Quantitative Researcher
CurrentMy time is divided between exploring/devising signals for traders, and tackling ad-hoc quantitative problems for either clients or the trading desk. Some projects I've done recently include: -Built a covariance matrix of all IG bonds using a nested factor model which leverages the fundamental sector classification hierarchy. The estimated matrix has led to greatly improved optimized portfolios as well as beta estimates. -Devised a short term relative value trading strategies for IG corporate bonds using a variant of gaussian process regression.-Modeled equity option implied vol surface using heteroskedastic gaussian processes. The use of heteroskedastic assumption allowed one to quantify the amount of uncertainty of the surface over areas of extreme strike prices, and facilitates relative value trading along the surface. -Constructed a methodology for optimal index tracking portfolios by combining standard covariance methods with statistical clusters. The addition of statistical risk buckets led to increased robustness out of sample.-Developed a quantitative framework to evaluate equity tail-risk hedging strategies for clients.
