What MQL5 Reliability measures, and what it does not

Every signal on the MQL5 marketplace carries a Reliability percentage. It is the first number many subscribers read, and the easiest to misread. This post explains what the score is, what the signal page shows next to it, and which risks neither of them shows.

What the score is

The MetaTrader 5 help describes it in two sentences:

Reliability evaluates in % the risks of this signal relative to others. The higher the variable, the more reliable the signal.

The help adds that it “considers profits, drawdowns, risks and many other criteria”. The exact formula is not published. Readers on the MQL5 forum report that a high deposit load and very high monthly growth lower the score, and that moderate growth and load raise it.

Two things follow from that description. First, the score is relative: a signal is reliable compared with the other signals on the list, not against a fixed standard of safety. If most signals on the list take the same risk, that risk does not lower anyone’s score. Second, because the formula is not public, nobody outside the marketplace can say which risks it weighs, or how much.

What the signal page shows next to it

A signal page shows a lot: growth, the balance and equity curves, profit, how many weeks the signal has run, subscribers, trades per week, average holding time, maximum drawdown, deposit load, profit factor, best and worst trade (payoff ratio), winning and losing streaks, and slippage per broker.

All of those are useful. But each one is a summary of the whole history, and some of the most important risks live in how the trades relate to each other, which no summary shows.

What neither of them shows

We ran the Monetio engine over 2,321 anonymized live forex signal histories. The table lists five checks the signal page does not show, and how often each one fired across the corpus.

Question	On the signal page	In the 2,321 histories
Did the strategy add positions as price moved against it?	Not shown	80%
Did lots grow after a loss?	Not shown	29%
Was money deposited right after a drawdown?	Not flagged	25%
Did the later record repeat the earlier one?	Not shown	24%
Would a little more spread erase the profit?	Slippage per broker, no cost test	12%

Positions added in a grid80%

Bigger lots after a loss29%

A deposit inside a drawdown25%

Early and late results disagree24%

A small extra cost erases the profit12%

A grid is by far the most common. Most signals add positions as price moves against them, so a score that ranks signals against each other barely sees it: it is the normal case. We explain why grid depth matters in Grid depth: the number that predicts the blow-up, and how deposits make a drawdown look shallow in How deposits hide drawdown.

What to do with your own shortlist

1. Use Reliability, growth and drawdown to make a shortlist, as they are meant to be used.
2. Export each shortlisted signal’s trade history from its page. It is a CSV file.
3. Check four things in the trades themselves: how deep the grid went, whether lots grew after losses, whether deposits arrived inside drawdowns, and how much extra cost the profit survives.

Monetio does step 3 for you when you drop the file, in your browser, and explains each finding with the number behind it.