IBA · OCT 2025 – PRESENT
IB Agent
A systematic intraday Nasdaq-futures strategy that combines opening-range structure with dealer-gamma exposure maps. Paper-trading live on IBKR and mirrored to prop accounts.
NQ 1m · Example tradeIBA
LOADING_
1 Sep 2026 (backtest, real NQ). The opening range is marked, price pulls back into the entry zone and it buys with the map bias; it scales out at the first node, and both runners reach their range extensions.
Backtest return, % of account
LOADING_
$100K account, compounding. Amber: after live position limits and measured execution costs. Grey: gross of both. Dashed: QQQ buy-and-hold over the same dates. The strategy lines are flat from December to mid-February because only six sessions of gamma-map history exist in that stretch: missing data, not a quiet patch.
Strategy logic
- Marks the initial balance, the 30-minute opening range
- Reads indice gamma exposure maps to form a directional bias
- Weighs nearby blocking nodes against the pull of nodes further out
- Needs agreement across maps
- Waits for a pullback into a retracement zone, with map confluence
▶ Enters on the pullback · scales out at key GEX nodes · runners to range extensions · flat by the close
- Annualised return
- +50%
- Max drawdown
- 6.1%
- Sharpe
- 2.1
- Profit factor
- 1.90
- Avg trade
- +0.50%
- Win rate
- 49%
* Backtest, 212 trading days (Oct 2025 – Oct 2026) on real 1-second NQ prices, 72 trades. $100K account after live position limits and measured execution costs; compounding, no withdrawals. Gross of both: +81% annualised, PF 2.30.
Overview
A deterministic intraday engine for Nasdaq futures. It turns a published initial-balance method into explicit rules, adds a directional bias read from three gamma exposure maps, and tags every rule by source so my additions can be tested separately from the original.
The same engine code makes every decision in backtest and live, once per 5-second tick, so there is no separate research implementation to drift. Live, it trades an IBKR paper account with a resting stop and mirrors into four prop accounts through an MT5 bridge, with tighter loss limits than the firm’s own.
Highlights
- Out-of-sample: rules chosen on one window and scored on an untouched hold-out. An earlier version held a profit factor of 2.16 on days never used for tuning.
- Data realism: the first study used a proxy price (QQQ scaled to NQ) and showed a 69% win rate. On real 1-second NQ prices the win rate is 51% — lower, but the edge survived, and every result since uses the real tape.
- Robustness: map-noise injection, removing the five biggest winners, and a sensitivity test showing the 10:00 bias flips under a 2bp spot shift on 38% of days — so I plan on roughly 60% of backtest P&L.
- Execution costs measured, not assumed: spread, entry and stop slippage are taken from the prop accounts’ own fill logs and charged on every trade in the amber curve.
- Risk: per-trade risk chosen from a Monte Carlo of the prop firm’s rules, weighing failure rates against the account’s margin limit.
- Payout policy: bootstrapped years across four accounts compared withdrawal rules. Withdrawing at +20% while keeping +10% as a buffer was chosen — modelled payout of about 37% of allocated capital a year, with the chance of losing an account cut from 10% (withdraw everything at +15%) to 5%.
- Execution: reconciling live fills against the engine across prop accounts showed the signals were profitable while the fills were not. The causes were execution bugs, now fixed and checked daily.
- Safety: refuses any non-paper account, only ever reduces a position after entry, counts only its own fills where several bots share an account, and rolls contracts before expiry.
Notes & caveats
- Exact bias maths, zone bounds, stop rules and sizing are deliberately left out — this strategy is trading live.
- Reversal setups were built, lost on the hold-out and are switched off (see RLOG).
- Live paper trading covers only a handful of trades so far — far too few to judge against the backtest.
Stack
PythonpandasPyArrowib_insyncIB GatewayMQL5DatabentopytestlaunchdTask Scheduler