matthiasheymann

Aided in the firm’s transition to its new interest rate risk modeling infrastructure. Documented the firm’s Barra-based equity factor risk model. Devised & implemented novel risk management methods related to the equity factor risk model. Recreated and cross-validated the monthly investor risk report based on new data sources.

Directing the group’s transition to AWS cloud computing (setup & teaching sessions). Aiding the group in their Covid-19 and US election research (based on Twitter & Google data).

Codeveloped the mathematical framework of a revolutionary new interest rate model that resolves key shortcomings of other leading approaches. Published as a book in 2018; 2nd ed. (extension to FX) in 2020, 3rd ed. (faster derivation) in 2024. Identified model limitations and uncertainties for a variety of risk models and quantified their materialities, frequently in the context of CCAR/DFAST stress tests. Designed statistical analyses and alternative approaches to test and challenge the firm’s models. Responsible for top-level VaR-related models (and others). Participated in Fed meetings, helped the firm pass all of its annual CCAR/DFAST stress tests. Onboarded and trained 130+ new team members & interns; interviewed 50+ candidates. Mentored ~10 newcomers, oversaw their first projects, and gave them frequent one-on-one lessons. Became the go-to person for questions about mathematical modeling, LaTeX, and IT.

Helped design and program the new ALM (Asset & Liability Management) infrastructure. Goals: Given portfolio data, send weekly liquidity assessment reports to PMs and Corporate Treasury. Prevent bankruptcy in future liquidity crises (e.g., Lehman Brothers in 2008). Wrote 15,000 lines of object-oriented code in a production environment, participated in code reviews. Established and led a biweekly learning series with 20+ talks by invited speakers.

Wrote a research monograph on a problem in probability theory (large deviation theory), published as a book in the Springer series 'Lecture Notes in Mathematics' (186 pages). Taught graduate- and undergraduate-level math classes in a variety of subjects.