suggested historical data window for re-optimisation — average of 3 scientific methods
—
✦ RE-OPTIMIZATION FREQUENCY
—
how often to re-run optimization
DERIVED FROM PERIOD ⓘ
STATISTICAL
ⓘ
—
Rolling variance stability
BENCHMARK
ⓘ
—
Asset class & timeframe
SPECTRAL
ⓘ
—
Return series memory
REGIME
ⓘ
—
Volatility adjustment
SHARPE RATIO
—
annualised
MAX DRAWDOWN
—
peak-to-trough
REGIME CHANGES
—
3σ structural breaks
DATA PERIOD
—
historical coverage
RAW STATISTICAL
—
before vol adjustment
ASSET RATIONALE
Observational analysis only. Values derived from loaded backtest data. Not financial advice.
SIMULATION PARAMETERS
RESAMPLING METHOD
All three rebuild your equity curve thousands of times. They differ in how the trades are drawn.
WINNERLOSERSHADE = SIZETIME RUNS DOWNWARD
PermutationRESHUFFLE
What if these same trades had arrived in a different order?
One line in, one line out. Every trade used exactly once.
KEEPSYour exact set of trades. Final balance identical every runBREAKSThe order, which is the point of itUSE FORSequence risk. Was the smooth ride real or helpful ordering
BootstrapRESAMPLE
What if a fresh sample of trades had been drawn from this strategy?
With replacement. #4, #6 feed two lines each. Faded ones (#3, #5) never drawn.
KEEPSThe shape of your trade distributionBREAKSThe exact set, so balance varies. Assumes trades independentUSE FORSampling uncertainty. How much rested on the trades you got
Block BootstrapADVANCED
What if a fresh sample were drawn while keeping runs of consecutive trades intact?
Lines arrive in parallel bundles. Each ring is a run lifted whole.
KEEPSLocal structure. Losses that bunched together still canBREAKSLong range order, holding short range dependenceRUN LENRead from your trades, not a setting. Reported with the result
COMPARE TWO METHODS IN ONE RUN
Bootstrap vs Permutation
Separates sequencing luck from sampling luck.
Bootstrap vs Block
Both draw with replacement, so the only difference is whether runs were kept together. The clean way to see what clustering costs.
Your Edge Matrix Score is computed from the same method every time, whichever you select, so strategies stay comparable. When a strategy shows no clustering the run length falls to roughly one trade, and block bootstrap then describes the same thing as bootstrap. The result panel says so rather than implying a difference that is not there.
Bootstrap: resample with replacement. Best for general robustness.
Permutation: reorder trades. Best for overfitting checks.
Block Bootstrap: resample in short runs so clustered losses survive. Block length is read from your trades.
Skip random trades
Robustness Analysis
MONTE CARLO
Stress-test your strategy with randomized equity paths
SHUFFLE
Randomize
Reorder trade sequences to remove path dependency
PROJECT
Simulate
Generate thousands of possible equity curve outcomes
QUANTIFY
Measure
Estimate drawdown risk with statistical confidence
> RUN SIMULATION
Configure options in the left panel to begin
Equity Distribution Projection
—
/100
ROBUSTNESS
GRADE—
DD STABILITY
—
PATTERN ANALYSIS
—
INTEGRITY CHECK
—
⚑ FLAGS
PROFITABLE
—
ORIGINAL DD ⓘ
—
DD 50% CONF ⓘ
—
DD 95% CONF ⓘ
—
DD 99% CONF ⓘ
—
All statistics computed from the full simulation set · Fan Chart & Percentile Bands use exact server-side bands ·
methodology
STRATEGY INSIGHTS Trading EA |
EXPECTANCYE = (WR × AvgW) − (LR × AvgL)
CAGRCAGR = (VF/V0)^(1/n) − 1
SHARPE RATIOS = (R − Rf) / σ × √252
PROFIT FACTORPF = ΣWins / Σ|Losses|
KELLY CRITERIONKB = W − (1−W) / RR
WIN RATEWR = Wins / Total × 100
SORTINOSo = (R − Rf) / σd
Z-SCOREZ = (X − μ) / σ
CALMAR RATIOC = CAGR / MaxDD
RECOVERYRI = DD / Expectancy
RISK/REWARDRR = AvgWin / |AvgLoss|
STRATEGY INSIGHTS
DECODING TRADING PATTERNS
Click ▶ RUN ANALYSIS to begin
🏆 COMPLETE OPTIMIZATION GUIDE
📊 Calmar Ratio
×
The Calmar Ratio measures risk-adjusted return by comparing annualized returns to maximum drawdown.
