Phase 1 target
+10%
Close positions once the target is reached
FTMO 2-Step · Historical Research
A historical comparison of risk allocations under the modelled challenge rules, alongside the documented result of the first attempt. These challenge-study allocations are separate from the proposed IC Trading Book risk weights.
Historical replay: €80,000 account · 2020-04-01 → 2026-07-31
Rule set
Phase 1 target
+10%
Close positions once the target is reached
Phase 2 target
+5%
New account; close positions at the target
Maximum loss
10%
Static floor below each phase’s initial capital
Maximum daily loss
5%
Previous daily close less a fixed initial-capital amount
Minimum activity
4 days
Four active days required in each phase
Each sampled date starts a fresh €80,000 account. Positions are treated as closed when a target is reached. Verification then restarts with a new account and a +5% target. Rates exclude unresolved windows. Start dates overlap and daily-loss checks use a proxy, so these are sample outcomes rather than future pass probabilities.
Engine A
0.1% risk per trade
Pass rate
100%
Fail rate
0%
252 pass · 0 fail · 98 unresolved of 350 windows
Engine B
0.1% risk per trade
Pass rate
100%
Fail rate
0%
226 pass · 0 fail · 123 unresolved of 349 windows
Engine C
0.5% risk per trade
Pass rate
96%
Fail rate
4%
349 pass · 16 fail · 31 unresolved of 396 windows
Combined (A+B+C)
A 0.1% + B 0.1% + C 0.5% per trade
Pass rate
86%
Fail rate
14%
273 pass · 44 fail · 23 unresolved of 340 windows
Tested range: A/B at 0.1–0.2% and C at 0.1–0.5% per trade. Results are ranked by historical pass rate, then median completion time. These scaled historical comparisons do not establish an optimal allocation; exact-size backtests are a separate check.
Under FTMO 2-Step, A 0.1% · B 0.1% · C 0.1% has the highest historical pass rate among resolved windows (98%). “Balanced” stays within 3 percentage points of that result; “Faster” is the shortest median pass among allocations at or above 90%.
| Allocation | Pass* | Fail* | Median pass | Resolved / total |
|---|---|---|---|---|
| A 0.1% · B 0.1% · C 0.1% · Highest pass rate | 98% | 2% | 173d | 283 / 340 |
| A 0.1% · B 0.1% · C 0.2% · Balanced | 96% | 4% | 162d | 292 / 340 |
| A 0.1% · B 0.1% · C 0.3% | 93% | 7% | 141d | 300 / 340 |
| A 0.2% · B 0.1% · C 0.1% · Faster | 91% | 9% | 128d | 287 / 340 |
| A 0.2% · B 0.1% · C 0.2% | 89% | 11% | 118d | 294 / 340 |
| A 0.1% · B 0.2% · C 0.1% | 89% | 11% | 124d | 285 / 340 |
| A 0.1% · B 0.2% · C 0.2% | 88% | 12% | 113d | 294 / 340 |
| A 0.1% · B 0.1% · C 0.4% | 88% | 12% | 120d | 309 / 340 |
| A 0.1% · B 0.2% · C 0.3% | 87% | 13% | 105d | 300 / 340 |
| A 0.1% · B 0.2% · C 0.4% | 86% | 14% | 95d | 310 / 340 |
| A 0.1% · B 0.1% · C 0.5% | 86% | 14% | 107d | 317 / 340 |
| A 0.2% · B 0.1% · C 0.3% | 86% | 14% | 108d | 300 / 340 |
| A 0.2% · B 0.1% · C 0.5% | 85% | 15% | 84d | 318 / 340 |
| A 0.2% · B 0.2% · C 0.1% | 85% | 15% | 91d | 289 / 340 |
| A 0.2% · B 0.2% · C 0.2% | 84% | 16% | 85d | 299 / 340 |
| A 0.2% · B 0.1% · C 0.4% | 84% | 16% | 97d | 309 / 340 |
| A 0.1% · B 0.2% · C 0.5% | 83% | 17% | 85d | 321 / 340 |
| A 0.2% · B 0.2% · C 0.3% | 82% | 18% | 74d | 305 / 340 |
| A 0.2% · B 0.2% · C 0.5% | 80% | 20% | 64d | 322 / 340 |
| A 0.2% · B 0.2% · C 0.4% | 80% | 20% | 69d | 313 / 340 |
* Rates exclude unresolved attempts. Start windows overlap and are not independent trials; daily-loss rules use the proxy described below.
Compare historical pass rates with median time to pass across the tested allocations. Both measures depend on the same overlapping sample.
Each point is a constant allocation. The dashed line connects allocations not dominated by another allocation on both speed and historical pass rate. Rates exclude unresolved windows. Historical starts overlap and are not independent trials.
