Transaction costs, liquidity, and asset pricing

Liquidity is one of the most important concepts in finance. Proxies of liquidity are transaction costs. The rigorous studies of transaction costs are only at the beginning due to technical complexity. My dissertation studies market liquidity and transaction costs, including fixed costs, proportional costs, and market impact costs for informed traders from several perspectives using continuous-time models. I have developed new analytical and numerical methods to obtain solutions. My dissertation is composed of three chapters. In particular, the third chapter explains several financial anomalies in an integrated framework.; The first chapter studies the value of private information in the presence of transaction costs by considering the optimal intertemporal consumption and investment policy of an active investor who observes a mean-reverting private signal process. The paper shows that the value of private information is determined mainly by the expected squared Sharpe ratio (SR) and the speed of mean-reversion of the private signal under transaction costs. We show that a higher SR leads to more aggressive trading and a larger net value (better performance), while a shorter horizon leads to more aggressive trading but a smaller net value (poorer performance). Hence, more trading activity does not always result in better performance. For a given SR, when the speed of mean-reversion of the signal is larger than a threshold, the best strategy is "buy-and-hold" and there is no information advantage in the presence of transaction costs. In addition, transaction costs cause the orders of the investor to be positively autocorrelated. Our model also has implications for the underperformance of mutual funds.; The second chapter presents an infinite horizon model of strategic informed trading when private information and pubic dividend announcements follow a VAR process. There are three types of traders: a risk-averse informed trader whose trading incurs trade-size related costs, competitive risk-neutral market makers, and liquidity traders with possible overreaction (underreaction) to dividend announcements. I demonstrate that there exists a linear equilibrium. The price change depends on the entire path of order flows and dividend announcements surprises and a public news surprise initiates informed trading. This paper shows that as public news becomes more intense, the bid-ask spread, the instantaneous return variance and the trading volume all increase. The orders of the informed trader are positively autocorrelated over a short period of time but negatively autocorrelated over a longer period of time. In addition, after a shock, the informed trader is trend-chasing in the short run, while he is a contrarian in the long run. Market depth is constant and can be negative in some cases. Our result can also be used to explain the observation that hedge funds perform relatively better in more volatile situations.; The third chapter is an extension of the second one. In this chapter, the market makers are risk verse. The paper shows that trading cost and market impact cost lead to the short-term positive autocorrelation of stock returns (momentum), that the underreaction of liquidity traders to public news and the momentum trading of the informed trader lead to short-term post-earnings drift, and that the risk-aversion of the informed trader and the mean-reverting properties of private information contribute to the long-term negative autocorrelation of stock returns (reversal) and to long-term opposite post-earnings drift. In addition, our model shows that the variance of order flow contributes to the high volatility of stock returns relative to fundamental values. The model also has implications for market liquidity, trading volume, price volatility, and the incentives for traders to collect private information.