Formula: Calmar = Annualized Return ÷ Max Drawdown %
Rating Scale:
• ≥3.0 = Exceptional (professional quality)
• 2.0–3.0 = Excellent
• 1.5–2.0 = Professional
• 1.0–1.5 = Acceptable
• 0.5–1.0 = Marginal
• <0.5 = Risk-Heavy
Higher is better — indicates more return per unit of drawdown risk.
CLOSE
STRATEGY COMPARISONNo strategies
EMStrategy Comparison0 strategies loaded
Compare. Validate. Decide.
Load MT4, MT5, or cTrader backtest files and let Edge Matrix run full validation — 19 statistical tests, Monte Carlo simulation, and automated ranking — on every strategy at once.
✓
19 Validation Tests
Temporal, edge quality, holding time, cliff detection...
READY
○
Monte Carlo Simulation
500 equity curve permutations per strategy
READY
≡
Ranked Comparison
Auto-ranked flip cards with strengths & warnings
READY
↓
HTML Export
Full comparison analysis report
READY
▤
Correlation Matrix
Pearson r · daily P&L · portfolio diversification score
READY
EMStrategy Queue
RUNNING ANALYSIS
ELAPSED 0:00
Read files
parsing…
Computing
queued
Render
queued
▐ RETURN CORRELATION MATRIX
HIGH r>0.70
MOD r>0.40
LOW r<0.40
NEG
Pearson r · daily-bucketed P&L · portfolio method (no-trade days = $0) · click any cell for full explanation
×
CORRELATION STATS
PEARSON r
daily P&L
STRENGTH
classification
OVERLAP
shared variance
WHAT THIS MEANS
DAILY P&L SCATTER —
← daily P&L →
Metric
—
—
WHAT THIS MEANS FOR THIS STRATEGY
○
Trading EA
REPORT ENGINE
EDGE MATRIX
Modules Ready
Trade Data
Edge Matrix
Monte Carlo
Projection
Optimal Period
Insights
◆
Generate
Full Report
0 of 6 modules ready
Report Sections
Executive Summary
Analysis & Projection
Edge Matrix (19 Tests)
Monte Carlo
Strategy Insights
Final Verdict
Load a backtest file to begin — report generates all modules automatically
Load a backtest in the Import tab to run What-If analysis.
CHOOSE A REALITY
CUSTOM CONTROLS
90%95%99%
0.0
0.0
1.0
10
HrDay
0
0
CntProfit
3.0
BTLow
Stress scenario selected.
EDGE MATRIX — WHAT-IF
Stress-test the whole strategy
01
Win-Rate Confidence
True rate at the bottom of its range?
—
02
Execution Cost Sweep
Broker worse than the backtest?
—
03
Time Concentration
Edge only lives in a few hours?
—
04
Outlier Dependency
Best trades don't repeat?
—
05
Inevitable Streak
The streak your WR guarantees?
—
ⓘ
Time finding
—
BACKTEST vs WHAT-IF
EQUITY — ORIGINAL vs WHAT-IF
OriginalWhat-If
PROP FIRM
Presets verified June 27, 2026. Prop rules change; verify on the firm's site before relying. Static-drawdown firms only.
PICK A PROP FIRM
Load a backtest in the Import tab to run Prop Firm Validation analysis.
Prop Firm Validation
WHAT DO YOU WANT TO ANALYZE?
One backtest, two questions. The path forks here.
▲ PROP CHALLENGE PASS
"Would my strategy have passed a prop firm challenge?"
Fast or aggressive strategies, 30 to 60 day horizons. Pass probability, days to target, failure breakdown. For traders attempting a fresh evaluation.
SELECT THIS PATH→
▼ LONG-TERM PROP STABILITY
"Would my strategy have held a funded account without blowing up?"
Stable or conservative strategies, 1-year horizon. Suitability score, drawdown survival analysis. For funded traders or conservative monthly grinders.
SELECT THIS PATH→
Not sure? Pick CHALLENGE PASS if you are attempting a fresh evaluation. Pick LONG-TERM PROP STABILITY if you already have a funded account or trade conservatively.
Choose a preset on the left to load its rule set, or pick Custom to enter your own. You can adjust any value, and toggle off rules that don't apply to your challenge.
