QuantJourney Backtester

QuantJourney Backtester

Share product feedback

Thank you.

Your note is now in the QuantJourney inbox.

Validation services

Statistics, not opinion

A strategy, a track record, or an external manager's pitch is a claim about the future built on a finite sample of the past. We adjudicate that claim — with statistics, not opinion — applying the same permutation tests, walk-forward discipline and selection-bias corrections documented in our research, independently, to your material, and returning a signed, reproducible report.

Three engagements, productized

Fixed scope, fixed timeline, one deliverable each — quoted up front at the intake call, no open-ended retainers.

Strategy Validation Report

For teams and individuals with a strategy — code or return series.

You provide the strategy (Python against the engine, or just the returns). We run the full battery: dependence-aware Monte Carlo permutation tests, bootstrap confidence intervals, Deflated Sharpe, canonical CSCV PBO when a complete trial matrix exists (rolling rank stability otherwise), date-bounded walk-forward with explicit pre-OOS purging, cost ladder and crisis-window analysis.

DeliverableSigned PDF report with a reproducibility fingerprint — every number replayable from the archived config and seeds.
Timeline2–3 weeks
PricingFixed-scope quote at the intake call

Track Record Review

For family offices and allocators evaluating an external manager.

You provide the track record and the manager’s stated process. We analyze what the returns can actually support: luck-versus-skill significance, selection-bias exposure, consistency of the realized series with the declared strategy, drawdown and crisis behavior versus the peer claim.

DeliverableA concise review memo your IC can read in twenty minutes — findings, red flags, and the questions to put to the manager.
Timeline1–2 weeks
PricingFixed-scope quote at the intake call

Research Process Review

For funds preparing for operational due diligence.

We audit the research pipeline itself: data provenance and point-in-time discipline, look-ahead surfaces, validation standards, search logging, reproducibility of published numbers. The output is a gap analysis against the standard an institutional reviewer will apply.

DeliverableGap-analysis report with prioritized remediations, plus a re-audit pass once they land.
Timeline2–4 weeks
PricingFixed-scope quote at the intake call

What you receive

A document your investment committee can interrogate — not a tear sheet. Every claim traces to an archived artifact.

The verdict leads. One page, in plain language: what the evidence supports, what it rules out, and — the discipline most reports omit — what it cannot decide at this sample size. Behind it sits the proof: every permutation test with its null distribution, every walk-forward fold, and the deflation of your headline Sharpe against the full trial history that produced it.

The fingerprint is what makes it evidence. Config, data lineage, seeds and package versions hash into a single identifier printed on the report; anyone holding the archive regenerates every number bit-for-bit — months later, on other hardware, for a regulator or an allocation committee. A validation you cannot replay is an opinion with formatting.

A negative verdict is a result, not a failure. A share of engagements end in "consistent with luck" — and that finding, delivered before capital is committed rather than after, is the least expensive risk decision on the table.

How it works

01

Intake call

Thirty minutes: what you have, what decision the result feeds, what format the material is in.

02

NDA and material

NDA first when needed. Code, returns or track record arrives — or never leaves your infrastructure at all.

03

Analysis

The full pipeline runs with seeds pinned and every artifact archived. We flag interim blockers, not surprises.

04

Report and review call

The signed report plus a working session: what the numbers mean, what they cannot mean, and what we would do next.

Your alpha stays private. For sensitive strategies the code never leaves your infrastructure: we run the pipeline on-prem, or your team runs it under our supervision and we validate the outputs. The engine is open source — you can audit every line the verdict rests on.

Selected engagements

Multi-strategy fund · United States

Validation gates for a discretionary-to-systematic desk

We installed statistical gates — permutation tests, nested walk-forward, Deflated Sharpe — on every candidate. Weak ones were retired before allocation; survivors now ship with evidence packets the IC can interrogate.

Family office · Gulf region

From spreadsheet research to an auditable engine

We migrated spreadsheet-and-notebook strategies onto the engine, reconciled against the historical numbers, and handed over a monthly review cadence. Every allocation now traces to an archived, seeded run.

Systematic PM · United States

Execution realism for a futures book

A persistent sim-vs-live gap. We rebuilt the backtest on the orders path and found the culprits: rebalance cadence and roll costs. Simulation now tracks live within tolerance.

Independent trader · Europe

From one strategy to a research workflow

One strategy, one question: is it real? We ran the full pipeline, then installed the same workflow on their machine. The strategy survived at reduced size; two queued improvements did not.

Representative work from five years and 50+ clients — described by role and region, never named without written permission.

Start with the intake call

Tell us what you have — strategy code, a return series, a manager's track record, or a team that needs the discipline installed. We respond within two business days with a scope and a quote.