Bayesian Hierarchical Models
Undergraduate Capstone Project
Financial information, like stock market prices, are known to be notoriously hard to predict. We wanted to take a Bayesian approach to try and tackle a similar situation: predicting the future earnings of S&P 500 companies. In this project we seek to model future earnings using other financial information about a company, like previous earnings and sales. We explore a few Bayesian hierarchical models, as well as a SARIMA model using the bayesforcast package to try and identify one that can provide insight and better predictions for future company’s earnings.