Algorex Trading Solutions

Test an idea before it costs money

In plain words

Many trading ideas sound convincing and come with a few good charts. This free tool lets you try an idea in the MetaTrader Strategy Tester in minutes, and it shows plainly whether the idea holds or only looks good.

Every trading idea comes with a story: breakouts, crosses, RSI turns, the opening range, each with a few good charts. What it rarely comes with is a fair test, and a place to try what makes it better.

I wanted a tool that lets anyone try an idea in minutes and keep improving it, with the same questions we ask of our own systems: does it hold in both halves of the test, without its few best trades, and with higher costs? Does it beat plain chance? And which filter, side, exit or market brings it to life?

Algorex Combo Lab

Combo Lab is built from blocks. Eight entry blocks: breakout, fractal break, EMA cross, RSI reversal, time of day, opening range, previous day's high and low, and a random entry as the baseline. Every signal can be followed or faded, long and short or one side only. Filters in any mix: trend on a higher timeframe, volatility, session hours, Fridays, spread, trades per day. Exits: stop by ATR, points or swing, target in R, break-even, ATR trailing stop, time exit, opposite signal, session end.

A profit factor alone says little. A test can look good because of one lucky half, three outsized trades or costs that are too low. So after every test the lab report asks four questions: does it hold in both halves, without the ten best trades, with higher costs, and how long was the longest losing streak? The weakest of the first three is the lab score, and the report gives one plain verdict.

While your entry block trades, the other seven run alongside it as virtual trades, with the same filters, exits and costs. The scoreboard ranks all eight. One run tells you which block is worth a closer look on your market.

Algorex Combo Lab Video
Watch the film: Combo Lab at work

The random entry is the benchmark that matters. On the Nasdaq it ended at a profit factor of 0.87: that is what trading costs. In the first runs with plain defaults, seven of eight ideas did better than chance. That is the raw material. Which condition makes one of them work is what the lab is for.

The manual walks through installation, the first test in five minutes, six experiments and what our own research found: Combo Lab manual (PDF). A short film shows it at work: watch the video.

Free, for MetaTrader 5 and MetaTrader 4. A laboratory, not a finished system: no setting in it is a recommendation, and the defaults are a starting point.

Links

All notes  ·  The tool on the Tools page