This dissertation presents two essays in financial econometrics. These chapters show how to adapt and use econometric techniques to study various questions in financial economics. The first chapter: Revealed Preference for Green Stocks: An Asset Demand Approach'' combines a traditional portfolio construction problem with demand estimation techniques to estimate the demand for green stocks of US institutional investors. The methodology presented innovates along two dimensions with respect to recent influential work on asset demand estimation. First, in this framework investors have heterogeneous portfolios not only through differential beliefs about future returns, but also because they place varying importance on the non-financial characteristics of the portfolios they construct. Second, by using a mixed logit demand specification, we can estimate asset demand that delivers more realistic substitution patterns across assets. This chapter uses data on the environmental performance of firms and quarterly stock holdings data from institutional investors to estimate the demand for stocks accounting for environmental scores and return-related stock characteristics. Estimates show that taste for green stocks fluctuates over time and by investors' assets under management. Finally, this chapter presents a counterfactual exercise to study the equity price effects of a ban on green investing for pension funds. Results show that a portfolio with the top brown stocks is estimated to have positive capital gains due to the policy, while a portfolio with the top green stocks is estimated to have capital losses. The second chapter: Exchange Rate Supervised Topic Modelling,'' shows how to use a hybrid of supervised and unsupervised learning models to go from text from news articles to an FX news index that can be used to enhance traditional models from the FX literature. To do so we rely on Supervised Latent Dirichlet Allocation (sLDA) which combines information about a supervising variable with topic extraction over a corpus of text in a single-stage estimation. Although this estimation can be done in two stages, this chapter documents with a Monte Carlo simulation that there are efficiency gains from a single-stage approach. The empirical application suggested is centered around the Monex Market, the main Costa Rican platform for FX trade; accordingly news articles are gathered from the main Costa Rican newspapers. The exchange rate of interest is the Costa Rican Colón (CRC), the local currency, and the United States dollar (USD). Using the CRC/USD exchange rate as the supervising variable this chapter relies on sLDA to extract the topics from the news article corpus that are relevant as covariates for the exchange rate over short frequencies. Results provide evidence that these topics have explanatory power in and out of sample, and can be used to provide interpretability to episodes of high volatility of the exchange rate.