TRADING DAYSActivity requirements during the challenge
HOLDING RESTRICTIONSWhen trades cannot be held open
Flag any trade open through Friday close.
Requires news calendar. Not enforced in v1.
Uncheck any rule that doesn't apply to your specific challenge. Disabled rules are skipped in the analysis. When you're ready, click RUN ANALYSIS at the top.
VALIDATION SUSPENDED
Statistical integrity issues detected in this backtest
Prop Firm Validation analysis is suspended because the backtest data shows statistical patterns that make any rule-compliance claim unreliable. Running the analysis anyway would produce numbers that look meaningful but aren't.
INTEGRITY WARNING
Statistical anomalies detected. Treat the results below with caution.
CHALLENGE PASS SCORE
-
%
-
WHY THIS SCORE →
- Monte Carlo paths passed every enabled rule
Stationary block bootstrap simulation (Politis-Romano 1994). Each path samples 30 trading days using random-length blocks from the backtest, preserving day-to-day serial dependence (volatility clustering, losing streaks). The score measures how often the strategy passes under randomized but realistic conditions, not just on the historical sequence.
95th PCTILE MAX DD
-%
MEAN MAX DD
-%
FORWARD 30-DAY WORST DD
-%
TARGET REACH
-
DAYS TO TARGET
-
Q1 · MEDIAN · Q3
TIME LIMIT
- days
-%
-
LONG-TERM SUITABILITY
WHY THIS SCORE →
EDGE-RELIABILITY WEIGHTED
- of - simulated 1-year paths stayed within the Max DD rule
1-YEAR DRAWDOWN INSTRUMENTS · AGAINST THE 10% RULE
REALISTIC WORST ⓘ
-%
MEDIAN ⓘ
-%
TYPICAL YEAR
EXTREME ⓘ
-%
RARE TAIL
MEAN ⓘ
-%
AVERAGE YEAR
ANNUAL RETURN ⓘ
-%
GEOMETRIC MEAN
HORIZON ⓘ
1 yr
365 DAYS
WHERE PATHS FAILED
—
0
0 / 0 paths (0%)
Max DD Breach
0
0 / 0 paths (0%)
Daily Loss Breach
0
0 / 0 paths (0%)
Did Not Hit Target
0
0 / 0 paths (0%)
Min Days Not Met
INACTIVITY RULE
SIMULATED CHALLENGES · CLICK FOR DETAILS
Four representative paths from the 1,000-path Monte Carlo, with background sample
CLICK CHART TO EXPAND
WHAT THE ANALYSIS FOUNDDESCRIPTIVE · NOT ADVICE
These notes describe patterns in historical backtest data under simulation. They are not financial advice and do not predict future results. Any change to a live strategy is the trader's own decision.
No observations were generated for this analysis.
← BACK TO SUMMARY
Run an analysis to see how the score was built.
PER-RULE PASS RATE
Independent pass rate of each enabled rule
Each engine checks one specific rule across all 1000 Monte Carlo paths. The headline shows the relevant metric, the threshold shows your rule limit, and the margin shows how much room (or how much over) the strategy has.
Per-Rule Pass Rate
Headline
Threshold
Margin
A8 · KILLER TRADES IDENTIFIER
3 historical trades violated at least one enabled rule
3 FLAGS
Specific trades from your backtest that would have triggered a rule breach. Sorted by severity. Past trades only, not predictions.
Date
Symbol
Cause Type
Severity
P/L
Rule Violated
DISTRIBUTIONS · SHAPE OF THE 1,000 SIMULATIONS
Four views of the Monte Carlo simulation distribution
The Summary tab shows headline numbers (Mean DD, Pass Rate). These charts show the underlying shape of the distribution. A single number can hide whether outcomes are clustered tight or spread wide. Click any chart to expand.
MAX DD DISTRIBUTION
Drawdown reached on each of the 1,000 paths
FINAL BALANCE DISTRIBUTION
Account balance at end of each simulated horizon
DAYS TO TARGET (PASSERS ONLY)
Time taken to reach profit target across passing paths
CUMULATIVE PASS RATE BY DAY
Fraction of paths resolved as PASSED on or before each day
WHEN PATHS FIRST BREACHED THE DD RULE ⓘ
M1
M2
M3
M4
M5
M6
M7
M8
M9
M10
M11
M12
PROP FIRM VALIDATION
—
—
—
Ready — Load data to begin
Statistical analysis tool only — not financial advice