● Learn

How to spot a fake trading bot track record

Most trading bot results you see online cannot be checked, and some are simply made up. Here are the signs, and what real evidence looks like.

Seven red flags

1. Only screenshots. An image of a balance or a trade list proves nothing. Anyone can edit one.
2. A curve with no bad stretches. Real strategies have losing runs and drawdowns. A perfectly smooth line usually means cherry-picked trades or a curve drawn after the fact. Run the worst drop you can see through our drawdown recovery calculator to see what gain it actually demands back.
3. A convenient start date. If the record begins right after a crash or right before a rally, ask what the years before it looked like.
4. “Tested on every coin.” If a backtest uses today’s list of coins, it has quietly left out every coin that went to zero. That flatters any strategy.
5. No costs. Results with no fees, no slippage and no funding are results nobody could have traded.
6. No code and no data. If you cannot rerun it, you cannot verify it.
7. Guarantees. Nobody can promise returns. A seller who does is telling you something about themselves, not the market.

What real evidence looks like

• The code that produced the numbers, which you can run.
• Data from a public source, pinned so you know it is the same data.
• Every cost included: fees, slippage and funding.
• Results labelled clearly for what they are, whether a backtest or a live record, with the dates shown in full.
• Rules fixed and dated before the results were judged.

How we publish ours

APEX and CARRY publish backtests, and say so next to every figure. Each package ships the code and the command that regenerates its figures from public data on your own machine, the costs are inside that calculation, and the research pre-registrations are anchored in the Bitcoin blockchain. If you would rather not take our word for any of it, you do not have to.