Why past performance is genuinely hypothetical (in CFTC terms)

You have read the sentence a hundred times and skimmed past it every time: "Hypothetical performance results have certain inherent limitations." It looks like legal throat-clearing, the fine print you scroll through to get to the numbers. It is not. Every clause of that disclaimer is a specific, true statement about why a backtest is not a track record, and the people who wrote it were describing exactly the trap that catches new traders.

What "hypothetical" means here

Hypothetical performance is any performance that did not come from a real account with real money in it. A backtest is the clearest example: you take a strategy, run it over historical data, and record what it would have done. No orders were sent, no money was at risk, no broker filled anything. The result is a simulation, and a simulation of trading is a fundamentally different object from a record of trading. The distinction is spelled out in what a backtest is.

Where the disclaimer comes from

The wording is not invented by marketers. In the United States, the display of hypothetical trading results is governed by the Commodity Futures Trading Commission, principally under CFTC Regulation 4.41, and by the National Futures Association under Compliance Rule 2-29. Between them they require that simulated or hypothetical results be clearly labeled as such and accompanied by a cautionary statement, because regulators watched a great many people mistake a backtest for a promise.

That regulatory backdrop is part of why automated trading is a perfectly legal, ordinary activity in the US when it is described honestly, a topic covered in is automated trading legal in the US. The rules do not ban showing a backtest. They ban showing one as if it were a live account.

The disclaimer, clause by clause

Here is the heart of the standard CFTC language. It reads like boilerplate until you translate it.

Simulated results do not represent actual trading. Since the trades have not been executed, the results may have under- or over-compensated for the impact of certain market factors, such as lack of liquidity. Simulated programs in general are also designed with the benefit of hindsight.

Three real claims are packed in there.

Why each part is literally true, not legal cover

Take hindsight first. When a strategy is developed, the developer can see the whole chart. It is genuinely difficult to build rules that are not, in some small way, shaped by knowing how the story ended. That is the same danger described in overfitting and curve-fitting: a strategy tuned too closely to its own history can look brilliant on the data it was born from and fall apart on data it has never seen.

Then execution. A backtest fills your order at a price on a chart. A live order, especially in the fast seconds around the 9:30 AM ET cash open, can slip, get a partial fill, or miss entirely. The gap between the assumed fill and the real one is not theoretical, and it is broken down in slippage and commissions. It always works against you, never for you.

And risk. The simulation has no emotions, no rent to pay, and no urge to switch itself off during a drawdown. It takes trade number 41 with the same discipline as trade number 1. The person running the live version is the variable the backtest cannot model, and usually the reason live results fall short of simulated ones.

A backtest is the score of a game played against a defense that never showed up.

Why publish a backtest at all

If a backtest is this limited, why lead with one? Because the alternative honest options are worse. A brand-new system has no live track record to show, and a long live record is exactly the thing that can be quietly cherry-picked or faked. A backtest has one property the others lack: it is reproducible. You can load the same strategy into your own NinjaTrader 8, run the same period, and see whether the report prints the same result. If it does not, you have learned something before spending a dollar.

That is also the whole reason every figure here is published net, with the commission and a tick of slippage already deducted, and the profit factor printed alongside the worst drawdown. A gross number is the easiest line on any fact sheet to flatter, because nobody can see what was left out of it; a net one you can regenerate on your own machine is not. The point is not that the numbers are big. It is that you can check them. How to read a track record for the opposite, the un-checkable kind, is in how to spot a fake track record, and the honest version of the money question is in how much a trading bot actually makes.

How to read any hypothetical result

When you see a simulated result, from anyone, run it through the disclaimer instead of around it. Ask three questions. Was it designed with hindsight, and is there any out-of-sample data to counter that? Does the number assume perfect fills, and what happens to it under realistic slippage? And does it quietly assume a human who never flinched? A result that survives those three questions honestly is worth something. One that dodges them is decoration, and you now know exactly why.

Frequently asked questions

What does "hypothetical performance" mean?

It means performance that did not come from a real account with real money at risk, most commonly a backtest run over historical data. The strategy was simulated rather than traded, so no orders were actually executed and no capital was exposed. Regulators require these results to be labeled as hypothetical because they systematically differ from live results, usually by flattering the strategy.

Is a hypothetical performance disclaimer required by law?

For commodity and futures trading in the United States, the display of simulated results is governed by CFTC Regulation 4.41 and NFA Compliance Rule 2-29, which require hypothetical results to be clearly identified and accompanied by a cautionary statement. The rules do not prohibit showing a backtest, only presenting one as though it were an actual trading record. The familiar disclaimer language grew directly out of those requirements.

Why are backtested results usually better than live results?

Three reasons compound. The strategy was designed with the benefit of hindsight, so it fits its own history unusually well. The simulation assumes clean fills and cannot fully model slippage or thin liquidity, which always cost real money. And the backtest never hesitates, while a live trader skips entries and moves stops. Each of these tends to make the live number lower than the simulated one.

Does "past performance does not guarantee future results" apply to real accounts too?

Educational content, not investment advice. Trading futures involves substantial risk of loss and is not suitable for every investor. Any reference to Algomyth results is hypothetical, simulated over historical data, and shown in full on the performance page; past performance does not guarantee future results.