Live demo · 7.6 MB payload · runs in your browser
A real iron condor backtest.
No server. No signup.
788,433 rows of actual NIFTY options data for 2024–2025 load directly into your browser. A DuckDB engine runs the SQL. The iron condor strategy backtests against real settlement prices. Nothing leaves your device.
Adjust the parameters, then click Run backtest to load the data and simulate weekly iron condors on Nifty, 2024–2025.
A word on these numbers
What this backtest does — and doesn't — tell you.
Entry fills use the Close price on entry day. Exit fills use the SettlePrice at expiry. Both are theoretical reference prices. In real trading you'd see slippage (0.5–2% per leg), commissions, and bid-ask spread you can't cross on thin strikes.
That means these results are systematically optimistic. A real iron condor with honest execution will have ~half the win rate shown here, and Sharpe closer to 1.5–2.5 instead of 5+.
This backtest is here because it runs on real data with real strikes and real expiries. It's a proof of concept for the pipeline, not a signal to trade. When Thuztra launches, the production backtest engine adds slippage, liquidity filters, and bid-ask modelling.
How it actually runs
- Your browser downloads a 7.6 MB parquet file (788k options rows)
- DuckDB-WASM spins up (~5 MB runtime, cached after first visit)
- A ~15-line SQL query picks weekly entries, matches strikes, joins exit prices
- Results render as an equity curve — all on your CPU, zero server cost