Open Source Versus a Black Box

A robot whose code you cannot see is judged only by the vendor's promises and by a report the vendor produced. The problem is not that the vendor is lying but that you have no way to check.

What closed code conceals

Three things that decide an account's fate. Position sizing: whether doubling after a loss is built in. The presence of stops: whether the robot closes losing positions at all or waits for a return. And entry conditions: whether they rely on data available at the moment of the decision. None can be established from a report, because a report with excellent statistics is produced equally well by a healthy system and by a grid without stops.

A report does not substitute for code

A beautiful curve of closed trades is compatible with catastrophic risk: martingale and grid systems produce exactly that until their final day. We demonstrated it in simulation — a grid closes only profitable positions while losing ones hang and never enter the statistics. There is one way to distinguish such a report from a healthy one: find out what the robot does with a losing position.

What to demand when source code is withheld

A minimum set: full trade history including floating drawdown rather than closed results only; the maximum number of simultaneously open positions; and behaviour during a run of consecutive losses. If a vendor will not show those three, the question about source code no longer needs asking.

The other side of open source

Open sources do not guarantee quality — they only make verification possible. Several thousand lines without comments are formally open and practically unverifiable. The value of openness is that it lets you ask specific questions: where is size calculated, where is the stop placed, what happens on a failure.

This material is educational and is not individual investment advice. Trading forex carries the risk of losing capital.