The signal-generation engine. Empirical research across global financial markets — equities and options, index and crypto derivatives, prediction markets, and corporate bonds — plus adjacent work in biology and AI. Each project pairs an economic mechanism with a defensible identification strategy, and backtests run on the same infrastructure that deploys live. Full papers, data, and methodology are available on request.
A comprehensive microstructure study of an on-chain perpetual-futures exchange — funding-rate dynamics, basis, and where price discovery happens.
A new microstructure framework for how market-makers actually earn in binary-settlement prediction markets.
Do binary prediction markets on Bitcoin contribute to its price discovery, or merely track spot?
How the introduction of perpetual futures affects spot-market quality, with a clean causal identification. Accepted at AFA 2026.
An independent replication of a latency-based method for identifying retail fractional-share trades in the consolidated tape.
Do Bitcoin moves predict returns of crypto-linked equities? A multi-horizon study of the asymmetry and how it has changed post-ETF.
Using a large language model to read earnings-call transcripts and generate Buy / Sell / Hold signals, backtested against realized returns.
A gamma-controlled, delta-hedged volatility-selling strategy on small-cap index options.
A systematic option-selling strategy with delta hedging on US equities, sized for consistency over absolute return.
Monte Carlo pricing of snowball autocallable structured products across volatility regimes — surfacing the hidden tail risk of high-coupon structures.
Quantifying the variance risk premium and backtesting systematic option-selling strategies through a volatility shock.
A liquidity-risk framework measuring execution slippage across crypto exchanges and market regimes.
Cross-exchange funding-rate dynamics and short-term predictability across venues.
Tick-level order-book analysis of Bitcoin — order-flow imbalance, spreads, and depth recovery.
Systematic signal extraction from prediction-market order flow and pricing.
Re-estimating post-earnings announcement drift on a modern sample, with a beat-vs-miss asymmetry.
A multi-horizon study of how news sentiment predicts cross-sectional stock returns.
A fully autonomous engine that designed, executed, and wrote up 50 systematic empirical studies end-to-end — hypothesis to data to econometrics to report — spanning microstructure, factor models, options, bonds, and crypto-equity linkages.
A systems-biology study of Type 2 Diabetes — protein-interaction networks, gene-expression meta-analysis, drug-target landscapes, and insulin molecular-dynamics simulation.
A navigable index of CVPR 2024–2025 talk recordings — speakers, timestamped talk-by-talk breakdowns, and cross-referenced models, datasets, and methods.
Full papers, data, and methodology for each project are available on request — get in touch.