When Should You Change Your Prop Account Settings? Don't Touch It During Evaluation — Always Change It the Moment You Get Funded

※ This article is organized around each firm's official rules as of September 18, 2026. It is not investment advice. Rules are updated frequently, so always check each firm's official site before purchasing.

The same temptation always shows up when an account starts sinking during a challenge.

"Should I raise my lot size and try to make it back in one shot?" "This time slot hasn't been good lately, should I switch?" "Maybe this method just doesn't work after all."

The conclusion up front: there is no statistical basis for making that call partway through the evaluation phase. What you must do instead is always change your settings the moment you reach a funded account. For some firms, not changing them is itself a rule violation.

Most people do these two things backwards.

The conclusion: what to keep fixed, and what to change

Fixed (until the account dies)... day of week / time of day / trade direction / entry and exit conditions
Variable (only at phase boundaries)... lot size / stop-loss distance / news filter

That's the whole distinction. Below is why it works this way.

Why you shouldn't touch anything during evaluation

Why "change it because you're losing" doesn't hold up

Say Method A and Method B exist, and A has a 40% pass rate while B has 25%. A is clearly the better method.

So — how many accounts would you need to actually detect that difference from real challenge results?

※ Calculated using standard statistical criteria (5% significance level, 80% power).

150 accounts. You probably have one or two.

This is the same story as a coin toss. Can you tell a slightly biased coin apart from a fair one by flipping each 3 times? No — even a fair coin lands heads 3 times in a row about 1 in 8 times. A challenge's pass/fail outcome carries information just as coarse as that.

In other words, both "it blew up, so this method is bad" and "it passed, so this method is good" are almost entirely talking about luck. Change your settings on that basis, and all you're doing is chasing noise.

Adding more accounts doesn't speed up verification

This is the counterintuitive part. You'd think "if I scale up to 10 accounts, won't I find out faster?"

No, you won't. Running the same method across 10 accounts still gives you the same single trade on the same day. If you open and close the same trade across 10 accounts simultaneously, that isn't 10 pieces of data — it's 1 piece of data copied 10 times.

What matters isn't the number of accounts — it's the number of signals.

No matter how hard you try, reviewing the method itself can't happen more than about once a year. Trying to judge it within a few months will always be too soon.

Some of this is also banned by the rules

Setting the statistics aside, mid-challenge changes also carry rule-related risk.

• FundedNext and TradingCult explicitly require maintaining the same strategy across the Challenge and Funded account. In particular, "pass with an EA during evaluation, then switch to manual on the funded account (or vice versa)" is banned by name
• Breakout Prop lists "changing your strategy after passing the evaluation" among its banned practices
• Alpha Capital can, at its own discretion, judge extreme scalping or oversized lots and drop you into a restriction tier called the Focused Trader Group. Landing there means leverage capped at 1:30, a 1% cap per position, and EAs disabled entirely

Raising your lot size on a sinking account is exactly the kind of move most likely to trip these discretionary clauses.

The moment you reach funded, the rules change on you

On the other hand, phase boundaries are an entirely different story. Here, rules that didn't exist during evaluation kick in all at once.

Hantec is the clearest example. A method you could build without worrying about news events during evaluation stops being viable the instant you go funded. And since closing a position also counts as a violation, any design that holds a position across a news release runs straight into it.

"I'm sticking to my original plan, so I won't change anything" is a direct route to a violation here.

"Changing your strategy" and "adjusting your lot size" are different things

This raises something that looks like a contradiction.

• FundedNext says "maintain the same strategy"
• Hola Prime says "cut your risk to under 2% once funded"

Which one do you follow? The answer is you follow both, because what you're changing is different in each case.

Strategy is...    when, in which direction, under what conditions you enter, and how you exit
Risk is...        how large you execute that method

Halving your lot size is not a strategy change. You're just shrinking the same method to fit a smaller container. Your entries and exits haven't moved a millimeter.

A genuine strategy change is a swap of the substance itself — like "EA during evaluation, manual once funded" or "scalping during evaluation, swing trading once funded." That's what's explicitly banned.

An account deep in drawdown is already a different product

There's one more point that's easy to overlook.

A sunk account is no longer the same account you bought. In numerical terms, it has effectively morphed into a completely different plan.

Looking at it as "target ÷ remaining room"

The difficulty of a prop account is largely determined by this one ratio.

Target ÷ remaining room = how much more % you must win ÷ how much more % you can afford to lose

The smaller this value, the easier; the larger, the harder. Calculated for a ¥20M account (8% target, 10% max drawdown), it looks like this:

The account at -¥1M sits at 2.60. That's more than 3x heavier than a brand-new account's 0.80. That's roughly the same difficulty as one of the industry's toughest plans (Breakout Turbo's 3.00).

Even though you bought the same product, it has turned into something else entirely.

How much value is actually lost

We measured the expected value at each stage of drawdown across 40,000 Monte Carlo runs.

※ For a method equivalent to PF 1.30 (52% win rate, 1.2 risk-reward). With zero edge, the -¥1M figure is 0.58; even with a quite strong method, it's 0.76.

Once an account is drawn down to half its max drawdown limit, treat it as having about 70% of the capacity of a brand-new one.

That's why "aiming for a recovery" doesn't pay off

This is the most important part.

Clawing back ¥1M on a damaged account only gets you back to "an account at breakeven." You haven't moved a millimeter closer to the target. And that "account back at breakeven" is something you could buy outright for a ¥110,000 fee on a brand-new account.

