Essays on using high-frequency data in empirical asset pricing models.

摘要

This dissertation explores using high-frequency data in empirical asset pricing models. Since 1990s, the progress of information technology has made tick-by-tick data available in some financial markets and also allows for empirical investigations of a wide range of issues.; The first chapter provides a survey on the use, analysis, and application of high-frequency data. I concentrate on the research using intraday observations on volatility measurement and forecast evaluation, especially after the realized volatility approach introduced by Andersen and Bollerslev (1998.; The second chapter explores how to estimate betas from high-frequency data. A market model is developed and a consistent beta estimator using high-frequency returns is derived. High-frequency intraday prices on individual stocks in the Dow Jones Industrial Average and S&P 500 futures contracts over a six-year period are used in the empirical analysis. I find the sum of lead 1, 2 period beta, the contemporaneous beta, and lag 1, 2 period beta can be used as the security beta estimator. The time-varying monthly and quarterly betas are computed using this method. In-sample and out-of-sample tests indicate that time-varying betas can substantially decrease the beta measurement error, but this has limited practical value for hedging, whether for individual stocks or some portfolios considered. Further analysis shows that the security beta is a weighted average of its intraday beta and overnight beta, where the weight is determined by the variance ratio of the intraday market return to the overnight market return.; In the third chapter, I consider the problem faced by a professional investment manager who wants to track the return of the S&P 500 index with 30 DJIA stocks. The manager constructs many covariance matrix estimators, based on daily returns and high-frequency returns, to form his optimal portfolio. Although prior research has documented that realized volatility based on intraday returns is more precise than daily return constructed volatility, the manager will not switch from daily to intraday returns to estimate the conditional covariance matrix if he rebalances his portfolio monthly and has past 12 months of data to use. He will switch to intraday returns only when his estimation horizon is shorter than 6 months or he rebalances his portfolio daily.