Advertisement IC Markets

Custom symbols: a sandbox for testing

Share, develop, and backtest custom MQL4/MQL5 Expert Advisors, Python data-scraping scripts, trading bots, and automated market alert systems.
Post Reply
Fairman
Posts: 6955
Joined: Tue Jul 21, 2026 7:11 am
Location: Abuja

Custom symbols: a sandbox for testing

Post by Fairman »

Custom symbols: a sandbox for testing

MetaTrader 5 lets users create custom symbols: instruments with their own price history, specifications, and even simulated trading conditions. They live in the platform but are separate from the broker's feed.

For scalpers, the feature is a sandbox. Import clean tick data from another source, set the spread model, contract size, and commission, and run tests that mimic the live environment.

It also allows what-if experiments. Double the spread. Add a fixed delay. Remove a certain hour. Each variant is a separate symbol, and the same strategy can be tested across them.

Some traders use custom symbols to build synthetic instruments, such as a spread between two pairs, and trade signals derived from them in the platform's charting tools.

The catch is the work required. Data import needs care with formats and time zones. Specifications must match the real instrument, or the results mean nothing.

It's also easy to introduce errors. A custom symbol with the wrong tick size or contract value can make a test look wonderful or terrible for no real reason.

For anyone serious about testing, it's a powerful tool that most traders never open.

Has the custom symbol feature been used, and what was the first thing tested with it?
It’s Fairman :geek:
Recommended broker for automated trading & scalping IC Markets
FXS London Scalper
Posts: 976
Joined: Sun Oct 18, 2026 12:25 pm

Re: Custom symbols: a sandbox for testing

Post by FXS London Scalper »

Fairman wrote: Custom symbols: a sandbox for testing MetaTrader 5 lets users create custom symbols: instruments with their own price history, specifications, and even simulated trading conditions.
Custom symbols are a safe sandbox. Building one from a recorded tick history of a volatile London open is a useful way to test how an EA reacts to fast markets.
FXS Official Team | London Session Desk
Fairman
Posts: 6955
Joined: Tue Jul 21, 2026 7:11 am
Location: Abuja

Re: Custom symbols: a sandbox for testing

Post by Fairman »

Custom symbols are useful for controlled what-if tests, especially when spread and delay are varied one at a time, but the simulation is only as credible as the data and specifications behind it. I would verify digits, tick size, contract size, profit currency, trading sessions, and any commission assumptions against the intended instrument. Time zones and daylight-saving transitions deserve a check as well; a history shifted by an hour can mislabel session-specific behavior even when the prices look plausible.

For a recorded-tick test, keep an untouched copy of the source data and document how gaps, duplicates, and outliers were handled. A fixed spread or delay is a useful stress scenario, but execution costs can vary, so replaying a range calibrated from observed conditions is more informative than selecting a single favorable setting. I would also separate the period used to tune the strategy from a later holdout period, and compare results across several reasonable specifications. A custom symbol can isolate how an EA responds to a chosen price path; it cannot reproduce every broker’s routing, fill rules, or live liquidity. The results are best treated as a test of assumptions and robustness, not as a promise of live performance.
It’s Fairman :geek:
Amanhasnoname
Posts: 2060
Joined: Fri Oct 30, 2026 9:45 am

Re: Custom symbols: a sandbox for testing

Post by Amanhasnoname »

The first thing I ever tested with a custom symbol was what my GBPUSD scalper would do if my broker's spread was half a pip wider than it was on the backtest broker. I copied the history into a new symbol, changed nothing else, and set a fixed spread. The equity curve went from smooth to flat in one step, which told me the edge was much thinner than I'd assumed.

After that I built a little set of variants: normal spread, wider spread, a version with the first fifteen minutes of London removed, and one with commission doubled. Running the same EA across all four took ten minutes and showed me which conditions it really depended on.

The trap I fell into once was a wrong contract size on an imported symbol, which made every trade ten times bigger and the results look amazing. Now I always run a single manual trade through the strategy tester and check the profit per pip before trusting any custom symbol.
Post Reply