Ready to go further
Next: how wide is the range on your win rate?
Enter the wins and the number of trades from your demo run to see the range the true win rate could sit in.
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
- A demo account is a practice account offered by a broker or exchange. It follows live prices, but your orders are filled by a simulation, not by other traders.
- It proves the bot connects, sizes orders correctly, places stops and recovers from errors. It does not prove the strategy earns.
- In the made-up example below, a 1.00 point edge per trade becomes 0.80 on the demo and 0.10 live. With fewer live fills, 100 signals give 80.0 points on the demo and 7.0 live (calculated by us on 29 Oct 2026, exact).
- Demo fills are usually kinder than real ones: no slippage, no queue, a tight spread and sometimes no fees. How kind varies by platform, so check how yours fills.
- Judge a demo run by the number of trades, not the number of days. At 30 trades a 55 percent win rate has a range of about plus or minus 17.8 points (approximation).
What a demo account is
A demo account gives you a balance of pretend money and a trading screen that looks like the real one. The prices usually come from the live feed. When your bot sends an order, the platform decides what would have happened and updates the balance. No other trader takes the other side. This is the same thing that other platforms call paper trading, and the lessons are the same.
There is one exception worth stating. Some demo servers run on a delayed or reduced price feed, or a different server from the live one. Ask what yours uses before you trust the timing of a fill.
What a demo account proves
- The bot runs. It logs in, reads prices, reaches the signal and sends the order in the right format.
- The sizing is right. The lot size or unit count is what you meant, and the stop and target are on the correct side of the price.
- It survives a restart. A bot left running for days meets dropped connections, server restarts, weekend gaps and daylight saving changes.
- It matches the backtest on the same days. If the demo trades differ from what a backtest says the bot would have done, a bug or a flaw in the test is likely.
What a demo account does not prove
Three things are different when real money is on the line.
Fills
On a demo, a market order usually fills at the price on screen. In a live market the price can move between the click and the fill, and that gap is slippage. In a fast market it is larger. A demo may also fill a stop at its exact level, when a live stop can fill worse.
Queue position
A limit order waits in a line behind other orders at the same price. A demo may fill it the moment the price touches that level. In a live book the price may touch your level and turn around before the orders in front of you are done, and your order stays unfilled. The reverse also happens: the orders that do fill are often the ones where the price kept going against you. A demo cannot show that.
Costs and your own behaviour
Some demo accounts use a tight or fixed spread, no commission and no swap or funding charge. A demo also cannot test what you will do when the loss is real, such as switching a bot off, or changing a setting in the middle of a bad run.
A worked example on made-up numbers
A bot has a gross edge of 1.00 points for each trade that fills. All costs below are made up. The demo account fills everything and the live account fills 70.0 percent of the limit orders, because of the queue. The table shows what is left per trade, and for 100 signals.
| Item | Demo | Live |
|---|---|---|
| Gross edge per filled trade | 1.00 | 1.00 |
| Spread paid | -0.20 | -0.40 |
| Slippage | -0.00 | -0.30 |
| Fees | -0.00 | -0.20 |
| Net per filled trade | 0.80 | 0.10 |
| Orders filled out of 100 signals | 100 | 70 |
| Total over 100 signals | 80.0 | 7.0 |
Demo: 1.00 - 0.20 - 0.00 - 0.00 = 0.80
Live: 1.00 - 0.40 - 0.30 - 0.20 = 0.10
Total = net x fills
Demo: 0.80 x 100 = 80.0
Live: 0.10 x 70 = 7.0
The strategy did not change. The demo shows 80.0 points and the live account 7.0, about 8.7 percent of the demo result. The made-up costs are chosen to make a point, so real accounts will differ. The lesson is that the same 1.00 point edge can survive the demo and nearly vanish live, and a small edge is the most exposed to costs.
How long to run a demo before judging a bot
There is no number of days that is enough. Count trades and market types. A bot that trades twice a week has 8 trades in a month, and the win rate range at 8 trades is very wide. The 95 percent range around a 55 percent win rate is about plus or minus 17.8 points at 30 trades, 9.8 at 100, 5.6 at 300 and 3.1 at 1,000. The win rate calculator shows the range for your own counts.
- Write the expected numbers first. The trade count, the win rate range and the biggest fall you expect. Then compare the demo against them.
- Cover different conditions. A quiet week, a news week and a trend are different tests. One kind of market tells you about one kind of market.
- Use pessimistic costs. Set the demo to charge the full spread and fees of the live account if it can, and take off a slippage allowance by hand if it cannot.
- Count only fills that would have happened. For limit orders, count a fill only when the price traded through the level, not just touched it.
- Stop when it is a demo of a demo. After the bot has run cleanly for a good number of trades, more demo time adds little. The next step is a small live account, as in forward testing.
What it does not tell you
A good demo result does not mean the bot will earn live, and a bad one does not always mean it will lose. It is evidence about the plumbing. For a bot that adds to losing trades, such as a martingale or a grid, a demo also understates the worst case, since the demo does not face a real margin call or a gap through the stops.
Keep Going
Related Tools and Guides
Keep exploring
FAQ
Frequently Asked Questions
What is a demo account?
A practice account with pretend money that follows live prices. Your orders are filled by a simulation, not matched against other traders.
Is a demo account the same as paper trading?
In practice yes. Both use live prices and simulated money. The name changes by platform, and so do the fill rules.
Why is my demo result better than my live result?
Demo fills are usually kinder: no slippage, a tight spread, no queue and sometimes no fees. The same trades cost more live.
How long should I run a demo account before going live?
Until the bot has made enough trades to compare against what you expected, and has been through more than one type of market. Trade count matters more than days.
Can a demo account test a martingale or grid bot?
It tests that the bot runs. It cannot show how a real account behaves in a margin call or a gap, so it understates the worst case.
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 about testing. The numbers are made up and your platform will differ.
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