AlphaLab MartingaleHedge EA
Disclosed-martingale scalper for index futures, wrapped in five layers of kill-switches.
A bidirectional martingale-with-hedge that takes many tiny session targets ($2–5 per session by default) and gates the strategy with a session SL, daily-loss lock, ATR regime guard, max session duration and time-of-day filters. Engineered for traders who want martingale exposure without naked martingale risk.
Strategy overview
A bidirectional martingale that pyramids both BUY and SELL at a configurable price step (default 5 points, optional ATR-adaptive plancher). The first trade in a session can come from an external trigger (LLM signal, manual entry, or auto-placed Buy/Sell-Limit pair at the M1 range extremes). Each subsequent trade on the same side is sized lot × 1.20, capped per trade and per session. A delta cap (default 3) keeps |BUY − SELL| bounded — net exposure stays small and the hedge produces a contrarian leg that softens runaway moves. The session closes when total floating PnL ≥ TpProfit (default $2.50) or PnL ≤ session SL (default −$50). Five additional kill-switches sit on top: a daily-loss lockout (force-close + lock until next CME 17:00 CT reset), an optional daily-take-profit lockout, a max session duration timer (default 1 h), a Forward-ATR Guard that detects post-entry volatility regime shifts and force-closes the session, and a Triple-Close-Same-Candle rule that closes three same-side trades opened in the same M1 candle. Time-of-day filters (session hours, weekday filter, hour blacklist, last-trading-time cutoff, break-at-time) gate placement of new initial trades. Optional Rejection-Close detects six classical reversal candlestick patterns (Hammer, Bullish/Bearish Engulfing, Piercing Line, Shooting Star, Bearish Belt Hold) and closes the rejected side on confirmation.
AlphaLab MartingaleHedge EA is a transparent martingale system. Most martingale EAs hide their assumptions, ship without a session stop-loss, and pretend the lot multiplier is small. This one documents every input on the product page, runs at a modest 1.20× lot multiplier (vs the typical 2× death-spiral), and refuses to compound losses past the configured kill-switches. It is built for short-window M1 scalping on index futures (NAS100, US500) and gold (XAUUSD), where micro-targets in the $2–5 range are reachable many times per session. We classify it R4 because the kill-switches transform what would otherwise be an R5 strategy into a tradeable one — the underlying martingale logic is real, and we say so.
Who it is for
- Traders with $2,000–$25,000 capital who want disclosed-martingale exposure on a single intraday symbol.
- M1 scalpers comfortable with high trade frequency and small per-session targets.
- Engineers and quants who would rather audit a documented martingale than trust an opaque "neural" EA.
- Index-futures and gold traders running CME-aligned sessions (Chicago time, 17:00 CT reset).
Who it is not for
- Traders unwilling to accept that any martingale carries tail risk — even with caps.
- Prop-firm challenge or funded-account holders. The strategy fundamentals (averaging into losers, no per-trade SL) clash with most firm rule sets, regardless of session-level caps.
- Long-horizon investors looking for hands-off, low-frequency automation — see Stability EA instead.
- Anyone who would deploy this without enabling EnableSL, EnableDailyLoss, and EnableForwardATRGuard. The risk rating assumes those switches are on.
Risk presets
Default: standard
| Preset | Step Pts | Lot Multiplier | Max Delta | Tp Profit Usd | Enable S L | Sl Loss Usd | Enable Daily Loss | Daily Loss Amount Usd | Enable Max Session Duration | Max Session Duration Sec | Enable Forward A T R Guard | Enable Atr Step | Max Daily Loss (%) | Max Overall Drawdown (%) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Cautious Tight session caps, gentle multiplier, ATR step ON. Best for first month live or sub-$5k accounts. | 7 | 1.15 | 2 | 2 | Yes | -25 | Yes | -50 | Yes | 1800 | Yes | Yes | 1 | 8 |
Standard Documented defaults from the EA. Recommended baseline once you have run cautious for two weeks. | 5 | 1.20 | 3 | 2.50 | Yes | -50 | Yes | -50 | Yes | 3600 | Yes | No | 2 | 12 |
Aggressive (informed users only) Wider session caps, larger TP, longer sessions. Only for users who have run Standard profitably for ≥ 30 trading days and understand the tail risk. | 4 | 1.25 | 3 | 5 | Yes | -100 | Yes | -150 | Yes | 7200 | Yes | Yes | 3 | 18 |
Illustrative equity curve
Synthetic walk for layout — not a real track record. See the backtest card below for disclosed methodology.
Backtest summary
Historical · not a forecastSample period
2024-10-01 – 2026-04-01
Trades
4,812
Win rate
87%
Profit factor
1.35
Max drawdown
14.2%
Initial capital
$5,000
Symbols
NAS100
Timeframes
M1
- Spread assumption
- Variable spreads averaging 0.6 points on NAS100 with widening to 1.4 points around the 09:30 ET equity-cash open and 16:00 ET CME settlement.
- Slippage assumption
- 1 point on entries (limit fills assumed at quoted price), 2 points on session-close stop-outs (worst-case at break-at-time), 3 points on Forward-ATR Guard fires (regime-shift exits).
- Source assumption
- IC Markets cTrader Raw Spread tick data via the MT5 strategy tester. Not affiliated. Replication on other brokers will differ — particularly market-maker brokers and dealing-desk routes.
Setup requirements
Broker requirements
MT5 only. ECN/STP broker with a hedging account (MT5 hedging mode is required — the EA simultaneously holds BUY and SELL on the same symbol). Symbol naming: NAS100 / US100 / US500 / GER40 / XAUUSD (suffix-tolerant). Spread of ≤ 1.5 points on the chosen index is recommended; tighter is better. Server location should match the broker primary data centre — typically London or New York for European/US brokers.
VPS recommendation
Required. The EA reacts on every M1 tick and pyramids on price-step crossings — running on a laptop will miss fills and break the grid. 2 GB RAM, 2 vCPU, < 50 ms latency to the broker server. Chicago, New York or London regions for CME-aligned brokers.
Known limitations
- • Like every martingale, the worst-case loss is bounded only by the session SL and the daily-loss lockout — both of which depend on the broker honouring stop-out levels under live spread/slippage conditions.
- • Performance is highly sensitive to spread widening. The system places frequent micro-TP orders; brokers with variable spreads at session opens or news events will see degraded fills.
- • Backtests on M1 tick data understate slippage on grid-fill events. Live results will differ from backtest equity curves; the gap widens with broker quality.
- • The Forward-ATR Guard catches volatility regime shifts but cannot prevent gap losses across server downtime or weekend gaps.
- • Lot escalation (1.20×) is gentle by martingale standards but still produces position sizes that are non-trivial after 8–10 pyramiding steps. The MaxLot cap is your hard stop on this dimension.
FAQ
The brand is anti-*hidden* martingale. We refuse to ship products with uncapped recovery logic, undocumented grid behaviour, or marketing that hides the strategy. This EA is the opposite: it documents every kill-switch line by line, exposes every input parameter, and refuses to run profitably without the session SL and daily-loss lockout enabled. We list it because a documented martingale-with-hedge is a legitimate strategy when used with risk caps — the dishonesty is in pretending it isn't.
Changelog
v1.0.0 · 2026-04-30
Initial release. Bidirectional martingale with hedge + delta cap, 1.20× multiplier, session TP/SL, daily-loss/TP lockouts, max session duration, Forward-ATR Guard, Triple-Close, Rejection-Close (six patterns), regime filters (rvol/volume/session/weekday/hour blacklist), CME-aligned daily reset (17:00 CT, DST-aware).