Backtest and AI Disclaimer

1Hypothetical results

Backtest results are hypothetical. They describe what a set of rules would have done against recorded data. They are not a prediction and not a promise.

Simulated performance carries a limitation real trading does not: it is prepared with the benefit of hindsight. The strategy was written by someone who already knew what the market did next, even if they were not consciously fitting to it.

2Why backtests overstate

• Overfitting. Tune enough parameters against one history and you will find a combination that fits it beautifully and generalises to nothing.
• Look-ahead bias.A rule that uses a bar's close to decide an entry within that bar cannot be traded, but it will backtest well.
• Survivorship bias. Instruments that were delisted, depegged, or collapsed often do not appear in historical datasets. The losers went missing.
• Slippage and fees. Simulations fill at prices real books do not always offer, and small per-trade costs compound into large differences.
• Liquidity. The size you tested may not be available at the price you tested it at.
• Regime change. Historical data contains the conditions that happened, not the conditions coming.

These are properties of backtesting itself, not defects in our engine. No backtester eliminates them, and one that claims to is worth less trust, not more.

3Data quality

Historical and live market data comes from third parties. It can contain gaps, bad prints, mis-timestamped bars, or corrections applied after the fact. We do what we reasonably can to detect and handle that, and we do not warrant that market data is accurate, complete, or timely.

4AI review and strategy analysis

NexPulse can send text you submit — journal notes, a strategy definition — to a large language model and show you what comes back. Understand what that is:

• It is a generated summary, produced by a statistical model. It is not a person, not an analyst, and not a reviewer with judgement.
• It can be confidently wrong. Models invent details, misread numbers, and produce fluent text with no basis in your data.
• It is not investment advice and not a recommendation, whatever it happens to say. If the output reads like advice, that is a property of how language models write, not a change in what it is.
• It is not deterministic. The same input can produce different output.

Never place a trade because the AI review suggested it. Verify anything it tells you against your own data before you act on it. Where the text you submit is processed is set out in the Privacy Policy.

5Figures shown on this website

Numbers published on nexlots.com — metric counts, feature capabilities, timing claims — describe what the software does or measures. None of them is a return you should expect, an average user outcome, or a performance claim. Where we show a number we can point at the code that produces it; where we cannot, we do not show it.

6Third-party content

News headlines, economic calendar entries and fundamentals come from external providers and are passed through for convenience. We do not verify them, we do not endorse them, and we are not responsible for their accuracy. External links are not recommendations.