Essays on liquidity constraints and stock price behavior

This dissertation has two main parts. The first part studies the preferences of liquidity-constrained agents and the implications of liquidity constraints for portfolio allocation at the household level. The second part deals with insider trading at the Mexican Stock Exchange.; The first essay evaluates preference heterogeneity between liquidity-constrained and unconstrained agents in the US economy. This essay finds evidence of a correlation between being liquidity constrained and having a lower discount factor (i.e., being “impatient”). Evidence on differences in risk aversion is, however, not conclusive. In the context of life-cycle portfolio decisions, these findings imply that liquidity-constrained agents prefer to have a smaller share of risky assets in their financial portfolios. Thus, preference heterogeneity can reinforce the negative effects of liquidity constraints on holdings of risky assets that the literature has found.; In the second essay, it is documented that liquidity-constrained households hold a smaller proportion of their portfolio in stocks and risky assets than unconstrained agents. It presents an empirical investigation that extends previous research by looking at three measures of income risk and two measures of liquidity constraints. It is found that liquidity constraints are an important factor for portfolio allocation, after controlling for demographic, financial and preference factors. In addition, background risk matters. The probability of having no income is, among the measures used, the most important determinant of the portfolio decision.; Contrary to the media's claims that insider trading is common practice at the Mexican Stock Exchange, the last essay finds statistical evidence of insider trading only in a small number of the corporations under analysis. The media's claims are based on evidence presented in prior research. Previous studies, however, have not dealt explicitly with the endemic problem of missing observations. The third essay uses Monte Carlo simulations in order to deal with the nuisance of missing data.