Three essays in financial economics
Chapter 1 posits that the market maker infers private information from the unexpected duration between trades in adjusting bid-ask quotes. Subsequently, I analyze the intra-day informational effects of the surprise in duration on bid-ask spreads and depths. The main findings are that unexpected early trades increase the quoted bid-ask spread and the adverse-selection component of the spread and that unexpected durations of buyer-initiated trades have a stronger effect on the spread than those of seller-initiated trades. I also find weak evidence that unexpected durations of trades increase quoted depths. Overall, the results on bid-ask spreads are consistent with the theories in Easley and O'Hara (1992) and Diamond and Verrecchia (1987) and complementary to the effect on the mid-quote revision in Dufour and Engle (2000). The result on depths suggests that the overall liquidity impact is more complicated than previously thought.
Chapter 2 tests the extant ex-dividend day theories with the Exchange-Traded Fund (ETF), which tracks a basket of securities and is typically very actively traded on the secondary market. Low transaction costs and risk help to differentiate between the tax hypothesis and the short-term trader hypothesis, the two dominant views of ex-dividend day behavior. The main findings are that the share price falls by the dividend amount on the ex-dividend day and that there is significant abnormal trading volume on the same day. The result on the price drop stands in contrast to that for common stocks, but the result on the abnormal volume is consistent with that for common stocks. Overall, the results are inconsistent with the tax hypothesis but support the short-term trader hypothesis.
Chapter 3 examines the quantitative effects of inflation and conservative accounting on the measurement of market-to-book ratio in a simple simulation framework. I focus on faster accounting depreciation as the primary source of conservative accounting. The main finding from the simulated data is that accounting bias is more important in accounting for the upward bias in the market-to-book ratio than inflation, but inflation and accounting bias have little effect on the correlation between the perfectly measured market-to-book and the one measured with error.