Does Averaging Down Beat the Consistency Rule? What a Real EA Backtest and a Rulebook Review Actually Show [Opinion]

※ This article is opinion/commentary. It is not investment advice. Backtests are tested against historical data and do not guarantee future results. Averaging down / martingale approaches can cause losses to balloon rapidly. Firm-specific rules change frequently — always check the official site of each firm. (Last updated: August 30, 2026)

The question: can averaging down get around the consistency rule?

In an earlier test, I found that the consistency rule specifically targets trend-following. The more profit concentrates into a handful of big wins, the worse "largest profit day ÷ sum of profitable days" gets — a setup with take-profit at 10x the stop-loss saw its pass rate collapse to 0.0% at a 20% threshold.

Meanwhile, averaging down was the one style where the numbers didn't move by a single point even as the threshold tightened. Because it closes out small profits every day, its largest-day ratio starts out low to begin with.

So — at firms with strict consistency rules, is averaging down the right answer?

I ran 12 configurations on an actual EA and also reviewed the rulebooks of several firms. The short answer up front: conditionally yes, but "averaging down alone" doesn't cut it.

Part 1: How is averaging down actually treated under the rules?

Surprisingly, it isn't named and banned

First I checked FTMO's official Forbidden Trading Practices. Neither martingale, grid, nor averaging down is named explicitly.

What actually applies is this clause:

"Opening substantially larger position sizes compared to your other simulated trades."

This is where the line gets drawn.

The common belief that "martingale is banned" isn't a named rule at all — it's a consequence of this clause. And averaging down at a constant lot size doesn't trip it under this logic.

Fintokei lifted its martingale ban in July 2025

Digging further, Fintokei had already reversed its stance.

"Martingale Trading & Aggressive Averaging is no longer listed among our prohibited trading practices." (July 28, 2025)

Reason: "existing protections — namely, the Maximum Risk limit on open trades and the Daily Loss Limit on challenges — already effectively mitigate the excessive risk typically associated with this trading practice." (Fintokei official FAQ)

This is a shift in design philosophy: "don't ban it by rule — let the drawdown limits naturally contain it." The reasoning is that with the 3% rule and a daily loss cap in place, martingale will self-destruct on its own, so there's no need to ban it individually.

That said, the same page also states plainly that it is "not recommended" and "still a red flag from a risk standpoint."

Where each firm stands

In our own comparison data, the firms that explicitly allow martingale are Fintokei, FundedHive, Fundora, Lark Funding, E8 Markets, and Funded Trader Markets — six firms in total (updated September 10, 2026). Funded Trader Markets in particular states in its official help center, "Yes, martingale and layering strategies are allowed within a single account," explicitly permitting it within a single account (spreading it across multiple accounts triggers a 5-minute rule). The rest either ban it or discourage it, and Funded7 and Eightcap Challenges name it explicitly as banned.

Grid trading, meanwhile, is handled inconsistently. Fixed-size grids are broadly tolerated by most major firms, but FundedNext explicitly names "Grid Trading" as banned (official help center). An implementation where lot size grows on re-entry is treated as martingale, not grid, and is a violation at nearly every firm.

In short: the rulebook risk isn't determined by "is it averaging down" — it's determined by "does the lot size grow."

Part 2: Is averaging down really immune to the consistency rule?

Test conditions

Evaluated across 249 challenge runs, shifting the start date one day at a time. Drawdown was judged on floating equity (including unrealized loss); the profit target and consistency rule were judged on realized P&L.

The immunity is real, but the pass rate is too low to use

No matter how tight the threshold gets, it doesn't move a single point from 13.2%. The immunity is real.

But the pass rate itself is only 13.2%. The breakdown of failures makes the reason clear:

• Max DD failures: 32 / Daily DD failures: 75 / Deadline expiry: 116

Positions stay stuck in floating loss without closing, realized profit never reaches +8%, and time runs out. Immunity is meaningless if you can't pass in the first place.

Part 3: Can it be improved (sweeping 12 configurations)

If "closing too slowly" is the cause, closing faster should fix it. That's the hypothesis I swept.

