Quantitative Researcher, Portfolio Construction & Risk Modeling

National Association of Women in Construction; Greenwich (CT); On-site; USD 150,000 - 230,000; Full time

Job summary

The National Association of Women in Construction in Greenwich, CT, is seeking a Quantitative Researcher to lead research on proprietary covariance and portfolio-risk models used to construct portfolios across its strategies. This investment-research role works directly with the CIO and partners and tests work in live trading.

Researchers own end-to-end research from methodology to implementation and contribute to related portfolio-construction and alpha-monetization challenges, within a small,

Qualifications

• Two to ten years of hands-on quantitative research experience in finance or related fields.
• Practice testing covariance, volatility, or portfolio-risk estimates and their impact on position sizing or portfolio construction.
• PhD in a quantitative field is required.

Responsibilities

• Develop and improve methods for estimating covariance/coherence/volatility across futures and equity universes.
• Establish baselines and compare methods using predicted vs realized portfolio volatility and effect on portfolio behavior.
• Measure how covariance estimates affect position sizes, diversification, concentration, and turnover.
• Own the research code, tests, diagnostics, and model recommendations, from implementation to results.

Skills

Multivariate statistics

Time-series estimation

Numerical linear algebra

Python programming

Education

PhD in statistics / math / physics / EE / econometrics

Tools

Python

MATLAB

C++

Job description

A systematic investment manager that has traded global futures for over a decade, and more recently expanded into U.S. equities, is seeking a Quantitative Researcher to lead research on proprietary covariance and portfolio-risk models used to construct portfolios across its strategies. This is an investment-research role, not risk oversight or vendor-model administration.

Researchers are viewed as core to the investment process rather than a support function, working directly with the CIO and partners and seeing their work tested in live trading. The team is small and lean, so candidates should be comfortable owning research end to end, from methodology through implementation, and contributing to related portfolio-construction and alpha-monetization problems.

Key Responsibilities

• Develop and improve methods for estimating covariance, correlation, and volatility across futures and equity universes, including how histories are selected and weighted, how estimates are shrunk and conditioned, and how models handle missing data, new instruments, and changes in the investment universe.
• Establish baselines and compare methods using predicted versus realized portfolio volatility, estimation error, matrix conditioning, eigenstructure, and resulting portfolio behavior, determining when evidence supports changing the current methodology.
• Measure how covariance estimates affect position sizes, diversification, concentration, and turnover, identifying changes attributable to estimation noise and investigating methods that reduce avoidable turnover.
• Own the research code, empirical tests, diagnostics, and model recommendations, carrying work through implementation and evaluating results against research predictions.

Requirements

• Two to ten years of hands-on quantitative research experience in finance, such as systematic strategy research, derivatives, futures, equities, risk estimation, or portfolio construction.
• Practical experience testing covariance, volatility, correlation, or related portfolio-risk estimates and measuring their effects on position sizing or portfolio construction.
• A PhD in statistics, mathematics, physics, electrical engineering, econometrics, operations research, or a comparably quantitative field required.
• Strong command of multivariate statistics, time-series estimation, and numerical linear algebra, including the assumptions, limitations, and failure modes of high-dimensional covariance methods.
• Strong Python and the ability to write tested, maintainable numerical research code; MATLAB or C++ experience useful but not required.