Ready to go further
Next: size the trade the builder will use
Enter your balance, the percent you accept to lose and the stop in pips to see the lot to type into the builder.
What is included
- What it is, in plain words
- The formula or mechanism, with exact arithmetic
- How it hits grid, martingale and DCA bots
- A check you can run before you switch a bot on
Who it is for
- People who run or plan to run a trading bot
- Anyone reading a backtest or a bot sales page
- Readers who want the numbers, not the pitch
- Not advice, and not a signal service
Key facts
- An EA builder is a tool, often a visual one, that turns rules you choose into an expert advisor file without you writing code. This page is general and names no product.
- A builder only builds. It does not judge. A rule that loses money is built as faithfully as one that does not.
- Builders make it cheap to try many settings. That raises the chance of tuning to old data. In the made-up example below, 1,000 combinations tried at a 5 percent luck chance each makes at least one lucky result almost certain (calculated by us on 29 Oct 2026, exact).
- What you check is the output: does it trade the rules you meant, with what lot, at what cost, and what is its worst case.
- No tool can promise a result. Your own test is the evidence.
What an EA builder does
A normal expert advisor is written in a programming language for the platform. An EA builder replaces the typing with a form, a set of blocks or a plain-language box. You pick an indicator, a condition, a stop and a lot size. The builder joins the pieces and produces a file you can load on a chart.
What it usually offers is a set of ready-made parts:
- A list of indicators and the conditions to compare them, such as one average crossing another.
- Entry and exit choices: stop loss, take profit, a trailing stop, a time of day.
- Sizing options: a fixed lot or a percent of the account.
- Filters for spread, session and day of the week.
- Often a way to run a test and to vary a setting across a range.
Used well, this is a fast way to turn a clear idea into something that can be tested. It also helps someone who can read rules but not code. The saving is real. It is the saving of typing, and only that.
What a builder cannot do
- It cannot tell you whether the idea has an edge. Rules built in a minute are tested the same as rules built in a month. See expectancy for the arithmetic of an edge.
- It cannot see your costs. Spread, commission and swap come from the platform and your account type. A builder test run with none of them is too kind.
- It cannot stop overfitting. It makes trying many settings so easy that it invites it. More on that below.
- It can only build what its blocks allow. If your idea needs a rule that is not on the menu, you either bend the idea to the menu or you need code. Bending the idea to the menu is how people end up testing something other than what they meant.
- It cannot know your account. It does not know your balance, your risk limit or the other trades you hold, unless you set them.
- It cannot protect you from a rule that is dangerous. A doubling lot after a loss is a menu choice like any other, and it is built as asked. The martingale page shows what that choice does.
Worked example: three settings, a thousand combinations
This is made up. Say a builder lets you set three things on a moving average cross EA: the fast period, the slow period and the stop distance. You try 10 values for each. That is 10 x 10 x 10 = 1,000 combinations, and the builder tests them all in one run.
Suppose, for the sake of the arithmetic, that none of these settings has any real edge, and that each combination has a 5 percent chance of showing a good-looking result on one stretch of past data purely by luck. How many look good?
Chance at least one looks good = 1 - 0.95n
| Combinations tried | Expected to look good by luck | Chance at least one does |
|---|---|---|
| 10 | 0.5 | 40.1% |
| 100 | 5.0 | 99.4% |
| 500 | 25.0 | 99.9+% |
| 1,000 | 50.0 | 99.9+% |
With 10 combinations there is a 40.1% chance of a lucky one. With 1,000 there are about 50.0 of them and it is a near certainty. The best-looking result in the list is then the luckiest of a lucky group. It will not repeat, because the luck was in the data, not in the rule. This is overfitting, and a one-click run of a thousand settings makes it quick to do by accident.
The size of the trade matters as well. A builder usually asks for a lot. A fixed 0.10 lot with a 20 pip stop on a pair at about $10 a pip per standard lot loses $20 when stopped (made-up). On a $2,000 account that is 1.0 percent of the account for each stopped trade. Check what the builder put in that box.
How to test what a builder gives you
- Read the rules back. Write the strategy in two or three plain sentences, then compare them with the rules the builder shows. If they differ, the test is of something else.
- Check the first few trades by hand. Open a chart, find the first five or ten trades the test took and check each entry against your rule. Trades that should not exist, or missing trades, show a wrong setting or a wrong time frame.
- Set real costs. Put in the spread and commission of the account you would use, and some slippage.
- Keep data back. Test on part of the history, then once on the part you did not touch, as in forward testing. If you tried many settings, expect the second result to be worse.
- Prefer a broad plateau. If a small change in one setting turns a good result into a bad one, the good result was probably luck.
- Find the worst case. Look at the largest position, the largest loss and the drawdown.
- Run it on a demo or paper account and watch whether it trades as the test did. The install steps cover loading it.
What this page does not tell you
It does not rate or name any builder, and it does not say an EA from a builder can or cannot make money. The numbers are made up. Menus and features differ by product, so check what yours does.
Keep Going
Related Tools and Guides
Keep exploring
FAQ
Frequently Asked Questions
What is an EA builder?
A tool that makes an expert advisor from rules you pick, usually with blocks or forms instead of code. It outputs a file that you load on a chart.
Does an EA builder make the strategy good?
No. It builds the rules you choose. Whether they have an edge after costs is a separate question that only a proper test can begin to answer.
What can a no-code builder not do?
It cannot judge an idea, know your costs, stop overfitting, build a rule that is not on its menu or protect you from a risky choice such as a doubling lot.
Why is testing many settings a problem?
With enough combinations some look good by luck. The best of them is then picked from the luckiest and is unlikely to repeat.
How do I check what a builder made?
Read the rules back, check the first trades by hand, use real costs, keep data back, look at the worst case and then run it on a demo account.
Are the numbers on this page real?
Example numbers are made up to show how the maths works, and the page says so. A figure only counts as exact when the page says it was calculated on a stated date. Nothing you enter in a calculator leaves your browser.
Can I use this to decide whether to run a bot?
It can show whether a bot's numbers hold together and how much an account can lose. It cannot say what the market will do, and it is not advice. Use it to ask better questions, then read the terms of your own broker or exchange.
Is this financial advice?
No. It is general education. The numbers are made up and no product is named or rated.
Know the Risk Before You Run It
Free calculators and guides, added every week. Leave your details and we will email you when new ones go up.
Trading carries a risk of loss