① The deciding factor was "adding a stop-loss"

Adding a stop-loss (SL5000 = $50) to averaging down tripled the pass rate from 13.2% to 41.2%, and the full consistency-rule immunity was preserved (ratio 13%). Annual P&L also improved, from +21.9% to +44.6%.

The trade count explains why: 124 → 280 trades. The stop-loss keeps positions turning over, so closings keep happening and realized profit reaches +8%. Deadline expiries stayed roughly flat (116 → 121), but daily-DD failures vanished from 75 to 0. Positions stop getting stuck in floating loss.

In other words: the weakness of averaging down is letting floating losses sit unrealized — not the concept of averaging down itself.

② Closing faster backfires

Counterintuitively, TP500 (cutting take-profit to a third) collapsed to an annual −24.7% and a 0.0% pass rate. The take-profit is too small to cover the losses picked up along the way down. Averaging down is a "win small, lose big" structure, so shrinking the wins even further kills it instantly.

③ The spacing and number of steps barely matter

Gap 1000/2000, max 5/10/20 steps — the results all came out nearly identical (13.2–13.6%). During this period, no pullback deep enough to reach the 10th step ever came. Only Gap4000 (wider spacing) slowed things down (25 days → 37 days) and dropped to 6.2%.

④ Martingale added nothing

Martingale at 1.5x came out at exactly the same 13.2% as the baseline, and annual P&L actually got worse (+21.9% → +20.0%). It never reached deep enough into the step ladder for the multiplier to have an effect. All risk, no reward.

Part 4: The real source of the immunity wasn't "averaging down" itself

This is the most important finding in this article. Raise the lot size, and the immunity disappears.

The immunity wasn't because it's "averaging down" — it was because it "risks small and trades on many closing days."

Raise the lot size and it wraps up in 6 days, concentrating profit into fewer days and pushing the ratio up to 39%. At the 20% threshold, that falls to 0.8%. This is the exact same principle that showed up in the earlier test — "the faster you finish, the harder the consistency rule hits you" — and it has nothing to do with which method you're using.

There's no special magic in averaging down. If you want immunity, the only way — regardless of method — is "risk small and pass slowly."

Part 5: Does it actually beat trend-following?

Lining up the best averaging-down configuration (SL5000) against trend-following:

Even at a 20% consistency threshold, trend-following's 52.9% beats averaging down's 41.2%. Immunity alone wasn't enough to close the gap.

So is there any scenario where averaging down wins?

Yes. When the trend-following setup is either "high RR" or "resolves quickly."

Against a trend-following setup that stretches its take-profit too far, or resolves too quickly, averaging down + a stop-loss does pass more often.

But be careful how you read this comparison. RR10 coming in at 0.0% isn't because "trend-following is weak" — it's because that particular setup is a terrible fit for the consistency rule. The same trend-following approach passes 52.9% of the time just by switching to a 50-bar lookback. Adjusting the trend-following setup is a better move than switching to averaging down.

Conclusion

The question: does averaging down work as a countermeasure to the consistency rule?

The answer: a qualified yes. But "averaging down alone" doesn't help — it needs "averaging down + a stop-loss." And in most cases, fixing the trend-following setup is the better move anyway.

To summarize:

1. The immunity to the consistency rule is real (the number doesn't move even at a 20% threshold)
2. But the real source of the immunity is "risking small and having many closing days," not something unique to averaging down. Raise the lot size and it gets hit just like anything else (0.8% at lot 1.0)
3. Averaging down alone isn't practical, with a 13.2% pass rate. Floating losses sit unrealized, realized profit never reaches +8%, and 116 runs expire on the deadline
4. Adding a stop-loss improves this to 41.2% while preserving the immunity. The key driver is daily-DD failures dropping from 75 to 0
5. Even so, at a 20% consistency threshold it still loses to trend-following's 52.9%
6. Martingale (doubling the lot) doesn't improve the pass rate by even a single point, and under the rules it runs straight into the size-consistency clause

The idea of "fleeing to averaging down because the consistency rule is strict" is a poor trade-off in cost-benefit terms — that's the conclusion the measurements point to. Putting the same effort into "don't get greedy with your take-profit distance" or "cut lot size and take more days" raises the pass rate more.

