Strategy Attribution
Definition
Strategy attribution is the verification feature where multi-strategy EAs publish each module's contribution to overall return separately on the verified live account. It distinguishes serious documentation discipline from aggregate-only claims that hide module-level performance variance.
In-depth: Strategy Attribution
Strategy attribution emerged as a verification expectation specifically for multi-strategy EAs because multi-strategy architectures are easy to misrepresent. A vendor can claim a 4-module multi-strategy product but actually have 3 modules producing flat returns and one module generating all the edge — making the product structurally equivalent to a single-strategy EA in terms of risk concentration. Strategy attribution prevents this misrepresentation by exposing per-module results.
What strategy attribution reveals:
• **Edge concentration**: if one module produces 90% of returns while three others are flat, the product is single-strategy with multi-strategy marketing. The buyer should expect single-strategy risk concentration rather than the diversification benefit the marketing implies • **Module decay**: if a module that was producing edge in months 1-6 has gone flat in months 7-12, the strategy is decaying and the vendor's retraining or update cadence is failing to maintain it • **Regime sensitivity**: attribution by module reveals which strategies are earning in which regime conditions; trend modules earn during trending months, mean-reversion modules earn during ranging months, etc. • **Risk allocation honesty**: vendors who advertise 30% allocation to module A should show that module A's risk contribution matches the marketing; mismatches indicate either misrepresentation or undocumented post-launch allocation changes
The 2026 editorial expectation: serious multi-strategy vendors publish per-module attribution monthly. The leading example is ThemisEA MT5, which pioneered per-module reporting in 2024 — the vendor maintains a public dashboard showing each strategy module's contribution separately. Market Trader AI Pro publishes similar monthly attribution analysis.
Vendors who don't provide per-module attribution face buyer scrutiny: either (a) the vendor doesn't track per-module performance internally (which suggests inadequate engineering investment in the architecture) or (b) the vendor tracks it but doesn't publish (which suggests the attribution would reveal unflattering concentration patterns).
For EA buyer workflow: when evaluating a multi-strategy product, ask the vendor explicitly for per-module attribution on the verified live account. Vendors who provide it immediately are typically serious about the architecture; vendors who deflect with "the modules all work together" framing are either masking concentration or hand-waving on an architectural claim they don't actually engineer to.
For product comparison purposes, per-module attribution lets buyers cross-shop across multi-strategy products on apples-to-apples bases. A multi-strategy product with 4 modules each producing 1-2% monthly is fundamentally different from one where 1 module produces 6% and 3 modules produce 0% even though both might headline at 6% monthly returns.