• Given 5155 stocks factors (style, alpha, and industry factors) data and stocks returns from 2007 to 2017, computed factor returns using OLS and WLS;
• Wrote EM algorithm in solving linear mixed-effect model and applied it to compute factor returns;
• Constructed constrained and unconstrained Markowitz portfolio by mean-variance optimization, calculated its daily cost using simple market impact model, and decomposed risks due to different factors in this portfolio;
• Constructed portfolio using Black and Litterman’s framework with one single view about its expected excess return from benchmark cap-weighted portfolio;
• Simulated a mean-reverting security price, designed and implemented simple Q-learning system to exploit the arbitrage.