These plans size the +10% Challenge and the fresh +5% Verification account separately. They directly test the idea of pursuing the first phase faster, then reducing risk before Verification. Funded-account sizing is deliberately excluded: it has no profit target and needs a separate survival-and-payout study.
| Plan | Challenge | Verification | Pass* | Fail* | Failure phase | Median pass | Modelled completion time with retries† | Resolved / total |
|---|---|---|---|---|---|---|---|---|
| Aggressive throughout | A 0.1% · B 0.1% · C 0.5% | A 0.1% · B 0.1% · C 0.5% | 86% | 14% | 31 challenge · 12 verification | 107d | 138d | 317 / 340 |
| Aggressive → balanced | A 0.1% · B 0.1% · C 0.5% | A 0.1% · B 0.1% · C 0.2% | 88% | 12% | 31 challenge · 6 verification | 124d | 163d | 307 / 340 |
| Aggressive → conservative | A 0.1% · B 0.1% · C 0.5% | A 0.1% · B 0.1% · C 0.1% | 88% | 12% | 31 challenge · 5 verification | 126d | 170d | 290 / 340 |
| Balanced throughout | A 0.1% · B 0.1% · C 0.2% | A 0.1% · B 0.1% · C 0.2% | 96% | 4% | 11 challenge · 0 verification | 162d | 180d | 292 / 340 |
| 0.5% nominal sensitivity check | A 0.1% · B 0.1% · C 0.3% | A 0.1% · B 0.1% · C 0.3% | 93% | 7% | 20 challenge · 2 verification | 141d | 166d | 300 / 340 |
* Rates exclude unresolved attempts. Start windows overlap and are not independent trials; daily-loss rules use the proxy described below. Failure phase counts show where the failed attempt ended. † Modelled completion time with retries models repeated, immediate retries after the observed failure duration, sampling from resolved historical windows; it assumes independent attempts even though the underlying windows overlap. The highlighted row is the proposed aggressive-to-balanced policy, not a recommendation or forecast.
Three risk policies compared under the same historical model. The results are not forecasts or selected production allocations.
Research objective: Comparing higher Challenge risk with lower Verification risk.
Trade-off: The tested policy uses lower risk in Verification; its historical outcomes are shown above.
Research objective: Comparing completion time when the model assumes immediate retries after failed attempts.
Trade-off: The fastest route in the immediate-retry model can also expose Verification to the most risk.
Research objective: Comparing the highest historical pass rate among these three policies.
Trade-off: Compare its historical failure rate and median completion time with the other policies.
This historical comparison uses: C at 0.5% per trade versus A 0.1% · B 0.1% · C 0.2% in both phases. It is not a matched-exposure comparison.
Engine C
0.5% risk per trade
Pass rate
96%
Fail rate
4%
Balanced Three-Engine Book
A 0.1% · B 0.1% · C 0.2% · both phases
Pass rate
96%
Fail rate
4%
C alone and the balanced three-engine book both passed in 96% of resolved historical starting windows. The balanced book reached the target sooner (162d versus 215d median), while using a lower 0.4% nominal maximum allocation when all three engines overlap, versus 0.5% for C alone.
For a speed-focused reference, the aggressive three-engine policy (A 0.1% · B 0.1% · C 0.5%) reached a 107d median pass, with a 86% historical pass rate among resolved windows.
Assumptions & Limitations
What is modelled, what is estimated, and how to use the results.
Rules. Phase 1 (challenge) passes at +10%, then Verification restarts at the initial capital and passes at +5%. Both phases enforce a static 10% maximum-loss floor from their initial capital and a 5% maximum daily-loss amount. The first violation before verification passes is a fail. Switch the rules tabs to compare the standard and stricter stress-test scenarios.
Rolling windows. Start dates are sampled every few trading days across the backtest. Each window is rebased to a fresh account at that date. Once Phase 1 passes, the next return sequence is rebased again to a fresh Verification account. Windows that reach the end of the backtest unresolved are counted as incomplete rather than fail.
Source data. The backtests use FTMO's historical backtest data. Its available history begins on 1 April 2020, so the study cannot test earlier market regimes.
Daily drawdown. Daily loss is measured from the previous daily close less a fixed percentage of initial capital. Days are grouped in UTC rather than CE(S)T, which can change daily-loss outcomes around the reset. Individual-engine tests use reported minimum-equity points when available. Combined and allocation tests use compounded balance changes without reconstructing simultaneous floating P/L, so their loss-limit checks can miss breaches caused by open-position losses. These simulations cannot fully reproduce real-time equity, swap and commission checks.
Minimum trading days.FTMO requires four days with a position opened in each phase. The source reports do not expose position-open events, so this model uses four distinct days with recorded account activity as a clearly labelled proxy.
Combined book. The combined series reuses the book simulation convention: per-trade return fractions from each engine are merged by timestamp and compounded onto one shared €80,000 account. It answers “what happens when all three run on the same account.”
Interpreting rates. Rolling windows overlap, so they are not independent trials. These rates describe this historical return sequence, not hundreds of unrelated future outcomes.
Advanced Diagnostics
Supporting evidence showing the long-run equity behaviour behind the simulations.
The default view focuses on the combined account, its peak-drawdown context, and daily-loss proxy breaches. Individual engine curves are available when needed.
Historical simulations cannot predict future outcomes, but they can show how risk allocations behaved across the same market history. This study is designed to make allocation decisions more informed and transparent—not to forecast returns.