🔍 Firm comparisons

Prop Firms That Allow Averaging Down (Nanpin): What the Rules Actually Ban Isn't 'Buying the Dip' — It's Increasing Lot Size

Published: 9/16/2026Updated: 9/27/2026

※ This article reflects each firm's public terms and help center as of September 16, 2026. Not investment advice. Averaging down and martingale are methods where losses can expand rapidly. Firm-specific rules are updated frequently, so always check each firm's official site before buying.

You often see the claim that "prop firms ban averaging down (nanpin)," but when you actually read the terms, that's usually not what's written.

What's actually banned is martingale — that is, "increasing lot size after a loss." Averaging down at the same lot size, or at a reduced lot size, is not a problem under the terms at many firms.

The firm that draws this line most clearly is Funded7.

Conclusion: the answer up front

FirmSame-lot/reduced-lot averaging downMartingale (increasing lot size)The constraint that actually bites
FintokeiAllowed (not on the banned list)Allowed (explicitly lifted in July 2025)Floating loss on the combined total of all open positions capped at 3% of balance / no commercial EAs
City Traders Imperium 1-StepAllowedExplicitly allowedBalance-based 5% trailing drawdown / third-party EAs also allowed
Funded7Allowed (explicitly stated)Named and banned under Rule 4Tiered risk caps of 1–3%
FTMONot namedNot namedThe "substantially larger position" clause / 200 simultaneous positions
E8 MarketsNot confirmed as namedSame as leftA discretionary clause that can cap risk at 1% per trade idea
Funding PipsNot namedNot namedMost-cited account closures in reviews are for Risk Per Trade Idea violations
Hantec TraderNot namedNot namedOne-sided bets banned / on a funded account, combined floating loss reaching 3% of the initial balance is an instant fail
Moneta Funded—BannedGrid also banned / EA must be self-built only
Eightcap Challenges—BannedGrid and hedging also banned / 10 simultaneous positions max
Alpha Capital——Fully automated EAs are banned outright (a non-issue before the method even matters)

If you're running an EA, Fintokei or City Traders Imperium; if you're trading discretionarily, Funded7 is also a candidate — that's roughly how it lines up.

How the terms separate averaging down from martingale

Funded7's wording is the clearest

Funded7's Rule 4 explicitly bans martingale by name. What's banned is "increasing lot size geometrically after a loss," and the official terms even give a concrete example.

  • BANNED: increasing from 1.0 lot → 1.5 lots → 2.0 lots
  • SAFE: same lot size, or averaging down at a reduced lot size (as long as it stays within Rule 2's risk cap)

In other words, what Funded7 bans is the lot-size curve, not averaging down itself. Summarizing this as "averaging down banned" throws away a method that's actually usable.

Fintokei lifted the martingale ban entirely in July 2025

Fintokei switched to explicitly allowing martingale as well. The original text of the official help article reads:

"As of July 28, 2025, Martingale Trading & Aggressive Averaging is no longer listed among our prohibited trading practices." "You will not be penalized for using Martingale or Aggressive Averaging."

The same page explicitly states it doesn't recommend it, and cites the Maximum Risk limit on open trades and Daily Loss Limit as the alternative guardrails. The policy shift is "we don't restrict the method — we restrict the amount of risk."

Checking the prohibited trading practices page on September 16, 2026, there are 9 banned items, and neither averaging down, martingale, nor grid is among them. What is banned: latency arbitrage, tick scalping, copy trading, third-party EAs, multiple profiles, shared devices, IP spoofing, third-party payment processing, and hedging across multiple accounts.

FTMO doesn't name it, but a different clause catches it

FTMO's list of forbidden trading practices doesn't name martingale or averaging down either. Instead, this is the clause that ends up doing the work:

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

Martingale's final step ends up several times the size of the first, so it runs straight into this clause. Same-lot averaging down, on the other hand, never creates a "substantially larger" position, so it's less likely to trip it. Here too, the lot-size curve is the deciding factor.

What actually determines how many steps you can take is "total lot size"

What actually blows up an account when averaging down isn't a banned clause in the terms — it's the floating-loss cap. And what drives the math isn't the average entry price — it's the total lot size.

