Strateva

Independent validation for systematic strategies

A second opinion for your backtest.

Automated independent validation for daily cash-equity, spot-FX and supported Binance spot-crypto strategies. We freeze the scope, reproduce the result, examine its fragility and issue a deterministic verdict before any prospective shadowing.

Connect the Strateva MCPView a sample report

No derivatives · No optimization to force a pass · No AI-generated verdicts

The product lives in your AI

Connect Strateva to Claude or ChatGPT

Add this exact MCP server URL as a custom connector. OAuth signs you in; no Strateva app or copied API key is required.

The copied value must end in /mcp. Do not copy the address of this web page.

  1. 01Open custom connectors or MCP settings in Claude or ChatGPT.
  2. 02Paste the exact URL shown here and complete OAuth sign-in.
  3. 03Tell your AI the rules or paste your code. The connector guides the validation to the PDF.

What can be validated

Daily cash equities and ETFs, daily spot FX, and supported Binance spot-crypto pairs. No intraday bars and no derivatives.

  • Equities and ETFs, spot

    One listed share or exchange-traded fund bought and sold outright, on daily data in the current automated path.

  • Crypto, spot — supported Binance pairs

    Supported spot pairs are acquired from Binance on daily 24×7 bars. Perpetual swaps, leveraged tokens and every other derivative remain excluded.

  • Currency pairs, spot

    One spot FX pair on daily data. The exposure is the currency itself, not a derivative wrapper.

No options, futures, perpetuals, CFDs, leveraged tokens or structured products.

See the full scope

A historical curve is not a decision

A historical curve can hide overfitting, unrealistic costs, dependence on one particular period, or a sequence of returns that happened to be exceptionally kind.

Strateva does not try to improve your strategy. It tries to refute it before the market does.

Backtesting and validation answer different questions

A backtester is the right tool for building a strategy. It is not the tool that tells you how much to trust the result it produced — that is a different question, and it needs a different instrument.

A backtester asks

How much would it have made?

Strateva asks

  • Is it reproducible?
  • Does it survive outside the observed sequence?
  • Does it hold up against costs and reasonable changes?
  • Is there enough evidence to move to shadowing?

Both are necessary. One builds the result; the other decides how much weight it can carry.

How a validation works

Four steps, from the material you already have to a decision you can act on.

  1. Define

    Send code, Pine Script, a trade list or written rules. Strateva turns the material into an exact scope, and you confirm it.

  2. Execute

    We freeze the rules, the data, the costs and the benchmark. The strategy runs in an isolated environment.

  3. Verify

    The engine reproduces the metrics and examines robustness, time dependence, costs, benchmark and alternative paths.

  4. Decide

    You receive the evidence, the limitations, and a result that tells you what the next step is.

Every validation ends in one of three results

  • Ready for shadowing
  • Revision required
  • Insufficient evidence

AI prepares the strategy. Strateva’s closed engine validates it. Where AI stops

See the full methodology

Three evidence levels

Know what kind of evidence you are buying.

Every report names the highest level actually reached. Higher levels require the protocol and time they claim; they are never inferred from a strong backtest.

  1. Included now

    1. Reproduced history

    The frozen strategy is replayed on the confirmed historical data, with declared costs, a benchmark when valid, robustness simulations and a reproducible evidence chain. This is historical evidence, not out-of-sample proof.

  2. When eligible

    2. Walk-forward OOS

    Chronological folds test decisions on later historical observations. Rolling WFA evaluates one frozen candidate; nested WFA is used only when multiple candidates were frozen before selection. It is stronger historical evidence, but still historical.

  3. Requires future data

    3. Prospective holdout

    The strategy, costs, benchmark, horizon and decision rule are sealed before new observations exist. Results remain hidden until the horizon completes. It cannot be delivered immediately or reconstructed from old data.

These are evidence levels within one service, not performance ratings. A result may be unfavourable or not evaluable at any level.

Proprietary implementation. Inspectable evidence. Deterministic verdict.

Use the assistant you already know

Claude or ChatGPT can collect your material, identify missing information and prepare a reviewable Validation Scope through MCP. The AI does not enter the quantitative engine.

