Concept site: what aiprofitbot.com could become. The domain is for sale. See the main page
The Plain Bot DeskMake an offer
A concept explainer

How trading bots actually work, and how they lose money

A bot is software that follows rules someone wrote. It is not a forecast, and no bot can promise a return. This guide explains the parts, the costs and the failure points.

1What a trading bot is

A trading bot is a program that watches a market and sends buy or sell orders when its rules say so. The rules can be simple, such as a price crossing a line, or they can come from a model trained on past data. Either way a person chose them, and a person is still responsible for what they do.

Calling a bot "AI" describes how its rules were made, not how well they will work. The market does not know or care how the rules were written.

2The loop every bot runs

Strip away the marketing and every bot repeats the same four steps, over and over.

Readthe data Applya rule Sendan order Paythe costs and again, every cycle The last box is the one marketing pages leave out.
Figure 1. The four steps of a bot. Costs are paid on every cycle, whether or not the rule was right.

3Where the costs come from

Every order has a price to the person sending it, and the price is charged win or lose.

What the rule saw What is left after Exchange or broker feeSpread, slippage, software Illustration only: no real figures Costs come out first. A rule must beat them before it gains anything.
Figure 2. The same move, before and after the costs. The widths are drawn for illustration, not measured.

Fees are charged per order or per period by the venue or the software. Spread is the gap between the price to buy and the price to sell. Slippage is the difference between the price the rule expected and the price it got. Software and data have their own subscriptions.

4How bots lose money

Bots lose for ordinary reasons, and none of them needs a villain.

Rules fitted to the past. A rule can be tuned until it would have fit old data, then fail on new data. A clean historical chart is not evidence about the future.

Markets change. A pattern that worked in one period can stop. The bot keeps following it.

Leverage. Borrowed money makes a small error large and can end an account fast.

Mistakes and outages. A bug, a lost connection or a stale price can send the wrong order. A bot repeats a mistake faster than a person can.

Keys and custody. A bot that can trade holds access to the account. Weak key handling, or a bot run by someone else, is a security risk of its own.

None of these depends on the bot being "bad". They are why a bot is a tool with risks, not a source of income.

5Before you trust any claim

  1. Does it promise income, profit or a return? Treat that as a reason to walk away.
  2. Can it explain its method in plain words, including how it can fail?
  3. Are all the costs listed in one place, before you sign up?
  4. Who holds your money and your keys?
  5. Does it ask you to recruit others, or to decide in a hurry?

The CFTC lists guaranteed-return AI trading bot claims among fraud red flags in release 8854-24 (January 2024). This guide is general education, not financial advice.

This explainer is a design. The domain that carries it is for sale.

Ask or make an offer