Let's look at a $100K account with a 3% floating-loss cap (= $3,000), averaging into XAUUSD at $10/oz intervals over 5 steps.

Same lot size (0.2 lots × 5 steps = 1.0 lot total)

By the time you've entered the final step, the average entry price is $20 below the first entry, and the current price is $40 below the first entry.

Code
Floating loss = 1.0 lot × 100oz × $20 = $2,000
Remaining room = $1,000 ÷ 100oz = $10/oz

Martingale (0.1 → 0.2 → 0.4 → 0.8 → 1.6 = 3.1 lots total)

At the same intervals and the same number of steps, the average entry price gets pulled down further, so the floating loss amount itself ends up in a similar range. But the total position size is 3.1x larger, so the weight of any further price move is completely different.

Code
Floating loss = 3.1 lots × 100oz × $8.4 = $2,604
Remaining room = $396 ÷ 310oz = $1.28/oz

Even at the same 5 steps, the room left right after the final step is $10/oz vs. $1.28/oz — a 7.8x difference. The reason martingale tends to get banned isn't a moral judgment — it's this number. The moment you place the final step, the account has almost exactly one price move's worth of runway left.

When designing this, don't ask "how many steps can I fit in" — ask "how many dollars of room is left right after the final step."

Firm-by-firm notes

Fintokei: the 3% cap applies to the combined total of every position open at the same time, regardless of symbol or direction. Running averaging-down positions on multiple pairs at once eats into the same shared limit, so it's safer to stick to one system per account. It also doesn't allow commercial EAs (item #4 on the banned list), so a purchased averaging-down EA can't be loaded even though the method itself is allowed. This 3% warning also accumulates across accounts — 3 warnings caps you at 1% risk, and 6 warnings restricts every account you have.

City Traders Imperium 1-Step: drawdown is Balance Based (highest balance minus 5% of the starting balance), so floating loss doesn't move the floor. This is the only setup here that squarely matches a design that holds a floating loss and waits for a bounce, the way averaging down does. Third-party EAs are allowed with no proof of ownership required. Note this condition is 1-Step only — the 2-Step, Instant, and Direct plans are a different setup entirely ("self-built EA only, martingale not allowed").

Funded7: risk caps are tiered (Gold 3% / Silver 2% / Bronze 1%), and if you don't set a stop-loss, assumed risk is calculated as ATR(14) × 1.96. Running an averaging-down EA without a stop-loss makes it easier to hit the cap under this conversion.

Hantec Trader: has a one-sided bets ban clause, and continually adding to positions in one direction, as averaging down does, may prompt a request for explanation. On a funded account, combined floating loss across all positions reaching 3% of the initial balance is an instant fail (no such rule during evaluation). Habitually risking more than 3% on a single trade can also get you discretionarily placed into the Risk Management Group, lowering your leverage.

Funding Pips: account closures for Risk Per Trade Idea violations are the most-cited complaint in reviews. With no explicit ban named, the real question is how you weigh the risk of a discretionary closure.

What about the middle ground (reduced-lot averaging down)?

A design that averages down at a reduced lot size is the safest category under the terms. Funded7 explicitly labels it SAFE, and it doesn't trip FTMO's "substantially larger position" clause either. Since total lot-size growth is contained, the room described in the calculation above stays intact.

On the expected-value side too, our own testing found that adding a stop-loss to an averaging-down setup raised the consistency-rule pass rate from 13.2% to 41.2% (see our validation article). Not "averaging down alone," but "averaging down + stop-loss + fixed lot size" is the design that works with both the terms and the expected value.

FAQ

Q. Aren't prop firms banning averaging down?

Most firms don't explicitly ban it. What's banned is martingale — increasing lot size after a loss. Funded7 spells it out with concrete examples: "increasing from 1.0 → 1.5 → 2.0 lots is BANNED; averaging down at the same or a reduced lot size is SAFE." Summarizing this as "averaging down banned" throws away a method that's actually usable.

