Essays in Empirical Asset Pricing: Rollover Risk, Liquidity and the Financial Crisis

This thesis consists of three essays in empirical asset pricing, focusing on rollover risk, liquidity and the financial crisis. In the first chapter, I investigate the asset pricing implication of rollover risk, the risk that a firm might not be able to refinance its due liabilities. I focus on the long-term debt rollover risk and use the proportion of long-term debt falling due within the year to measure a firm's exposure to rollover risk. I find that firm-specific rollover risk coupled with deteriorating market wide credit conditions indeed increase a firm's credit spreads. During the peak period of the financial crisis, namely the second half of 2008, one-year CDS spreads for a firm that needs to refinance more than 20% of its long-term debt are 72 basis points higher than the spreads for a similar firm with low refinancing needs. The differences are smaller, around 32 basis points, but still significant at the early and final stages of the financial crisis. Compared to one-year CDS spreads, CDS spreads for longer maturities are also sensitive to rollover risk information but at lower magnitudes throughout the financial crisis. During normal periods, CDS spreads are mostly explained by fundamental risk while rollover risk is not a significant determinant of credit spreads. Similar rollover risk effect is also observed in other financial markets, including the corporate bond, stock and option markets.;In the second chapter (coauthored with Jun Pan and Jiang Wang), we propose a broad measure of liquidity for the overall financial market by exploiting its connection with the amount of arbitrage capital in the market and the potential impact on price deviations in US Treasurys. When arbitrage capital is abundant, we expect the arbitrage forces to smooth out the Treasury yield curve and keep the dispersion low. During market crises, the shortage of arbitrage capital leaves the yields to move more freely relative to the curve, resulting in more "noise." As such, noise in the Treasury market can be informative and we expect this information about liquidity to reflect the broad market conditions because of the central importance of the Treasury market and its low intrinsic noise---high liquidity and low credit risk. Indeed, we find that our "noise" measure captures episodes of liquidity crises of different origins and magnitudes and is also related to other known liquidity proxies. Moreover, using it as a priced risk factor helps explain cross-sectional returns on hedge funds and currency carry trades, both known to be sensitive to the general liquidity conditions of the market.;In the third chapter (coauthored with David R. Kuipers and Yong Zeng), we propose a general nonlinear filtering framework with marked point process observations incorporating other observable economic variables for ultra-high frequency (UHF) data. The approach generalizes several existing models and provides extensions in new directions. We derive filtering equations to characterize the evolution of the statistical foundation such as likelihoods, posteriors, Bayes factors and posterior model probabilities. Given the computational challenge, we provide a powerful convergence theorem, enabling us to employ the Markov chain approximation method to construct consistent, easily-parallelizable, recursive algorithms to calculate the fundamental statistical characteristics and to implement Bayesian inference in real-time for streaming UHF data. The general theory is illustrated by a specific model built for U.S. Treasury Notes transactions data from GovPX. We show that in this market, both information-based and inventory management-based motives are significant factors in the trade-to-trade price volatility.