Essays in Asset Pricing and Microstructure of Fixed Income Markets

Abstract

This dissertation consists of three essays on asset pricing and market microstructure topics within the U.S. corporate bond market. The first essay investigates asymmetry in price pressure between customer buy and sell orders and demonstrates that it is a valuable measure of downside liquidity for corporate bonds. While evidence of a characteristic premium for illiquidity in the cross-section of corporate bonds is mixed, aggregate liquidity asymmetry has high explanatory power for the time series of market returns. Its statistical and economic significance justify it as a credible asset pricing factor. Average market-wide liquidity asymmetry comoves with interest rate and credit spread changes, investor sentiment, funding liquidity, dealer inventory, exchange-traded fund flows, and post-crisis regulatory change. The second essay documents the properties of market-wide corporate bond liquidity and demonstrates that liquidity risk is an important determinant of returns. In market downturns, transaction costs rise for sellers and fall for buyers. The negative relation between buyer and seller liquidity motivates a new across-measure liquidity factor that incorporates an asymmetric liquidity component. Shocks to market-wide liquidity explain a large portion of bond return variation in the time series. Primarily driven by the asymmetric component, the liquidity factor attracts a cross-sectional risk premium that is robust to controls for credit, equity, and interest rate factors, as well as the illiquidity level. The third essay provides new evidence of retail investors’ ability to predict returns based on transactions in U.S. corporate bonds with equity-like risk. Retail order flow is persistent and contrarian, and it predicts future returns in the cross-section. The profits of an equal-weighted, long-short strategy that buys (sells) bonds that experience high (low) net retail buying are economically meaningful. The alpha based on decile portfolios is significant at the 10% level when controlling for common equity and bond risk factors. However, due to high transaction costs and because retail purchase volume is concentrated in underperforming bonds, retail traders lose money in aggregate.