce biases and estimation errors. In out-of-sample forecasting experiments we find that models that exploit the theoretical restrictions produce better forecasts than unconstrained models.

The Sampling Distribution of Variance Components in Stock Return Decompositions (2007, work in progress)

This work examines finite sample properties of the VAR methods used to decompose variations in stock returns into news about risk premium and cash flows.

Demographic Trends, Health Economics, and Rapid Response in Southeast Asia: Focus on Thailand and the Point of Care (with Kost, Wongboonsin, Peungposop, & Chen, published, 2003)

The authors introduce relevant demographic trends and health economics in Thailand and Southeast Asia, and emphasize the need for collaborative teamwork in arriving at acceptable strategies to fulfill societal, economic, and medical goals.