Once you confirm the scope, Strateva’s closed engine processes the frozen inputs and produces the metrics, simulations, evidence and deterministic verdict.

The engine implementation is proprietary. The scope, assumptions, evidence and decision trail are documented.

Where AI stops, in full

What gets validated

  • Methodology

    Causality, labels, validation splits, benchmarks and experimental design.

  • Economics

    Commissions and fees modelled with their source and effective date, plus how sensitive the result is to cost and turnover. Spreads, slippage, capacity and latency are reviewed as declared assumptions: if they are missing or implausible, the report says so.

  • Statistics

    Uncertainty, stability, sensitivity, multiple testing and incremental value.

  • Implementation

    Code, financial accounting, data integrity, tests and reproducibility.

What you receive

A twelve-part report: the verdict first, then the scope you confirmed, the metrics, the benchmark, the alternative paths, the findings, the limitations and the next step.

  • Verdict

    One result, its reason, and what happens next.

  • Validation Scope, confirmed

    The exact claim, asset, timeframe, costs and benchmark you agreed to.

  • Permuted Monte Carlo

    The same returns, reordered — how much of the risk was sequence luck.

  • Limitations

    What this validation could not establish.

See a sample report

Optional support

Validation is sponsored. Supporting Strateva is optional.

A contribution does not buy a validation, unlock a case, improve a verdict or move anyone ahead in the queue. It only helps cover infrastructure and research costs.

Make a voluntary contribution

Public case studies

  • RF100Validation auditRevision required

    RF100 — Can a Random Forest beat simple momentum?

    Market
    US equities, spot
    Strategy type
    Cross-sectional machine learning

    What was checked

    A causal cross-sectional Random Forest found statistically positive predictive structure but failed to beat a simple momentum benchmark under equivalent execution assumptions.

    View repository on GitHub
  • EWMA_63_StratValidation auditRevision required

    EWMA63 — Does beta selection add value beyond de-risking?

    Market
    US equities, spot
    Strategy type
    Volatility-based allocation overlay

    What was checked

    The audit showed that most of the historical Sharpe and drawdown improvement came from lower market exposure, while the incremental value of the EWMA63 allocation layer was approximately zero.

    View repository on GitHub
  • F2-ANTICIPATOR-V2Research exampleRevision required

    Historical research — outside the current commercial scope

    F2 Anticipator V2 — Can a late-book signal anticipate 5-minute BTC Up/Down markets?

    Market
    Polymarket prediction markets
    Strategy type
    Order-book microstructure signal

    What was checked

    A causal, book-only model estimated the final probability of Up and traded only when expected edge exceeded costs. The temporal OOS showed that the predicted edge did not survive realistic execution assumptions.

    Read the case studyView repository on GitHub
  • Simple-StrategyResearch exampleNo formal outcome — published as research

    Volatility Term-Structure Strategy

    Market
    SPY ETF, spot
    Strategy type
    Volatility term-structure regime filter

    What was checked

    A transparent SPY allocation example combining a lagged VIX term-structure filter, realized-volatility targeting, leverage limits and explicit turnover costs.

    Read the case studyView repository on GitHub

See all case studies

Independent, and explicit about its limits

  • No stake in the answer

    No commission, no product to sell alongside the verdict, no upside in a strategy passing. A negative outcome is worth exactly as much as a positive one, and is delivered as readily.

  • The limits are published, not disclosed on request

    Every report states what it could not establish. A capability published without its limit reads as sales; published with it, it can be checked.

  • Not investment advice

    A validation reports what held up under the agreed checks over the period examined. It is never a recommendation to allocate capital and never a promise of future performance.

  • Your material stays yours

    Nothing you send is published, reused or shown to anyone else. The public examples on this site are strategies belonging to Strateva itself.

Before you risk capital, check what the result actually rests on.

Check compatibility

Validation is sponsored. Optional contributions never affect access or verdicts.

Contact

Have a strategy, backtest or model that needs an independent review?

Contact by email

The information you send will be used to assess your request. Read the privacy policy.