Oct 2015 – Jul 2020 · 4 yrs 9 mos · Mumbai, Maharashtra, India

• Designed and developed a simulation environment in python, the team’s first backtesting architecture to backtest midfrequency strategies using multiple datasets. Implemented in C++ with more functionality.
• Developed and backtested intraday momentum based strategies on stock futures traded on NSE.
• Deployed them in team’s existing C++ based HFT framework. Improved them to hold overnight positions too and capture longer term momentum, which ran a variable booksize of up to 200 Cr in gross exposure.
• Successfully built and implemented in C++ a fully systematic bottom-up quantitative Long Short market
• neutral strategy in Indian equities using Price Volume/Open Interest data.
• Adapted interday alphas for intraday trading and used them as prediction variables to predict single stock returns and hence used the weights to combine them in single stock strategies for stock futures traded on NSE.
• Also involved in active managing of the portfolio and writing scripts for automating manual processes, getting and cleaning data, alerts for sanity checks/crash reports, parameter optimizations for strats, sensitivity analysis of params and allocation sizes, order execution analysis, slippage analysis and post trade analysis for improvement of strats.

Jan 2013 – Dec 2014 · 1 yr 11 mos

• Developed alpha models for spot G10 and EM currencies (sharpe ratio > 1.5) using various data sources: News Sentiment using (RavenPack News Analytics), FX Dealer Order Flow using (EBS data), Fund Allocation - EPFR data.

Jul 2011 – Dec 2012 · 1 yr 5 mos

• Researched stat-arb signals for dollar neutral equity strategies in US, Europe and Asian markets generating abnormal returns at high Sharpe (2.5+ for turnover< 40% and 1.5+ for turnover< 10%) using various data sources:
• Fundamental using (Compustat, Worldscope),
• Market Microstructure using (limit order book and tick data),
• Analyst consensus earning estimates,
• Customer Supplier momentum using Revere and
• Securities lending market (short interest) using Dataexplorer.
• Built 300+ alpha models, used by PMs for their strategies in equities; spanning over $40M investment at peak weight in portfolio (0.4%)

May 2010 – Jul 2010 · 2 mos · Bangalore, India