The three chapters in this dissertation examine issues related to liquidity and asset pricing.

In the first chapter, I examine the equity premium puzzle by incorporating transactions costs into the Campbell and Cochrane (1999b) habit formation model. The objective of this chapter is to determine whether the incorporation of such costs can sufficiently reduce the required level of relative risk aversion while at the same time retaining the main properties of Campbell and Cochrane (e.g., high equity premium, low riskless rate, and the market Sharpe ratio). I find that incorporation of transactions costs reduces the required level of relative risk aversion at the steady state from 35 to 15. Thus, I conclude that transactions costs seem able to reduce, but not completely solve, the equity premium puzzle.

In the second chapter, I examine the determinants of liquidity covariances risks. In particular, I closely examine the large return premium associated with the covariation between firm liquidity costs and the market returns. Using a standard return decomposition (Campbell and Shiller, 1988b), I show that the liquidity premium comes from two sources: covariation of liquidity costs with (i) macroeconomic shocks and (ii) market risk premium shocks. In 1964–2017 US stock market data, both components are priced but the expected return premium associated with the former is approximately three times larger than that for the latter. Liquidity volatility is primarily incorporated in stock prices via its common variation with macroeconomic shocks.

In the third chapter, which is motivated by the second chapter, I construct six liquidity return factors based on the stock liquidity covariance risks (i.e., the macroeconomic shock beta and the financial shock beta) and examine the performance of liquidity-augmented factor models. Consistent with the second chapter, liquidity risks are priced and the liquidity augmented factor pricing model performs no worse than the five factor model of Fama and French (2015).

The third chapter has two main findings. First, the liquidity return factors capture additional risks that other Fama-French factors do not: the risk-adjusted returns associated with these factors are large and statistically significant. Second, after controlling for liquidity effects, the risk-adjusted return for the Fama-French size factor (SMB) is eco- nomically small and statistically insignificant, suggesting that liquidity effects can explain the size effect (but not the other way around).