🇯🇵

This page is currently available in Japanese only. Parameter names, download links, and all tools work in English.

⚠️

Eight rules for a martingale EA — design it assuming ruin

💴 ~$700 recommended

The win rate is high. That is exactly why it is dangerous.

A martingale (grid) EA trades a high win rate for the rare loss of the whole account. A 90% win rate does not mean "one loss in ten" — it means the one loss takes back the profits and the deposit together. What follows is not how to win, but how to decide in advance where it dies.

📋Eight things to settle before you start

1. Fund it only with money you can lose

Treat this account as one that will eventually reach zero. Keep living costs and your core capital out of it — the moment they are in, you stop deciding clearly.

2. Cap the lot multiplier at 1.5x

The multiplier is a geometric series, so 2.0x lifts the required margin sharply. Surviving the same number of entries then costs an order of magnitude more, which makes the multiplier the dial that sets how far you can take it.

3. Put a ceiling on the number of entries

Unlimited averaging down is the same thing as never cutting a loss. Always enable a depth limit such as MaxNanpinCount and stop adding once it is reached.

4. Always add a margin-level emergency close

A broker stop-out fires at the broker's convenience and at a price you do not choose. Closing first at a level you set (200%, say) puts the size of the loss back in your hands.

5. Withdraw profits instead of letting them sit

Profit left in the account lets the EA justify bigger lots, so the open gain becomes fuel for the next blow-up. Take it out on a schedule and keep the balance flat.

6. One EA and one pair per account

Put two martingale EAs in one account and the drawdown of one eats the margin of the other, taking down positions that had nothing to do with it. Kept apart, the damage stops at one account.

7. Have a rule for switching it off

Martingale breaks in markets that keep running one way. EAs with a regime filter handle that internally; EAs without one need a written rule for stopping them by hand around major releases and shocks.

8. Write the restart condition before you need it

"Blew up, deposit more, carry on" is the most expensive decision available. Decide on paper — while you are calm — whether you restart, re-tune, or stop.

From opening the account to going live

Decide the amount you can lose. Everything else follows from this.
Open a dedicated account holding only that amount — never mixed with an existing one.
Get a VPS. A reboot or a dropped line at home is fatal mid-sequence.
Run it on demo for one to two weeks and watch it add and close as expected.
Go live with InitialLot at the minimum (0.01). Do not scale up at the start.
Re-check that the margin-level emergency close and the entry ceiling are actually enabled.
Withdraw each time you are in profit and return the balance to its starting size.
Record the deepest open loss once a month. If it is worse than you assumed, cut the lot.

The steps people skip — and regret — are 4 (demo) and 7 (withdraw). Skip 4 and a misconfiguration goes unnoticed; skip 7 and the account quietly grows.

The martingale EAs published on FXEA365

EURJPY Nanpin KAMIKAZE

Flagship

EURJPY / weighted-average averaging down

⚠ Assumes ruin

The plainest form: add to the losing position to pull the average entry closer. A depth limit and an emergency margin close decide how it ends.