Deriv automated trading

What automation genuinely buys you

The common thread is discipline rather than intelligence. A bot's advantage over a person is not that it knows more; it is that it does not get frustrated, hopeful or tired, and those three states account for most of the money retail traders lose.

• Setups are never missed because you were asleep, at work, or without signal.
• The same rules execute identically every time, with no hesitation and no second-guessing.
• Risk limits are enforced in code, so a daily loss cap actually stops trading instead of being overridden in the moment.
• Losing positions close at the stop — the thing human traders find hardest.

What it cannot fix

Automation cannot make a losing strategy profitable. If the rules have no edge, automating them loses money faster and more consistently, because software never hesitates.

It does not know a news announcement is coming, and it cannot adapt to a market that has fundamentally changed unless it was specifically built to detect and respond to that.

And it does not remove the need for judgement — it moves it earlier. You still decide the strategy, the risk limits and when to stop. The bot just executes what you decided when you were calm.

The costs that decide the outcome

On short timeframes, costs are most of the result, and this is the calculation most people skip.

Every trade pays the spread. A bot placing fifty trades a day pays the spread fifty times a day whether it wins or loses, so trade frequency is a cost decision as much as a strategy one. Multiplier contracts also carry a commission on the notional, and leveraged positions held past the daily rollover accrue a holding fee.

A backtest that does not charge spread and commission is not a backtest, it is a demonstration. Any result you are shown should state what costs were applied, and if it does not, assume none were.

Choosing where to run it

The practical question is whether the bot needs to run while you are not there. If it does, browser-based automation is the wrong tool regardless of how good the strategy is.

• Deriv Bot (DBot) — free, no code, browser-based. Stops when you close the tab.
• Deriv MT5 with an Expert Advisor — deep tooling and a built-in backtester, but the terminal must stay open, so a VPS is a running cost.
• Deriv cTrader with a cBot — C#, modern API, smaller community.
• A server-hosted API bot — runs continuously without your machine, and can express strategies no block builder can.

Testing before real money

There are three stages and the middle one is the one people skip.

• Backtest on real historical data with spread and commission charged.
• Validate on a separate period the strategy was never tuned on. This is the number that predicts live behaviour; the backtest number does not.
• Run on a demo account for weeks — long enough to see a losing run and find out whether you would have interfered.
• Worth adding: compare against random entries on the same data. If the strategy's results sit inside the range that coin-flip entries produce, its timing carries no information whatever the profit figure says.

Keeping control of your account

Connect through Deriv's official sign-in or an API token you generated and can revoke. Never give any bot your Deriv password.

A trading connection should be able to read your account and place trades and nothing else. You can remove an authorised app from your Deriv account yourself, without asking whoever built it, and that is the property that makes running one reasonable.

Common questions