Backtest Lens
Инструменты
Анализ локально в браузере
EA CORRELATION ANALYZER

EA correlation analyzer

Upload EA reports and compare strategy behavior over shared calendar periods. The correlation matrix reveals strategies that tend to win and draw down together, even when their symbols and trading logic appear different.

What you get
MT4 / MT5 HTMLNo installLocal first analysis
PURPOSE-BUILT ANALYSIS

Test diversification with evidence, not strategy names

Two EAs trading different currency pairs can still lose during the same months. Monthly-return correlation describes that shared behavior, while portfolio drawdown shows its financial consequence.

01

Strategy matrix

Compare every EA pair using overlapping calendar months.

02

History overlap

See whether each estimate is supported by a meaningful common period.

03

From correlation to portfolio

Evaluate the matrix alongside combined drawdown, Equity Stress and strategy contribution.

PORTFOLIO EXAMPLE

AO Extremum: five strategies on a shared timeline

The public AO Extremum portfolio combines five selected strategies and 3,522 trades. Its matrix compares monthly strategy returns only over overlapping periods, while the portfolio charts reveal the capital consequence of that relationship through combined return, drawdown and Equity Stress.

Open the AO Extremum correlation →
5strategies
3,522trades
Monthscomparison unit
Overlapshared coverage
01

What the coefficient means

A positive value means monthly strategy results tended to move in the same direction; a negative value means they tended to move oppositely. A value near zero indicates no strong linear relationship in the observed sample, but it does not prove independence during an extreme market event.

  • Read direction and strength, not only the cell color.
  • Check how many shared calendar months support the value.
  • Compare relationships over full and common-history periods.
02

Why monthly returns instead of individual trades

Different EAs do not need to trade at matching timestamps or hold positions for the same duration, making trade-to-trade pairing economically weak. Calendar-month returns give systems with different frequencies a shared time axis while preserving the direction and scale of each period.

  • High- and low-frequency EAs receive a comparable timeline.
  • Months without trades remain part of coverage.
  • Weak overlap should not produce a confident conclusion.
03

Correlation does not replace portfolio drawdown

Low average correlation does not rule out simultaneous losses during one stressed window. Use the matrix with the combined curve, Balance DD, Equity Stress and strategy contribution. Allocation decisions should be based on the full evidence set rather than one correlation threshold.

  • Identify contributors to the maximum stressed interval.
  • Compare equal and adjusted lot multipliers.
  • Validate the selected composition on OOS periods.
SCOPE AND LIMITS

What this analysis does not prove

  • Correlation describes historical linear association and can change between market regimes.
  • A short common period makes the coefficient unstable.
  • Negative historical correlation does not guarantee protection from future joint drawdown.
HOW IT WORKS

From MetaTrader report to actionable evidence

1Save EA reports in one folder
2Open folder portfolio analytics
3Compare correlation, coverage and combined drawdown
FAQ

Questions before you upload

How is EA correlation calculated?

Backtest Lens compares monthly strategy returns using only calendar months available for both reports.

What is a good correlation value?

There is no universal cutoff. Interpret it with overlap length, combined drawdown and the portfolio objective.

Does negative correlation guarantee protection?

No. Historical relationships can change, particularly during stressed market regimes.

Can I compare symbols inside one account history?

Yes. Account-history analysis also provides a separate correlation view by instrument.

BACKTEST LENS

Analyze the report before you trust the result.