AR Alexander Roesler QR-focused MFE candidate

Alexander Roesler

Quantitative researcher building alpha, execution, and ML-driven research systems.

I am a Berkeley MFE candidate with an applied mathematics and statistics foundation, currently building NLP alpha research in an industry project with Trexquant. My strongest fit is quantitative research, with adjacent interest in quant trading, quant development, and data-science roles that value rigorous experimentation.

Cross-sectional alpha research Execution and microstructure modeling ML and NLP signal development Research pipelines and validation
Open Resume GitHub LinkedIn

Best fit: QR. Also interested in QT, QD, and data-science roles with strong modeling and implementation depth.

Now Berkeley MFE candidate working on a Trexquant NLP alpha industry project
Signal Cross-sectional stat-arb, factor mining, intraday execution, and walk-forward validation
Coursework Stochastic processes, stochastic calculus, Monte Carlo, machine learning, algorithms, regression and data mining
Credentials UCLA Applied Mathematics, Statistics & Data Science minor, and Actuarial Exam P passed

I care most about work that is statistically disciplined, implementation-aware, and easy to audit. The strongest environments for me are research teams that value careful validation, robust code, and concise communication.

alex_roesler@berkeley.edu Latest resume PDF Optional: view the creative portfolio version

About

Research-first, with a strong implementation bias.

My work sits at the intersection of quantitative research, software engineering, and model validation. I like turning noisy real-world problems into research pipelines that can survive scrutiny: realistic costs, no-lookahead discipline, reproducible experiments, and reporting that makes the result easy to inspect.

  • Strongest signal in alpha research, systematic trading ideas, and implementation-aware backtesting
  • Comfortable moving between statistics, optimization, machine learning, and production-style research code
  • Bias toward systems that are rigorous, legible, and useful to the next person reading them

Current Focus

What I am working on right now.

Trexquant industry project

I am currently building an NLP alpha signal pipeline that turns SEC EDGAR filing text into tradable, cross-sectional return predictors. The work combines embeddings, statistical and ML modeling, and walk-forward evaluation with decay, turnover, and regime-robustness analysis.

  • End-to-end text-to-signal workflow
  • Cross-sectional prediction and formulaic alpha construction
  • PnL, decay, turnover, and robustness diagnostics

Current MFE track

  • MFE 230A: Investments and Derivatives
  • MFE 230E: Empirical Methods in Finance
  • MFE 230Q: Introduction to Stochastic Calculus
  • MFE 230T: Positioning Yourself for Opportunities in the Financial World
  • MFE 293: Individually supervised Trexquant industry project

Projects

Projects that best signal QR, QT, and data-science readiness.

Independent research 2026

NLP Alpha Signal Research

Building an end-to-end pipeline that transforms SEC EDGAR filings into tradable cross-sectional signals through embeddings, predictive modeling, and formulaic alpha construction.

  • NLP embeddings and text features
  • Walk-forward backtesting and signal decay
  • Regime robustness and turnover profiling

Python, statistical learning, NLP, cross-sectional backtesting

See on resume
Independent research 2025

Cross-Sectional Statistical Arbitrage

Market-neutral crypto stat-arb research on a 174-asset panel with PCA risk neutralization, signed-graph clustering, and strict out-of-sample portfolio construction.

  • 1.76 Sharpe, 29.2% annualized return net of 25 bps/side
  • Near-zero market exposure and ~27.5% daily turnover
  • No-lookahead daily OOS pipeline with liquidity filters

Python, pandas, scikit-learn, statsmodels, graph clustering

View repository
Independent research 2026

Alpha Factor Mining Framework

Built a US equities factor-research framework with point-in-time universe construction, embargo windows, transaction-cost modeling, and gated promotion criteria for factor selection.

  • Deflated Sharpe adjustment for multiple testing
  • Subperiod stability and sector concentration checks
  • Programmatic factor generation via constrained LLM prompts

Python, cross-sectional research, factor validation, tooling

View repository
Independent research 2025

Intraday Optimal Execution

Modeled temporary market impact from minute-level order book data, smoothed intraday liquidity with penalized B-splines, and solved for the cost-minimizing execution schedule.

  • Piecewise impact model with a concave power-law tail
  • GCV-tuned liquidity curve estimation
  • KKT-based execution schedule with Lagrange multiplier bisection

Python, optimization, microstructure, execution research

View repository

Resume

Condensed evidence for quick hiring review.

1.76 Crypto stat-arb Sharpe
29.2% Annualized stat-arb return
Exam P Passed Sep. 2025
2027 Expected Berkeley MFE graduation
  • Current Trexquant industry project: end-to-end NLP alpha pipeline from SEC text to tradable cross-sectional predictors.
  • Independent QR work in stat-arb, factor mining, and execution modeling with implementation-aware validation.
  • Strong academic signal from Berkeley MFE plus UCLA applied mathematics and statistics/data science training.
  • Relevant coursework spanning stochastic processes, stochastic calculus, machine learning, algorithms, Monte Carlo, optimization, and data mining.
Download PDF

Professional and research experience

  • William Blair private wealth management internship building a Sharpe-ranked quant-fund screener and producing alpha/beta-based investment memos.
  • Oxford Algorithmic Trading Programme with vol-scaled momentum, debiased backtests, walk-forward OOS evaluation, and VaR/ES stress testing.
  • Research style centered on no-lookahead discipline, realistic costs, risk neutralization, and robustness under multiple testing.

Skills and Education

Tools, coursework, and training behind the signal.

Python NumPy pandas SciPy statsmodels scikit-learn PyTorch C++ R SQL Bash Docker Git Linux Jupyter Machine Learning NLP Monte Carlo Optimization Stochastic Calculus Pattern Recognition
2026-2027 UC Berkeley Haas

Master of Financial Engineering, with current coursework in derivatives, empirical methods, stochastic calculus, and a Trexquant industry project.

2021-2025 UCLA

B.S. Applied Mathematics, Minor in Statistics & Data Science, GPA 3.82, Dean’s List.

2022 Oxford Algorithmic Trading Programme

Vol-scaled momentum, debiased backtests, walk-forward out-of-sample research, CAPM/FF3 alpha, and risk stress testing.

Contact

Reach out directly.

Email alex_roesler@berkeley.edu GitHub github.com/Aroesler1 LinkedIn linkedin.com/in/alexroesler Resume Open the current PDF resume