Q. Are there firms that officially allow martingale?

Fintokei, and City Traders Imperium (1-Step). Fintokei removed it from its banned list on July 28, 2025, explicitly stating "you will not be penalized for using it." CTI's 1-Step rules table states "Martingale: Allowed." Neither recommends it — both rely on risk caps as the guardrail instead.

Q. Can I buy a commercial averaging-down EA and use it?

Yes, on City Traders Imperium's 1-Step (third-party EAs are allowed with no proof of ownership required). Fintokei bans commercial EAs, so even though averaging down itself is allowed, you can't load a purchased EA — it has to be self-built, or one where you've adjusted the parameters yourself.

Q. How many steps am I allowed to take?

Decide based on "how much room is left right after the final step," not the number of steps. On a $100K account with a 3% floating-loss cap, averaging in at $10/oz intervals over 5 steps at the same lot size (1.0 lot total) leaves $10/oz of room right after the final step. At the same 5 steps but with martingale, total lot size becomes 3.1 lots, and room drops to just $1.28/oz. Same number of steps, but a 7.8x difference in runway.

Q. Is Fintokei's 3% per symbol, or across the whole account?

Across the whole account. It applies to the combined floating loss of every open position, regardless of symbol or direction. Running averaging-down positions on multiple pairs at once eats into the same shared limit, so it's safer to stick to one system per account. This warning also accumulates across accounts — 3 warnings caps you at 1% risk, 6 restricts every account, and 10 ends your service with the firm.

Q. Doesn't adding a stop-loss defeat the purpose of averaging down?

On the expected-value side, it actually worked the other way. In our 12-configuration test, averaging down alone passed the consistency rule only 13.2% of the time — not usable on its own. Adding a stop-loss raised that to 41.2% while preserving its immunity to the consistency rule. It also makes the risk cap under the terms clearer, so a design with a stop-loss in place is a better fit for a prop account.

Q. Are grid trading and "torarepi" (repeated range trading) treated the same way?

Often they fall under a separate clause. Moneta Funded and Eightcap Challenges both name and ban grid trading explicitly. Fintokei's list of 9 banned items, on the other hand, doesn't include grid. For more detail, see Prop Firms That Allow Grid Trading and Torarepi.

Sources

Written by

Hosono P | the prop firm strategist

I buy challenges with my own money and record everything through to the payout. Recorded payouts: ¥6.1M in total from Fintokei, Fundora and Funded7, plus $4,776 from The5ers (as of September 2026). Author of the semi-discretionary EA "ELDRA".

Profile and payout recordX @hosono_p

📚Related articles

🔍

Only Two Prop Firms Can Actually Run Grid/Trailing-Repeat EAs: "Not Banned" and "Usable" Are Different Things

If you want to run a grid or trailing-repeat ("trailing stop and repeat" / トラリピ) type EA on a prop account, checking only whether the ban list mentions it doesn't tell you anything. What actually kills the account is the "floating-loss cap" and whether drawdown is measured on equity or on balance. City Traders Imperium's 1-Step is the unique combination of balance-based DD, no daily DD, off-the-shelf EAs allowed, and grid explicitly permitted. Fintokei doesn't list grid among its 9 banned practices, but off-the-shelf EAs aren't allowed and a combined 3% cap across all open positions becomes the real limit on how many grid levels you can run. A comparison table across 10 firms, plus the formula for working backward from the floating-loss cap to figure out how many levels you can stack. Confirmed against each firm's official help center on September 16, 2026.

Firm comparisons9/16/2026
🔍

FTMO Futures vs Topstep: FTMO Wins on Expected Value, Topstep Wins on Monthly Cost. If You're Running a Bot, VPS Access Is the Deciding Factor [September 2026]

We re-confirmed the rules on both the futures prop leader Topstep and FTMO Futures, which entered the market in September 2026, on September 23, 2026, and compared them. The monthly fee is less than half at Topstep, but running the same lot size through evaluation to payout for a year gives FTMO Futures Pro the higher expected value ($50K, zero edge: +$4,836 vs. +$2,519). Covers activation fees, the fact that you can only withdraw 50% of the balance, an optional daily loss limit, resets that shift your billing date, VPS eligibility, and running a bot on a live account.

