Using a different risk profile after passing a challenge can make sense, but the decision needs a test. It should not rest on a slogan that evaluation must be aggressive and funded trading must be conservative. The account rules, payout conditions, horizon and strategy distribution determine whether a change helps.
The clearest first experiment keeps entries and exits unchanged and varies only position sizing. This isolates the effect of risk. If you also change entry filters, trading hours and targets, you are testing a different strategy rather than a different risk profile. Both experiments are legitimate, but their conclusions should not be mixed.
Write down the objective of each stage
During evaluation, the objective may be reaching a profit target before a loss boundary. A program may or may not have a formal deadline. Even without one, you can evaluate completion within a planning horizon because fees and time still matter. Do not assume that every challenge requires a target within 30 days.
After evaluation, the relevant objective may become expected cash payouts over a defined period while complying with continuing rules. A funded label does not by itself establish whether the account trades live capital or remains simulated. Use the provider's actual contract and model the rewards and obligations that apply to that account type.
Rules can change between stages. Loss limits, consistency tests, position caps, payout buffers and permitted trading times may differ. A sizing policy that fits the first stage is not automatically compatible with the second. Build separate rule configurations, then connect them in a complete purchase-to-payout analysis.
Use one signal history for a controlled comparison
Consider a hypothetical strategy with ten sequential net outcomes in risk units: +1.5, −1, +0.5, −1, +2, −1, −1, +1.5, −1 and +2.5R. Their sum is +3R. For this teaching example, assume outcomes scale linearly with cash risk and that there are no quantity constraints or additional costs.
The cumulative sequence is 1.5, 0.5, 1, 0, 2, 1, 0, 1.5, 0.5 and 3R. The largest decline from an earlier peak is 2R, from the fifth-trade peak to the seventh-trade trough. These values let us compare fixed cash sizing without changing the strategy's signals or assuming that the same percentage of current equity is risked.
| Fixed cash risk per R | Net result over all ten trades | Largest closed-equity drawdown | Peak-to-trough loss in R |
|---|---|---|---|
| $250 | $750 | $500 | 2R |
| $500 | $1,500 | $1,000 | 2R |
| $1,000 | $3,000 | $2,000 | 2R |
Both profit and drawdown scale in this unrestricted example. The underlying expectancy per R did not improve when size increased. Now introduce a $1,500 account loss boundary or a $2,000 target and paths stop at different times. The contract can convert a simple scale change into a different realized sequence of outcomes.
Compare complete policies rather than isolated risk numbers
A risk profile should specify its basis. Is $500 fixed throughout the attempt, is risk a percentage of current equity, or does it shrink with remaining buffer? Does it stop after two daily losses? Does it reduce exposure near a payout date? These rules change the path and should be implemented before the test.
Suppose the evaluation policy risks $500 per trade and the funded policy risks $250. If the same signal process produces a 2R drawdown, the cash decline halves. However, the time needed to accumulate a fixed payout threshold may increase. A smaller position can reduce one kind of failure while extending the period during which other failures can occur.
The policy also needs a transition rule. A clean version changes profiles only after the evaluation is formally completed and the next account is activated. A vague instruction to reduce risk once things look good allows discretionary changes that cannot be reproduced. State the trigger and preserve it across simulations.
Measure the metrics that match each objective
For evaluation, report the probability of passing by the chosen horizon, probability of failure, unfinished share and time to pass among successful paths. For the funded stage, report probability of receiving a payout, expected net payouts, survival duration and remaining headroom after withdrawals. These are related but different quantities.
A policy with the highest evaluation pass rate might reach the funded stage with an unfavorable trading habit or be poorly matched to payout restrictions. Conversely, a policy with lower passing probability can have better expected cash value if it produces more durable funded outcomes. The complete decision requires both stages rather than a ranking on one number.
Do not compare a 30-day evaluation simulation with a two-year funded simulation and attribute all differences to risk. Use appropriate but clearly stated horizons, and where possible also include a common-horizon comparison. Report costs and unresolved attempts consistently.
