"How much does a trading bot make?" is not answerable as asked, and the people who answer it anyway are the ones you should be most careful with. A dollar figure without an account size, a contract count and a time period attached is not information. It is decoration.
So here is the honest version. First why the question is broken, then the full set of numbers behind the Algomyth engine, one of the engine behind this site, including the parts that still come off the top.
Three things have to travel with every profit number or it says nothing.
And a fourth, which is the one that separates a real number from a brochure: is it a backtest or a live account? Everything below is a backtest: a simulation over historical data. No real money was at risk in producing these figures.
An average of $18,814 a year does not mean you receive $18,814 a year. It means the total of the record divided by the years in it came to that. Some stretches were far better, some were losses. The average is a summary, not a paycheck.
The main published report runs day by day from January 2022 to September 2026, on a mid-size account at the Medium sizing level, trading the micro Nasdaq one window a day at 9:30 AM ET.
| Line | How it is calculated | Figure |
|---|---|---|
| Period | Day by day, historical data | January 2022 to September 2026 |
| Account modeled | Mid-size, Medium sizing level | Mid-size |
| Total trades | Strategy Analyzer report | 1,164 |
| Profit factor | Dollars collected / dollars paid out | 1.46 |
| Worst drawdown of the period | Deepest peak-to-trough | $5,273 |
| Costs | $1.90 per contract round turn plus a tick, charged inside every trade | ≈ $2.40 a trade on MNQ |
| Net profit | What is left once that is paid | $87,359 |
| Net per year | Average across the record | $18,814 |
| Net per trade | $87,359 ÷ 1,164 | ≈ $75 |
That last row is the one worth staring at. Seventy-five dollars a trade, after costs. That is the whole edge, repeated eleven hundred times. Nothing about this is a jackpot. In the test it is a small statistical advantage applied with no days off. The cost side of that calculation is broken down in slippage and commissions, and the full report with the live signal log that archives every new signal with its outcome is on the performance page.
Hypothetical performance. Every figure on this page comes from a backtest of Algomyth over historical data, not from a live account. Simulated results are prepared with hindsight, carry no financial risk, and cannot fully reflect real execution, slippage or liquidity. Past performance, real or simulated, does not guarantee future results.
No month in that backtest made exactly the average. Months are lumpy. There were months well above it, months near zero, and losing months. A system whose target sits at three times its risk, and which therefore only needs a 25% hit rate to break even, produces losing streaks as arithmetic, not as malfunction.
The number that describes the experience better than the average is the worst drawdown of the period, $5,273. That is the deepest the equity curve fell from a high before recovering, in simulation, on a mid-size account. Before you look at any monthly average, decide whether you would have kept the system running through that. Most people who fail with a functioning system fail exactly there, and the honest treatment of it is in risk.
The average tells you what the system did. The drawdown tells you what you would have had to sit through to get it.
Across the whole record, the Medium level on a mid-size account averages $18,814 a year net. That is the figure to quote, and it is an average of four and three-quarter years, not a description of any one of them. Some years inside that record ran well above it and some ran well below.
It would be easy to pull the best twelve months out of the curve, lead with those, and quietly drop the rest. We are doing the opposite, because a single year is the less useful number. Twelve months is a small sample from one set of market conditions. Volatile, directional opens are exactly what this system is built for, and they are not permanently available. A good recent year is a fact about the recent past, not a forecast. Build your expectations on a strong year and the first ordinary quarter will feel like a failure when it is just the average showing up.
Here is a cost most published trading numbers leave out entirely, and it is a real one if you trade funded capital.
Funded accounts have hard drawdown rules (usually trailing ones) and those rules close accounts. Not only bad accounts. Even a system that makes money over time will still have accounts taken out, because a drawdown that the strategy recovers from is a drawdown the account rule does not wait for. When an account closes, you buy another evaluation.
How often that happens depends on the firm's rule and on where in the sequence you start, so there is no single number to publish for it and we are not going to invent one. What the report does give you is the size of the hole: a worst drawdown of $5,273 over the period, against an average of $18,814 a year. A drop of that size ends a 50k evaluation outright.
State the consequence plainly, because it is the actual downside of the funded model: evaluation fees are money you pay in advance, in cash, whether or not the account ever pays you back. In a flat six months you pay them and receive nothing. The mechanics are described properly in how funded accounts work.
For reference, typical evaluation costs run about $100 for a 50k, $250 for a 100k and $400 for a 250k.
Two levers, and they both move in both directions.
The sizing level. The system publishes three, and they scale together. On a mid-size account, Medium averages $18,814 a year net with a worst drawdown of $5,273. Low is about half of both, $9,407 and $2,637. High is about double, $37,629 and $10,546. The pain scales exactly as fast as the profit, which is why starting at a level you can sit through is not timidity, it is the only way the numbers you looked at still apply to you.
Number of accounts. Running the same system across several of your own funded accounts multiplies whatever the result turns out to be, up or down, and multiplies the evaluation fees and commissions with it. The mechanics are in copying trades across funded accounts.
Change the account size, the number of contracts and the commission rate, and see what the same backtest looks like on your numbers instead of ours.
The net figure above accounts for commissions and slippage, which are charged inside the run. It does not account for evaluation fees, nor for the rest of the operating costs, which are small but real and all yours:
For this system, on this account size, over this period, in simulation: $87,359 net from January 2022 to September 2026 on a mid-size account at the Medium level, an average of $18,814 a year, about $75 a trade with costs already taken out, a worst drawdown of $5,273 along the way, and a target at three times the risk that needs only a 25% hit rate to break even.
That is a real number with all its qualifiers attached, and it is deliberately less exciting than what you will be shown elsewhere. If somebody quotes you a monthly percentage with no account size, no contract count and no drawdown, they are not being optimistic. They are describing something that does not exist. The wider context for all of it is in what a trading bot actually is.
No. Every figure on this page comes from a backtest: a simulation of the strategy over historical market data using NinjaTrader 8's Strategy Analyzer. Simulated results are prepared with the benefit of hindsight, involve no financial risk, and cannot fully reproduce real execution, slippage or liquidity conditions. They are published because you can reproduce them yourself, not because they predict anything.
No, and you should treat that figure as a description of the past rather than an expectation. It is an average across four and three-quarter simulated years that included losing stretches and flat ones, and any individual year can land far above or far below it. Past performance, real or simulated, does not guarantee future results. Plan around the drawdown figure, not the average.
Because a gross number is the flattering one and the easy one. Leave the commission field at zero in the Strategy Analyzer and a bigger figure prints, without anybody having to lie. Here the toll goes in before the report comes out, at a rate that is written down: $1.90 per contract round turn plus a tick of slippage, about $2.40 a trade on the micro Nasdaq. Set those same rates in your own Strategy Analyzer and the published figure reproduces; put your broker's real rate into the calculator and you get your own version of it.
One is enough to find out whether you can actually live with the system, and starting with one is the sensible move. People run several because a single account caps contract size and because spreading across accounts reduces the chance that one drawdown rule ends everything at once. Each additional account is another evaluation fee paid up front, so scale only after the first one has behaved as expected for a while.
Starting with more contracts than the tested size, and switching the system off during a normal drawdown. The first one multiplies a drawdown you have never experienced; the second ensures you take the losses and miss the recovery. Both are decisions made by the owner, not by the software, and they account for most of the failures we see. Start where the backtest starts, and read getting started before you fund anything.
Everything we write comes back to one idea: verify, then decide. The live terminal exists for exactly that.
Unlock Algomyth Live