Firm comparisons9/22/2026
🔍

Breakout Prop vs. Fintokei: Which Should You Buy? 'Narrow Room, No Rules' vs. 'Wide Room, Many Rules'

A head-to-head comparison of Breakout Prop, the Kraken-backed crypto prop firm, against Fintokei, the firm that runs end-to-end in Japanese. The fee as a percentage of account size is nearly identical (Breakout Pro 0.545% vs. Fintokei Sapphire 0.549%), but the substance is opposite. Breakout has just two rules — daily loss and max drawdown — with no consistency rule, no minimum trading days, and no per-trade risk cap, but its max drawdown is a narrow 3–6%. Fintokei has a wide 10% max drawdown, but a cap on floating loss at 3% of balance applies to the sum of all open positions. Looking at target ÷ max drawdown, how easy it is to pass differs by 1.7x to 3.3x. Measured on both official sites on September 16, 2026.

Firm comparisons9/16/2026
🔍

FTMO Futures Complete Guide: Automated Trading and VPS Both Officially Allowed as FX Giant Launches Futures in September 2026

A hands-on check of FTMO's official pages for FTMO Futures, the futures prop program FTMO launched on September 1, 2026. Its forbidden-practices page explicitly states "Automated trading is permitted," making it one of the few firms that also explicitly permits VPS and VPN use. Covers full pricing for every size of Growth and Pro, how EOD trailing drawdown works, the consistency rule that only applies during evaluation, the JST close-out deadline, the three platforms (NinjaTrader, Tradovate, TradingView), and whether it's usable from Japan.

Firm comparisons9/22/2026
🔍

E8 Pro vs E8 One: Which Should You Buy? Comparing "Static 8% + 2%/day cap" vs "Dynamic 6% + 40% Consistency" at the Same $366, for Both Smoothed and Spiky P&L [September 2026]

E8 Markets' E8 Pro (static DD 8%, daily 2.5%, daily profit cap 2%, payouts are 50% of profit and the fail line moves to the starting balance) and E8 One (dynamic DD 6% locked at the starting balance, daily 4%, target 9%, 40% consistency rule in Performance stage) both cost $488 at $100K — $366 with code E8. The answer depends on the shape of your P&L. Traders firing multiple times a day to smooth out returns come out ahead with Pro in Monte Carlo (4.5x vs 1.4x at zero edge). Traders who trade once a day and earn most of their profit from a handful of big days come out 2–7x ahead with One in real-data tests (+¥3.09M/+¥5.85M vs +¥630K/+¥1.37M per year). The reason: Pro's "2% daily profit cap" shaves 8–27% off big-day profits, and the buffer disappears on the first payout. Also covers which configuration to pick (8% for Pro, 6% for One).

Firm comparisons9/20/2026
🔍

Web3 Prop Firms Special [September 2026] | Sorting 22 Firms — from On-Chain Settlement to Exchange-Backed and Prediction-Market Types — by 'Is This Actually On-Chain?'

New prop firms settling on Hyperliquid or Arbitrum surged in 2026. We checked 22 firms on September 20, 2026, against their official rulebooks, terms of service, and Trustpilot — from the on-chain settlement type (Propr, Hypernova, HyperPNL, Carrot, DojiFunded), the chain-recorded type (Vanta, Hyper Stack, Solana Funded), the on-chain payout type (Investabl, FundedHive, AI Prop), Kraken-owned Breakout / Kraken Prop / Kraken Funded, the crypto-native type (SizeProp, HyroTrader, Crypto Fund Trader, Bitfunded, TradeXProp, Upscale), to the prediction-market type using Polymarket (Funding Predicts, PropMarket). Even firms calling themselves 'on-chain' run every evaluation phase as a simulation, and each firm's terms state that whether real orders get routed to the market after funding is entirely 'at the operator's discretion.' We cover whether you can buy from Japan, what to do once you receive USDC, and where the fees hit.

Firm comparisons9/20/2026