A grid/recovery system for XAUUSD built the opposite way round: the brakes first, then the engine. Five independent layers of protection, no martingale, and an absolute loss floor that no combination of inputs can switch off by accident.
High-risk instrument. Read the full risk warning at the bottom of this page before making any decision.
Each layer fires at a defined threshold and acts on its own. The values below are the stock configuration — all adjustable, none removable by accident: the absolute floor is active in every risk profile.
Every cycle is written to an auditable CSV — depth, peak volume, worst floating loss, hedge activations and the reason it closed. What happened to your account is never a mystery.
Stock lot multiplier of 1.00. The grid adds levels with ATR-based widening spacing — it never doubles the bet to "win it back".
The classic grid trick is hiding losses in floating P/L for weeks. Here, the time floor closes any cycle older than 12 hours.
We publish the drawdown next to the return and label backtests as backtests. Anyone promising certain profit in trading is lying to you.
| Platform | MetaTrader 5 |
| Instrument | XAUUSD (gold) |
| Entry signal | 4h momentum in ATR + M5 pullback |
| Filters | H1 EMA · ADX · volatility regime |
| Grid | up to 6 levels · widening ATR step |
| Lot multiplier | 1.00 (no martingale) |
| Basket target | adaptive (ATR + spread) with trailing |
| Cycles per day | max. 8 |
| Loss floor per cycle | 20% (adjustable, never removable) |
| Loss floor per day | 12% |
| Maximum basket age | 12 h |
| TailGuard hedge | arms at 25% DD · 85% of volume |
| Deep deleverage | from 60% stress |
| Weekend close | Friday 21:30 (server) |
| Per-cycle log | auditable CSV |
| Account | hedging · multi-chart via magic numbers |
No. The lot multiplier is 1.00 out of the box: every grid level uses the same base lot. What grows between levels is the spacing (in ATR multiples), not the bet. And unlike classic grids, there is a hard per-cycle loss limit that closes everything — losses never sit hidden in floating P/L waiting for a miracle.
Probably not, and be suspicious of anyone who says otherwise. The published numbers come from 26 years of testing with modelled ticks; live execution has variable spread, slippage and conditions no test fully captures. We publish the backtest because it validates the architecture — including a 34.7% equity drawdown, which is right there next to the return. Always test on a demo account first.
A hedging-type MetaTrader 5 account (TailGuard opens positions against the basket). The EA checks at startup and warns you if the account doesn't support it. Position sizing tracks the balance, so it works from small accounts to large ones — with proportional risk configurable in the inputs.
Yes. Each instance isolates itself by magic number — different symbols work out of the box; on the same symbol you just assign distinct magics to the second instance. Daily floors measure the whole account, so one instance in trouble halts the others: a shared brake, by design.
Cycle state is persisted in terminal global variables: on restart, the EA reconciles what it finds in the account with what it had saved and resumes management. Even so, an actively managed EA belongs on a VPS with serious uptime — that's part of operating responsibly.
We ask every user to run Sentinel on a demo account before any live account — and we ship the production set with the weekend close enabled.
Trading forex, metals and CFDs involves a high level of risk and can result in the loss of all invested capital. Do not invest money you cannot afford to lose. GoldHub Sentinel is an automation tool — it is not investment advice, and goldhubTraders is not a registered financial advisory firm.
All results shown on this page come from historical tests (backtests) performed in the MetaTrader 5 Strategy Tester using modelled ticks. Backtest results have inherent limitations: they do not reflect variable spread, slippage, latency or real execution conditions, and past performance — simulated or real — is not indicative of future results. The maximum equity drawdown observed in the tested period was 34.7%; future drawdowns may be larger.
We strongly recommend prior use on a demonstration account, enabling the weekend close on live accounts, and sizing risk according to your financial situation. If you are unsure whether this product suits your profile, consult an independent financial adviser.