Research collection
Selected Topics in Quant Finance
Six standalone studies in quantitative finance: central-bank statement sentiment, intraday spectral structure, order-book microstructure, risk-signal change detection, cross-sectional intraday reversal, and a public-data replication audit.
These are independent investigations grouped by theme rather than a single arc, in no particular order. Each opens its full report below.
FOMC Statement Sentiment and the 2-Year Yield Reaction
An event-study testing whether FinBERT-coded changes in FOMC statement tone explain 30-minute 2-year Treasury-futures proxy moves around release times.
- Role
- Independent researcher and quantitative analyst
- Methods
- FinBERT, FOMC statements, event-study windows, HAC-robust OLS, cycle stratification, residual analysis, Python
- Outcome
- Found no measurable relationship in the statement-read window, while the press-conference window flips to the expected sign but remains statistically indistinguishable from zero.
- Alternative Data
- Event Study
- Fixed Income
Spectral Signatures in SPY and BTC Intraday Returns
A sample-limited comparison of intraday spectral band shares in SPY and BTC, with a cross-method audit showing how STFT and DWT specifications can dominate the headline asset gap.
- Role
- Independent researcher and signal-processing analyst
- Methods
- Python, STFT, discrete wavelet transforms, date-block bootstrap, Kolmogorov-Smirnov tests
- Outcome
- Found that the SPY-BTC high-frequency gap is small and specification-sensitive, while the stronger result is a methodological warning about residual versus low-pass band definitions.
- Signal Processing
- Market Microstructure
- Python
Order-Book Imbalance as a One-Second Predictor
A reproducible execution-research baseline testing whether top-five order-book imbalance, spread, and prior-one-second signed trade imbalance predict one-second-ahead mid-price direction.
- Role
- Independent researcher and execution-research analyst
- Methods
- Python, LOBSTER order-book data, chronological validation, logistic regression, calibration, ROC AUC, transaction-cost overlay
- Outcome
- Found modest directional discrimination on a single AAPL day, but the signal does not survive a taker-style spread-plus-fee cost model.
- Market Microstructure
- Execution Research
- Python
SPX Realized-Volatility Change-Point Shoot-Out
A pre-registered detector bake-off comparing HMM, CUSUM, PELT, and a fixed-threshold baseline on SPX log realized volatility from 2010 to 2026.
- Role
- Independent researcher and risk-signal analyst
- Methods
- Python, realized volatility, HMM, CUSUM, PELT, event matching, false-alarm analysis
- Outcome
- Established under one shared evaluation rule that a fixed-threshold baseline dominates the named detectors on both false alarms and F1, and delivered the reusable bake-off harness that settles the comparison.
- Risk Signals
- Change-Point Detection
- Python
Cross-Sectional Intraday Reversal at Multiple Horizons
A point-in-time re-measurement of intraday cross-sectional reversal on 373 million minute bars across 529 US large caps, sorted into deciles at 5-, 30-, and 60-minute horizons over 2018 to 2025.
- Role
- Independent researcher and quantitative analyst
- Methods
- Python, point-in-time universe construction, decile sorts, 5-day block bootstrap, decay characterization, turnover-aware transaction-cost overlay
- Outcome
- Found the effect statistically real but economically dead: it breaks even below half a basis point of round-trip cost and decays inside a single sort interval, the signature of bid-ask bounce rather than deployable alpha.
- Factor Research
- Market Microstructure
- Python
Real-Time Machine Learning in the Cross-Section: A Public-Data Reproduction
A leak-audited replication attempt on the Chen-Zimmermann sub-analysis of Li, Rossi, Yan and Zheng (JFE 2025), rebuilt as an expanding-recursive real-time pipeline on public, commercial-clean data.
- Role
- Independent researcher and model-governance analyst
- Methods
- Python, boosted regression trees, expanding-recursive validation, decile portfolios, six-factor alphas, SEC EDGAR XBRL fundamentals, ablation and cost overlays
- Outcome
- Locates the exact data seam that blocks the paper-faithful reproduction rather than papering over it, and reports a commercial-clean track whose equal-weighted Sharpe of 1.25 survives risk adjustment but does not hold under value-weighting.
- Model Governance
- Replication
- Python
