Pulkit Gupta

About Pulkit Gupta

Quantitative Analyst in Platform Strats at Morgan Stanley, supporting equity derivatives trading desks across vanilla and exotic products. Experienced in P&L attribution, risk modelling, and building scalable analytics infrastructure for trading and strats teams. Worked closely with equity derivatives and exotic strats to enhance transparency into P&L and risk drivers, and integrate pricing models, Greeks, and volatility-based analytics into platform systems. Strong foundation in stochastic processes, derivatives pricing, and Monte Carlo methods, with hands-on expertise in Python and large-scale quantitative systems.

Experience

1. Senior Associate

   Morgan Stanley

   Jul 2025 — Present · Mumbai, IN

   Quantitative researcher specializing in multi-factor equity risk models and statistical portfolio risk decomposition. Led Devlopment and enhancement of BARRA-style factor models, improving ex-ante risk attribution accuracy by 15% through refined factor definitions, improved cross-sectional exposure estimation, and more robust factor covariance and idiosyncratic risk modeling.Estimated and validated factor exposures via cross-sectional regressions, and modeled factor covariance matrices using time-series techniques with stability diagnostics, shrinkage, and out-of-sample validation. Developed and calibrated specific risk models for securities in Python, incorporating residual analysis, volatility clustering, and regime sensitivity to improve forecasted risk consistency.Conducted systematic factor return analysis and backtesting to detect factor decay, instability, and correlation breakdowns across market regimes, driving targeted model revisions and measurable improvements in risk forecasts. Partnered closely with portfolio managers to decompose P&L into factor and specific components, quantify unintended bets, and optimize active risk, reducing factor concentration and overall active risk by 20%.Designed scalable quantitative infrastructure, including a shared in-memory market data layer for EOD and select intraday data, enabling low-latency factor computation and large-scale daily risk runs with 70% lower peak-time latency. Automated quantitative risk diagnostics, stress testing, and reporting pipelines, accelerating research iteration cycles and reducing analysis turnaround time by 40%.

Education

• University of Westminster

  Master of Science - MS, Fintech and Business Analytics

• Guru Gobind Singh Indraprastha University

  Bachelor of Technology - BTech, Computer Science

  2016 — 2020