nelsonmu.com:~
溫故知新

Nelson Mu

I am an analyst at a multiasset fundamental long/short hedge fund, with a special focus on building AI-native workflows that utilize agents effectively, and improving risk & stress models.

[email protected]  /  CV

nelson@work
shell tape arxiv demos

What I do

Model risk & stress testing

Validation, scenario design, and stress tests you can rerun next quarter and get the same answer.

Stress tests

Agentic research workflows

Research, data and reporting structured into versioned, testable workflows that agents can run.

Workflows

Financials

Banks and financials: deposits, credit, capital, measured point-in-time.

Financials work

read the tape

SPY ···XLK ···XLF ···KRE ···XLE ···

nelson@work:~$ ls work/

Work

  1. 01Bank rate-shock and capital tests2026 Securities marks to AOCI to CET1 under key-rate shocks.
  2. 02Validation gates2026 Nulls, multiple-testing bars, known-answer checks, pre-registration.
  3. 03Agentic research workflows2026 Structured prompts, verification checks and reproducible scripts for fundamental and quantitative research.
  4. 04Point-in-time data panels2026 Every fact dated by when it became knowable. A security master that survives renames and delistings.
  5. 05Unattended daily pipelines2026 Agent-run jobs, safe to rerun, watched by an independent monitor.
  6. 06Financials data platform2026 Call reports, XBRL and prices joined point-in-time; banks, insurance, REITs, semis.

All work

now

On the bench

Agent-run daily pipelines, point-in-time bank fundamentals, rate-shock tests.

silent failure

Netflix is not an ETF

A name filter that matched the substring “etf” also matched Netflix.

nelson@work:~$ ls tools/ demos/

Interactive tools & demos

Live quantitative regime & credit monitors, bank peer & county maps, real-time macro nowcasts, and interactive stress-testing demos.

nelson@work:~$ cat arxiv/

Recent papers

arXiv: q-fin, econ

  • Fetching…

nelson@work:~$ open thesis.pdf

Thesis

MSc Economics, LSE, 2026.

The Sectoral Effects of Volatility Shocks

A residual synthetic control analysis of equity market reallocation.

When implied volatility jumps, does the whole market fall, or does money move between sectors? Weekly sector-minus-SPY outcomes around large VIX shocks, synthetic-control counterfactuals per sector, and a residual synthetic control that screens covariates and conditions on market state.

Read the PDF

慕
residual synthetic control VIX shocks sector reallocation permutation importance AugSynth benchmark event study

nelson@work:~$ whoami

About

I work between fundamental research and quantitative modeling: bank balance sheets, stress tests and the statistics that decide whether a result is real, plus the data panels, pipelines and agentic workflows that make the work rerunnable.

MSc Economics at LSE; economics and statistics at UC Santa Barbara. Fluent in English, Mandarin and Spanish.

[email protected] · CV