Recent progress in causal inference has opened a path, however difficult, for advancing financial economics beyond its current phenomenological stage. The goal of this article is to propose a hierarchy of empirical evidence, recognizing that not all types of observations have the same scientific...

In a number of applications, data may be anonymized, obfuscated, or highly noisy. In such cases, it is difficult to use domain knowledge or low-dimensional visualizations to engineer the features for tasks such as machine learning, instead, we explore dimensionality reduction (DR) as a...

Several features of financial research make it particularly prone to the occurrence of false discoveries. First, the probability of finding a positive (profitable investment strategy) is very low, due to intense competition. Second, true findings are mostly short-lived, as a result of the...

Investing can be characterized as a data science problem. While investment firms have attracted scientific talent, they have done a poor job at developing it. Firms hire specialists, but entice them to become generalists (e.g., portfolio managers). Under the ubiquitous silo/platform structure,...

Finance cannot become a rigorous science (in the Popperian or Lakatosian sense), however it can still operate as an “industrial science”. This article describes the scientific method by which industrial finance discovers through experimentation, and avoids false discoveries

Most firms and portfolio managers rely on backtests (or historical simulations of performance) to select investment strategies and allocate them capital. Standard statistical techniques designed to prevent regression over-fitting, such as hold-out, tend to be unreliable and inaccurate in the...

A substitution effect takes place when two or more explanatory variables share a substantial amount of information (predictive power).Under the presence of substitution effects, feature importance methods may not be able to determine robustly which variables are significant.This presentation...

Many problems in finance require the clustering of variables or observations. Despite its usefulness, clustering is almost never taught in Econometrics courses. In this seminar we review two general clustering approaches: partitional and hierarchical

Two random variables are codependent when knowing the value of one helps us determine the value of the other. This should not me confounded with the notion of causality.Correlation is perhaps the best known measure of codependence in econometric studies. Despite its popularity among economists,...

Traditionally, the development of investment strategies has required domain-specific knowledge as well as access to restricted datasets. This has meant that investment opportunities are not researched by the majority of data scientists, because they lack either or both of these requirements. In...