Disclaimer

This page states plainly what Quant Compose is not, alongside what it is, because a research tool working with financial data invites assumptions it needs to correct directly rather than leave implicit.

This is a research tool, not investment advice

Nothing produced by Quant Compose — a backtest result, a risk statistic, a portfolio optimisation, a validation verdict, a generated report — constitutes investment advice, a recommendation, or a solicitation to buy, sell, or hold any security, financial instrument, or asset class. The workspace is designed to help you test and evaluate quantitative research questions with appropriate statistical rigour. It is not designed, and should not be used, as the sole basis for a real financial decision. If you are making decisions about real capital, you should seek advice from a qualified, appropriately regulated financial adviser who can take account of your personal circumstances, which this workspace has no knowledge of and no ability to assess.

Every generated evidence report carries a verdict of supported, not supported, or insufficient evidence for the specific research question it addresses, and every one of those verdicts is accompanied by the explicit statement that it is not investment advice. This is not boilerplate added as an afterthought; it reflects what the workspace’s statistical output actually is and is not capable of telling you.

No execution, no live order routing

Quant Compose is a research and paper-mode product. It does not place trades. It does not route orders to any broker, exchange, or execution venue. It does not connect to a live or paper trading account anywhere in its current design. The live_order_routing status reported by the product’s own health check is, and will remain during this phase of the product, false. A strategy that reaches an “approved” or “live” stage in the workspace’s internal lifecycle has passed through a gated evidence-and-approval process; it has not been connected to any brokerage, and no order has been placed on your behalf as a result.

If a future version of the product were ever to introduce live order routing as an optional, explicitly enabled capability, that would be a materially different feature requiring its own clear disclosure, its own regulatory consideration, and your own affirmative choice to enable it — it is not a capability quietly waiting to be switched on today.

Past performance

Any backtested or historical performance figure shown in the workspace — a Sharpe ratio, a return series, a drawdown statistic — describes how a strategy or portfolio would have performed on the specific historical data it was tested against, under the specific assumptions about cost, slippage, and execution the backtest applied. Past performance, however carefully validated, is not a reliable indicator of future results. Markets change character over time, the assumptions embedded in any backtest are simplifications of what a real execution would actually involve, and a strategy’s edge, if genuine, can decay or disappear entirely for reasons no historical backtest could have shown in advance. This is precisely why the workspace reports validation statistics like the worst fold and the degradation grade alongside a headline return figure, rather than the headline figure alone.

Synthetic and sample data

Some data available inside the workspace, including its bundled reference dataset, is synthetic: deterministic, seeded, and generated without any network call to a real market data source. Synthetic data is provided so a researcher can explore the workspace’s mechanics, test a workflow’s structure, or learn how a method behaves, without depending on a live data connection. Synthetic data is never a basis for a claim about the behaviour of any real security, and any result produced from it should be understood as illustrative of the method, not descriptive of any actual market. Wherever the workspace surfaces a result derived from synthetic or sample data, it labels that data’s origin so it cannot be mistaken for a live market feed.

Data provider limitations

Real market data connected through the workspace’s providers — including free sources such as Yahoo Finance — carries the limitations of those providers, including restatement risk after corporate actions and the coverage gaps described for each provider in our product documentation. The workspace does not silently substitute synthetic data when a real data request fails; a failed fetch is reported as a failure, not masked with a plausible-looking replacement.

No guarantee of outcome

Quant Compose’s validation methodology — walk-forward testing, purging and embargoing, multiple-testing corrections, degradation grading — is designed to make a research result harder to overstate, not to guarantee that a validated result will succeed if pursued further. A strategy that passes every check the workspace applies has cleared a meaningful bar of statistical scrutiny; it has not been guaranteed to work, and no amount of validation can eliminate the fundamental uncertainty of financial markets.

Contact

If anything in the workspace’s output is unclear about what it does or does not claim, contact contact@quantcompose.com.