Systematic futures trader and builder. I turn signal research and regime models into tools that survive a live chart, and I tell you when the edge does not.
WHO IS
OPERATING
Systematic futures trader and developer out of Calgary. I research signals, model regimes, and build the tooling around them: Pine Script indicators with on-chart expectancy and MFE/MAE, Python research dashboards, and fast websites. Everything is walk-forward tested and slippage aware. Clean code you own, and honest answers on what holds up. If a concept has no edge, I show you the numbers and say so. That honesty is the whole point. (hover to decrypt)
WHAT GETS BUILT
PINE SCRIPT v6
Non-repainting signals, multi-timeframe logic, regime filters, on-chart stats engines (MFE/MAE, expectancy, sample gates, walk-forward), alerts for PineConnector, 3Commas, webhooks.
NINJASCRIPT · C#
Custom NinjaTrader 8 indicators and strategies in NinjaScript (C#). ATM and order automation, regime filters, alerts, and the same on-chart stats discipline.
ACSIL · C++
Sierra Chart custom studies and trading systems in ACSIL (C++). Fast studies, spreadsheet-system logic, and order-flow / DOM ready tooling.
PYTHON · FASTAPI
Cockpit-style research dashboards: backtests, walk-forward, equity curves, drawdown, fills, regime state. Live feeds, SQLite/Postgres, parquet, DuckDB.
FAST WEBSITES
Hand-coded, responsive, deployed. Real brands, not page-builder templates. Like the one you are reading.
RESEARCH & RISK
BACKTEST & WALK-FORWARD
Strategy testing done honestly: walk-forward, Monte Carlo resampling, expectancy, MFE/MAE, slippage and cost modeling. You get the verdict on whether the edge survives, not a sales pitch.
RISK & POSITION SIZING
Kelly and fractional Kelly, volatility targeting, risk of ruin, drawdown control, per-trade sizing off your daily loss limit. Sizing that keeps you in the game.
TAIL RISK & STRESS
VaR and CVaR (expected shortfall), extreme value theory for fat tails, Monte Carlo ruin simulation, scenario and shock testing. Know what the bad day actually looks like.
HEDGING MODELS
Protective puts, collars, cross-asset and beta hedges, ratio and basis hedging, vol hedging. Models and context for cutting exposure, not advice.
PORTFOLIO CONSTRUCTION
Mean-variance, risk parity, volatility targeting, correlation-aware allocation. Build a book that is more than a pile of correlated bets.
ML SIGNAL RESEARCH
Feature engineering, LightGBM and XGBoost, leakage-safe cross-validation, honest out-of-sample reporting. Signals tested hard enough to trust or discard.
DATA & ANALYTICS
Data pipelines (Databento, broker feeds), parquet and DuckDB, cleaning, plus trade journals and performance-attribution dashboards.
LET'S BUILD
Secure transmission. Tell me what you are building. I read every message and reply from hello@kjmartyn.com.