This repository contains all the files for 6 lectures and 4 assignments for the unit FINA4590 Research Methods in Accounting and Finance.
The first lecture introduces the practical arrangement of the unit and explains the assessments. I introduce the research process with CEO compensation data for the largest public US firms. I also provide a number of resources for the literature review and a first introduction to R, Rstudio, and Rmarkdown.
- Dates of assessments for 2022
- Add resources
- Section on workload management
The second lecture revisits the CEO compensation data and a new theory to predict the relation between firm size and total CEO compensation. We also look at the incentives of CEOs. The main question I focus on wether CEOs are paid for every $ in value creation or for a % in value creation. This leads to hypotheses about the relation between the sensitivity of pay to performance. I introduce how we can statistically test these hypotheses.
The section discusses variables we want to control for (confounders) and the specific controls of fixed effects. It also discusses the problem of bad controls (colliders). Finally, I introduce two stage least squares with an instrumental variable as an alternative.
- Update the initial example for DAGs with an example from the seminar
- Update IV diagnostics
- Explanation of negative Rsquared with IV
- Tiange's study as event study
- Clustered Standard Errors