Meanwhile, earning ¥1M on a brand-new account completes 62.5% of a ¥1.6M target.

The same ¥1M, but the value is nowhere close to equal. Here it is in numbers:

Even with zero edge, rebuying wins. If paying ¥110,000 gets you back ¥1.82M worth of capacity, there's no real dilemma here.

But that means "buy more while leaving it running," not "kill it and rebuy"

To be clear, this doesn't mean you need to go out of your way to kill a sunk account.

That account still carries ¥4M of expected value, and holding onto it costs nothing. There's no reason to kill it.

• If you can afford to buy more, buy more right away (it's a clear win, no need to compare)
• Let the sunk account keep running as is (it'll either die on its own or pass on its own)
• Never raise the lot size (recovering it only gets you back to brand-new status, while raising the odds of it dying)

The only situation that genuinely calls for a hard decision is when there's a cap on how many accounts you can hold at once. For example, at firms where warnings accumulate across accounts, a sunk account is occupying one of your slots. That's the only case where you should consult the table above.

Only 3 triggers for a review

To sum up, the review framework looks like this.

And the important part is that the act of "swapping out" doesn't really exist as a task.

Accounts die on their own if you leave them alone. So there's no need to decide "which one to cut." Just changing where your next purchase gets assigned lets the portfolio naturally rotate over time. No stop-loss decision, no exit decision required.

Summary

Here's the life cycle of an account:

Purchase → decide on a method
  ↓
Phase 1 ─┐
         │ Don't touch anything here (there's no basis for a decision)
Phase 2 ─┘
  ↓
★ Halve the lot size, turn on the filter (mandatory — because the rules change on you)
  ↓
Funded → withdraw once you hit the minimum payout amount

• The only time it's okay to touch anything during evaluation is when the reason comes from outside — a rule change, an EA bug, an abnormal instrument. "It's not going well" doesn't count as a reason
• Always change something at phase boundaries. Not changing is itself the violation risk
• The only things you change are lot size, stop-loss distance, and filters. The substance of the method stays fixed to the end
• A sunk account has about 70% of the capacity of a brand-new one. Don't chase a recovery — prioritize buying more instead

We've published the script. Download sim-damaged-account-value.py (Python + NumPy, fixed random seed). Rewrite ACCOUNT, TARGET_P1, MAX_DD, and RISK to get the same tables for your own plan and method.

FAQ

Q. Is changing your method mid-challenge a rule violation?

It depends on the firm. FundedNext and TradingCult explicitly require "maintaining the same strategy across the Challenge and Funded account," and in particular ban by name switching from an EA to manual (or vice versa) after passing. Breakout Prop also lists "changing your strategy after passing the evaluation" among its banned practices. Even at firms without an explicit rule, some — like Alpha Capital's Focused Trader Group — have discretionary clauses that can drop you into a restriction tier, so it's safer to err on the side of caution.

Q. Does changing your lot size also count as a "strategy change"?

No. Strategy means "when, in which direction, under what conditions you enter, and how you exit," while lot size is "how large you execute that." As long as your entry and exit conditions haven't changed, you're just executing the same strategy smaller. In practice, since firms lower your risk cap once you're funded, not lowering your lot size is what would actually be the rule violation.

Q. If losses start piling up, shouldn't I switch methods?

You won't have enough data to make that call. Telling a 40%-pass-rate method apart from a 25% one using real results requires 150 accounts per method. Concluding "this method is bad" from 1-2 accounts' results is the same as judging a coin's bias from 3 flips. The quality of a method should be judged by per-trade performance over a year-scale time frame, not by whether an individual account passed or failed.

Q. Does adding more accounts finish verifying a method faster?

No, it doesn't. Running the same method on 10 accounts at the same time still produces the same single trade on the same day. It's not 10 pieces of data — it's 1 piece of data copied 10 times. What determines the outcome is the number of signals, so adding more accounts doesn't speed up verification by even a millimeter. For a method equivalent to PF 1.30, it takes 224 signals (about 10 months at 5 trades a week) before you can call the edge real.

Q. Should I give up on an account that's deep in drawdown?

You don't need to give up on it, but don't expect much either. An account drawn down to half its max drawdown is only worth about 70% of a brand-new one (0.58 to 0.76 depending on skill). Since holding it costs nothing, it's fine to just let it keep running. But trying to raise your lot size to make it back is the worst move — even if it recovers, all you get is "the same capacity as brand new," which you could buy outright anyway.

Q. Isn't leaving a sunk account alone and buying a new one a waste of money?

The opposite. The gap in expected payout between a new account and one at -¥1M is ¥1.82M (for a method equivalent to PF 1.30), against a ¥110,000 fee — a 16.6x return. Even calculated at zero edge, it's still 1.76x, so rebuying still wins. That said, there's no need to kill the sunk account. Since holding it costs nothing, the optimal move is to leave it running and buy more on top of it.

Q. How often should I do a review?

Drive it by the calendar and by events, not by an account's outcome. Specifically: (1) immediately, if the firm changes its rules or a payout issue comes up (just stop buying more from that firm), (2) 5 minutes, every time you buy more, to decide what the next one gets assigned to, and (3) once a year, to compare the method's track record against the backtest. Swapping out the method itself only happens at point 3.

Related articles

• Challenge distillation: only let the survivors move forward
• What Dan Cheung means by "rotation"
• Position sizing and lot-size calculation for prop firms
• Protecting a funded account and reaching payout
• Comparison of prop firm rules