Dissertation Chapters

“Why They Buy: Primary Market Demand for U.S. Treasury Securities”
Download (Job Market Paper)
Abstract: The underpricing of U.S. Treasury auctions relative to their secondary market post-issuance price is lucrative to investors, but costly to the Treasury. The simple trading strategy that buys 2-year notes from the primary market and sells them at the secondary market price upon security issuance results in average excess returns of 10.0 bps, daily returns of 3.1 bps, and annualized returns of 7.83%; a higher yield than any 2-year note in the sample. In contrast, daily return volatility is more than 7 bps, so while the strategy is profitable, it also carries considerable risk. The contemporary view is that these returns result from supply shocks. However, I show that volatility explains the bulk of these returns, and argue that underpricing is in fact compensation for risk. In addition, primary market demand is sensitive to changes in expected risk-adjusted returns, highlighting the important relationship between underpricing and the ability to issue debt.
“Empirical Evidence for Carry Trade Liquidity Spirals”
Abstract: This paper provides an empirical test in support of carry trade liquidity spirals that result from the mutually reinforcing characteristics of market liquidity and funding liquidity. Volatility increases with market illiquidity, and when markets are illiquid, reductions in funding liquidity can result in liquidity spirals, which increase the negative skewness of speculative asset returns. Conversely, reductions in funding liquidity can reduce market liquidity and induce liquidity spirals, which increase volatility and negative skewness. It is well known that many financial asset return processes exhibit volatility clustering, which can imply an important dependence in the return time path. Since market liquidity and funding liquidity can be mutually reinforcing, liquidity spirals may be related to the dependence in the asset return time path. If liquidity spirals are related to the time dependence, then skewness should also be related. This paper tests this conjecture with a bootstrap or resampling method and rejects in favor of liquidity spirals and time dependence in carry trade returns.
“A Demonstration of Time Varying Volatility and Invalid t-Statistics”
Abstract: This paper demonstrates that using ordinary least squares (OLS) to estimate a model that exhibits time varying volatility can result in invalid t-statistics and a rejection of the true data generating process. Artificial data is simulated for ten separate models that all have the same linear relationship but different volatility structures. Parameters are estimated for each model using both OLS with Huber-White heteroskedasticity consistent standard errors (robust standard errors) and an alternative model that compensates for the time varying volatility. In each case, OLS with robust standard errors rejects the true data generating process by more than the allowed p-value. This result indicates that the efficiency gains from modeling volatility may reverse contemporary findings in economics and finance.