Limitations of this test

• Grid trading could not be tested. It requires specifying a price range, and a fixed range couldn't cover a full year of price movement, resulting in zero trades
• This is the result for one instrument (XAUUSD), one year, one EA. The averaging-down results in particular depend heavily on "no pullback beyond what the setup expected" ever occurring. With a 10-step ladder at $20 spacing, a pullback beyond $200 would have changed the results entirely
• The lack of difference across averaging-down spacing/step counts is for the same reason — a reflection of this period being calm
• Floating equity was reconstructed from execution history and 1-minute bars, and the annual peak likely underestimates by a median of 2.0 points (daily behavior was confirmed to match the EA's own daily-withdrawal settings)

FAQ

Q. Doesn't averaging down trip the consistency rule?

As long as you're risking small, it doesn't. In the test, the largest single-day profit at the point of passing was 18% of total profit, and even at a 20% threshold the pass rate didn't move a single point from 13.2%. However, raising the lot size makes it resolve in around 6 days, worsening the ratio to 39% and dropping the pass rate to 0.8% at a 20% threshold. The immunity only holds while the lot size stays small.

Q. Is averaging down banned by the rules?

Not many firms explicitly ban averaging down (DCA) at a constant lot size. FTMO's official Forbidden Trading Practices doesn't name martingale, grid, or averaging down explicitly either. What actually applies is the clause about "opening substantially larger position sizes compared to your other trades" — martingale, which doubles the lot size, hits this directly, but same-lot averaging down doesn't.

Q. Can I use martingale?

In the test, the pass rate came out exactly the same as the baseline, at 13.2%, and annual P&L actually got worse (+21.9% → +20.0%). More risk for nothing in return. On the rules side, it also runs into the size-consistency clause — of the 17 firms we track, only Fintokei and FundedHive explicitly allow it.

Q. Why did Fintokei lift its martingale ban?

In an official FAQ dated July 28, 2025, it explains that "the maximum risk limit on open trades and the daily loss limit on challenges already effectively mitigate the excessive risk typically associated with this trading practice." The judgment is that drawdown limits naturally contain it without needing an individual ban. That said, it also states plainly that it's "not recommended" and "still a red flag from a risk standpoint."

Q. How can I improve averaging down so it actually passes?

In the test, adding a stop-loss was the most effective fix. Adding SL5000 ($50) raised the trade count from 124 to 280, kept positions turning over, and lifted the pass rate from 13.2% to 41.2%. Daily-DD failures dropped from 75 to 0, and annual P&L improved from +21.9% to +44.6%. Conversely, shrinking take-profit (TP500) collapsed the results to an annual −24.7% and a 0.0% pass rate.

Q. In the end, should I pick trend-following or averaging down?

Even at a 20% consistency threshold, trend-following (52.9%) beat averaging down + stop-loss (41.2%). Averaging down only wins when the trend-following setup is high-RR (take-profit at 10x the stop-loss, for example) or resolves quickly. Even then, adjusting the trend-following approach's take-profit distance and lot size raises the pass rate more than switching to averaging down.

Sources

• FTMO – Forbidden Trading Practices (the position-size consistency clause)
• Fintokei – Martingale and Aggressive Averaging: now allowed but not recommended (the policy change of July 28, 2025)
• FundedNext – Restricted/Prohibited Trading Strategies (the explicit ban on Grid Trading)

※ Each firm's rules change frequently. Always check the latest version on the official site before buying.

Related articles

• What Percentage of Challengers Pass Phase 1 on a Coin Flip? An Exact Calculation Puts It Above 50%

• The Consistency Rule Specifically Targets Trend-Following: 10,000 Runs on a Real EA

• Complete Guide to the Consistency Rule: How It's Calculated, Which Firms Use It, and How to Work Around It

• Complete Guide to Drawdown: The Differences Between Static, Trailing, and Daily DD

• Fintokei Complete Guide