Skip to content

Case Studies

Real strategy backtests with strict no-lookahead validation.

Each case study replays a known option strategy through TradLyt's sizing model on historical market data, applies the ML position-sizing overlay, and reports P&L, Sharpe, and drawdown side-by-side with the flat (un-sized) baseline. Every leg is scored using only data that was public at its entry moment — no same-day lookahead, no feature peeking — and the verdict is a rolling-quarter walk-forward (sizing must beat flat across a majority of independent quarters).

We publish what we find — including the strategies where sizing doesn't help. If the result doesn't survive strict validation, we don't publish it.

Published

  • ML Sizing Overlay — validated results — June 2026 NIFTY + SENSEX, naked and hedged, intraday and positional, with De-risk sizing. The overlay lifts out-of-sample Sharpe materially (e.g. SENSEX straddle 1.26 → 1.84, NIFTY iron-fly 1.70 → 2.38) while cutting drawdown — at equal average exposure.

Coming soon

  • Net-of-costs case study — the same uplift after brokerage / STT / slippage
  • Live forward test — sized vs flat on out-of-sample live sessions

Have a strategy you want sized? Upload your AlgoTest CSV on the /backtest page in the app for an instant flat-vs-sized comparison, or email [email protected].