ESSAYS ON MICROECONOMETRICS AND FINANCE

Abstract

This dissertation includes four chapters which are four separate papers on Mircoeconometrics and Finance. The first two chapters establish estimators which are useful to study distributional effects of a continuous treatment and local elasticities, respectively. The tools then are applied on to intergenerational income mobility. Thus, Chapter 1 and 2 are on Microeconomics. Chapter 3 and 4 are on Finance. In particular, I apply the Oaxaca-Blinder decomposition, distribution regression and recentered influence function regression to decompose the portfolio returns between North America and Europe. Chapter 1, titled DISTRIBUTIONAL EFFECTS OF A CONTINUOUS TREATMENT WITH AN APPLICATION ON INTERGENERATIONAL MOBILITY (with Brantly Callaway), considers the effect of a continuous treatment on the entire distribution of outcomes after adjusting for differences in the distribution of covariates across different levels of the treatment. Our methodology encompasses dose response functions, counterfactual distributions, and ``distributional policy effects'' depending on the assumptions invoked by the researcher. We propose a three-step estimator that consists of (i) estimating the distribution of the outcome conditional on the treatment and other covariates using quantile regression; (ii) for each value of the treatment, averaging over a counterfactual distribution of the covariates holding the treatment fixed; (iii) manipulating the counterfactual distribution into a parameter of interest. We show that our estimators converge uniformly to Gaussian processes and that the empirical bootstrap can be used to conduct uniformly valid inference across a range of values of the treatment. We use our method to study intergenerational income mobility where we consider distributional effects of parents' income on child's income such as (i) the fraction of children with income below the poverty line, (ii) the variance of child's income, and (iii) the inter-quantile range of child's income -- all as a function of parents' income. Chapter 2, tiled LOCAL INTERGENERATIONAL ELASTICITIES (with Brantly Callaway), proposes a ``local'' intergenerational mobility parameter (LIGE) that allows the effect of parents' income to vary across different values of parents' income. We also extend this result to an ``adjusted'' local intergenerational elasticity (ALIGE) which adjusts for differences in the distribution of observed characteristics at different values of parents' income. We develop the asymptotic properties of the LIGE and ALIGE, and apply them to study intergenerational mobility using data from the PSID. We find that the intergenerational elasticity is much larger for low values of parents' income (indicating \textit{less} mobility) relative to high values of parents' income; adjusting for differences in characteristics reduces the local IGE at all values of parents' income as well as flattening it across different values of parents' income. Chapter 3, titled DECOMPOSING DIFFERENCES IN PORTFOLIO RETURNS BETWEEN NORTH AMERICA AND EUROPE, decomposes differences in mean and a series of quantiles of portfolio returns between North America and Europe into Fama and French's five factors. We show that the differences in risk premia on factors, especially on market and size factors, account for most of the differences and the differences in factor risks seem to play an insignificant role in aggregate. The results from Blinder-Oaxaca decomposition show that the differences in market and size factor risk premia explain 71.9\% and 22.8\% of the overall mean difference, respectively. We also show that the roles that the risk premia on market and size factors play vary at different levels of portfolio returns, implying the market and size factor risk premia vary at different levels of portfolio returns. Also, we find that the risks on some factors seem to vary at different levels of portfolio returns. Chapter 4, titled DECOMPOSING DIFFERENCES IN QUANTILE PORTFOLIO RETURNS BETWEEN NORTH AMERICA AND EUROPE USING RECENTERED INFLUENCE FUNCTION REGRESSION, decomposes quantile portfolio returns using recentered influence function regression. Chapter 1 decomposes the differences in quantile portfolio returns using distribution regressions. The main issue of using distribution regressions is that the decomposition results are path dependent. In this paper, we can obtain path independent decomposition results by combining the Oaxaca-Blinder decomposition and the recentered influence function regression method. We show that aggregate composition effects are all positive across quantiles and the market factor is the most significant factor which has detailed composition effect monotonically decreasing along quantiles. The main decomposition results are consistent with Chapter 3.