Preserve sequence dependence across the transition
If market regimes persist, the sequence after passing is not necessarily independent of the sequence before it. A trend-following strategy may pass near the end of a favorable trend and begin the next stage during a reversal. Independently drawing a fresh favorable sample for the funded stage can miss this relationship.
One way to examine it is to run the stages sequentially through historical or block-sampled sessions. Reset the account state when the contract calls for a fresh account, but continue the market process. Compare that result with an independent-stage model to understand how much the conclusion depends on regime persistence.
When using historical windows, do not select only starting dates that pass quickly. Include all eligible starts under a predeclared sampling rule. Otherwise you condition the analysis on successful evaluation paths and make the later stage look better than a new purchase actually deserves.
Model withdrawals and buffers explicitly
Suppose a hypothetical funded account has equity of $104,000 and an active floor of $99,000. It has $5,000 headroom. A $2,000 withdrawal, with the floor unchanged, reduces equity to $102,000 and headroom to $3,000. Continuing at the same $500 risk consumes a larger share of the remaining buffer.
Some contracts adjust floors or impose separate payout eligibility rules. Apply those definitions rather than assuming every withdrawal has the example's effect. The broader principle is that distributions to the trader and capital left inside the account interact. A high gross profit does not necessarily mean a large immediately withdrawable amount.
Expected payout analysis should include attempts that receive nothing. Estimating average cash only among successful payout recipients omits the probability of getting to that state. To compare profiles economically, follow each simulated purchase through its fees, evaluation, funded outcomes and actual eligible payments.
Do not turn a sizing test into hidden strategy selection
It is tempting to use the most aggressive variant that passed development, then the smoothest variant from validation. If those choices are made after inspecting both periods, the combined policy has been selected with hindsight. Freeze the policy definition before its final evaluation and count the alternatives that were tried.
Changing trade frequency is also more than a pure scale adjustment. Restricting the strategy to one signal per day changes which trades exist. Skipping news events changes the sample. Label these as separate input settings or code variants, and generate their own trades. Do not attach the old profit history to the new rules.
For reproducibility, save the code version and inputs behind each risk profile, plus the evaluation and funded rule configurations. A compact name is useful for navigation, but it cannot substitute for the exact calculation identity. Results belong to the configuration that produced them.
Check operational consistency
A theoretical two-profile system is only useful if the trader or automation can apply it consistently. Specify allowed quantities, maximum simultaneous exposure and how to handle a profile change with open positions. A switch that unexpectedly doubles one leg or leaves a stale stop can create risks absent from the backtest.
Paper testing can help verify the transition mechanics. Record intended and actual position sizes, orders, fills and account-state updates. The purpose is not to prove profitability from a short forward sample. It is to discover whether the operational policy matches the tested one.
Behavioral pressure can change execution, but do not assume every trader becomes reckless in evaluation or cautious afterward. Test observable behavior: missed signals, manual overrides and deviations from planned risk. If those occur, include their effects in the review rather than attributing everything to market randomness.
Keep stage labels separate from market data labels. Evaluation and funded describe contractual states, while development, validation and holdout describe the research role of observations. A funded policy can be designed on development data and tested on validation. Mixing these two naming systems makes it difficult to tell which evidence was genuinely independent.
A compact experiment plan
- Choose one version of the strategy and one untouched evaluation sample.
- Define a small set of executable sizing policies before running comparisons.
- Use the same trade paths to compare profiles under each stage's rules.
- Simulate the transition and payout events, including fees and withdrawals.
- Stress costs, weaker outcomes and persistent adverse regimes.
- Prefer a stable region of results over a narrowly optimized risk point.
The best outcome may be one policy for both stages, or a clearly defined change in size after evaluation. There is no need to introduce two strategies merely because the labels differ. Start with the smallest meaningful change and measure whether it improves the objective you actually care about.
A disciplined two-profile approach is a testable contract between the strategy and the account rules. It preserves the trading logic where possible, makes transitions explicit and evaluates cash outcomes beyond the first pass. That is more informative than assuming that faster evaluation and smoother funded equity can both be obtained simply by turning a risk dial.