The short answer
A trading engine is a fixed set of rules, run by software, that decides when a trade exists: which asset, which direction, where to enter, where the stop sits and where the target sits.
Human judgement goes in once, when the rules are written. After that, the rules decide. The engine cannot talk itself into a trade, and it cannot talk itself out of one.
An engine, not picks
A pick is someone's call on a given day. It depends on what they read that morning and how their last trade went. You cannot test it, because the same person will not make the same call twice.
An engine makes the same call every time the same conditions show up. That one property changes everything: it can be tested against years of history before anyone risks money on it, and its live results can be checked against what it said it would do.
What an engine does, and what it does not
An engine generates trades. A well-run one publishes every trade it generates: the entry, the stop and the target when the trade opens, and the result when it closes.
What an engine should not do is make your decisions for you without your say. Vault Protocol™ publishes trades; it trades no one's account. You decide whether to act, how much to risk, and on which regulated exchange.
How to judge a trading engine
- ·Published before the outcome: every trade is timestamped when it opens, before the market decides it.
- ·Losses stay up: a record that only shows wins is a highlight reel, not a record.
- ·Honest counting: if the stop is hit before the target, it is a loss, even if price reaches the target later.
- ·A stated method that does not change after bad weeks.
- ·A benchmark you could have bought instead: the engine's results set beside the S&P 500 over the same period, in the same units.
- ·Enough trades to mean something: judge it over dozens of trades, not over a good week.
Vault Protocol, an example
ChartsMeanCash™ built Vault Protocol™ — a systematic trading engine. The goal is to beat the S&P 500. Every trade the engine generates is published — timestamped at open, in full at close, losses included. It trades no one's account.
It runs a systematic multi-asset strategy across large-cap equities and major digital assets. In its backtest, 343 trades over 24 months, it won 63% of the time, with wins 33% larger than losses. Hypothetical backtest.
The live record, every trade since June 16, 2026, is published on the track record and scored against the S&P 500 over the same period.