Risk Disclosures

Version 2026-08-14. Please read this page carefully. It explains what the numbers on InvestLab (the “Service”) are — and, just as importantly, what they are not. It supplements, and should be read together with, our Terms of Service.

1. No investment advice, no recommendation

InvestLab is an educational and analytical platform. Nothing on the Service — no strategy, backtest, simulation, ranking, commentary, or article — is investment advice, a personal recommendation, or an invitation or inducement to buy or sell any financial instrument. Content is generic: it is computed from rules and market data, and it does not know or consider your financial situation, objectives, or risk tolerance. We are not your adviser, broker, or fiduciary. Before making any investment decision, consider seeking advice from an adviser who is licensed in your jurisdiction and who knows your circumstances.

2. No offer; jurisdiction

Nothing on the Service is an offer or solicitation to anyone in any jurisdiction in which such an offer or solicitation would be unlawful, or to anyone to whom it would be unlawful to make it. The Service is not directed at any person in any jurisdiction where its publication or availability would require us to hold a licence, registration, or authorisation we do not hold. If you access the Service, you are responsible for compliance with the laws that apply to you where you live.

3. Investment risk

All investing involves risk, including the possible loss of the capital you invest. In addition, some strategies shown on the Service use techniques that materially increase risk:

Leverage amplifies both gains and losses, and adds financing costs that vary over time. A levered strategy can lose more, faster, than an unlevered one.
Short positions lose money when the shorted asset rises, and that loss is in principle unlimited.
Volatility targeting and momentum rules change positions mechanically and can concentrate a portfolio or move it to cash at unhelpful moments.
Currency: instruments denominated in a currency other than your own add exchange-rate risk on top of market risk.

4. Hypothetical and simulated performance

This is the most important section on this page. Backtested results shown on the Service are hypothetical. They do not represent actual trading, and no account ever achieved them. Hypothetical performance has inherent limitations, including:

Hindsight. Backtests are designed with knowledge of how markets actually behaved. Strategy rules that look good over a period were, unavoidably, selected because they look good over that period. Live results are routinely worse than backtests for exactly this reason.
No capital at risk. Simulated positions were never held, so the results reflect none of the market impact, liquidity constraints, or forced decisions of real trading — and none of the discipline required to hold a losing strategy through a drawdown.
Modelled costs. Transaction costs, financing, and (where modelled) taxes are estimates applied by formula. Real costs differ by broker, size, and market conditions, and were often higher in the past than the model assumes.
Proxy and spliced data. History before an instrument existed is reconstructed from index or asset-class proxies. These splices are disclosed where used, but a proxy is not the instrument: tracking differences, costs, and availability are simplified.
Data revisions. Market data comes from third-party providers and can be corrected or restated; numbers on the Service can change when the underlying data does.

“Live” strategy pages are also simulations. They apply each strategy’s published rules to market data on an ongoing basis. They are not brokerage records, they are not audited, and they do not show anyone’s actual account.

5. Forward-looking simulations

Some tools (for example Portfolio Construction and the IPS simulator) show distributions of possible future outcomes generated by statistical simulation from historical data. These are illustrations of a model, not predictions. The ranges shown depend entirely on the assumptions used, real outcomes can fall outside them, and the future can differ from the past in ways no simulation of history captures.

6. Past performance

Past performance — whether realised or simulated — is not a reliable indicator of future results. A strategy that performed well historically can perform poorly in the future, and the conditions that favoured it may not repeat.

7. AI-generated content

Some written commentary on the Service (for example market notes and portfolio analysis narratives) is generated by artificial-intelligence models from the numbers we compute and, where stated, cited news sources. It is reviewed for method but not guaranteed to be free of errors, and it can mis-state or over-simplify. Treat it as a starting point for your own verification, not as a conclusion.

8. Data and accuracy

Market data on the Service comes from third-party providers. We work seriously on correctness, but we do not warrant that any figure, dataset, signal, or simulation is accurate, complete, or timely, and we may correct, restate, or remove content at any time. If you spot something that looks wrong, please tell us.

9. Contact

Questions about these disclosures: stephane@investlab.app.