AlgoLens Factory
AlgoLens AnalyticsBuild your own trading strategy.Without knowing any code.
Pick the idea, combine indicators and let the factory test thousands of versions. You get only the ones that survived, ready for MetaTrader 5.
An illustration of the flow. The labels are the product's own.
Why so many trading strategies break in a live account
It is not bad luck. It is three defects almost nobody checks before putting a strategy live.
You bought a black box
You cannot see the rules, you do not know why it passed its test and you have no way to tell when it stopped working.
The backtest was fitted to the past
Move the numbers until the curve looks good and it will, in the past. That is exactly what falls apart first in a live account.
The EA is not what was tested
Spread, slippage and refused orders change the result. Between the test and the account, nobody measures the difference.
AlgoLens Factory was built to close all three.
You pick the pieces. The factory tries every combination.
Not a line of code. You switch on what it may use, and it draws, assembles and tests the rest.
The pieces you switch on
Pick a base-idea palette and the factory switches on that family's setups. Then you adjust them one by one.
of 115 in the catalogue
- Break of the prior high
- False break of the high
- Head and shoulders
- Break of the Asian high
- RSI
- MACD
- Stochastic
- Awesome Oscillator
- Exponential MA (EMA)
- ADX
- Ichimoku Cloud
- Parabolic SAR
- Swing pivot
- Liquidity sweep
- Range (consolidation)
- Inside bars
- ATR
- Bollinger Bands
- Donchian Channel
- SuperTrend
- Bullish engulfing
- Hammer
- Doji
- Shooting star
- OBV
- VWAP
- Chaikin Money Flow
- Force Index
- Ratio against a peer
- Relative strength vs. a peer
You also set how many setups have to agree before a trade opens: at least two, at most four, whatever you decide.
When the strategy may trade
A strategy can be right and still be wrong at the wrong moment. These switches say which market it waits for.
What the rule may look at
It is not stuck on the chart it trades: it can require the bigger charts and fire off another instrument.
Strategies that watch several timeframes
The rule trades the 30-minute chart, but only after the bigger charts agree.
It fires off an instrument it never trades
What happens on the dollar index can be the reason a gold trade opens.
Ten ways to fail your strategy
It is only approved if it clears every stage you switched on. The first one it fails ends it there, and the reason goes in the report.
Sealed OOS2
The last 15% of the history stays locked through the whole search and only opens for the verdict.
Preferred mode, the factory’s default. Each candidate goes through the ten stages, including 400 Monte Carlo simulations.
A perfect backtest can be reading the future
The rule looks excellent: a clean curve, climbing without a pause.
Walk-forward and Monte Carlo do not catch this: both reuse the same backtest, and the future is still inside it. Only the trade-by-trade audit does.
Of 124,000 versions, six made it
Gold, 30-minute chart, two years of history, Preferred mode. The record of one real mining run, not an average.
One approved for every 20,667 mined
Where the 34 rejected ones stopped, with the reason the factory recorded:
- 23Walk-forwardthe most common reason: the thinnest run has 0 trades, below the 1 floor
- 9Monte Carlo · trade orderreshuffle: median Ret/DD (9.53) below 70% of the original (14.99)
- 1Trade scenariosWithout the 4 best trades (5%), −$24.34 is left.
- 1Neighboring timeframesprofitable neighboring timeframes (0) below the 1 floor · H1: -41 (43t) · H4: -37 (14t)
The other 6 stages: none rejected in this run.
In an earlier, harder run (every stage switched on, a year and a half of history), none made it: 115,800 mined, 40 candidates, 0 approved. The factory also says no.
Your next strategy starts from an idea you already have
I want to build my strategy →A bought EA or a factory strategy
The difference is not in the code. It is in what you know about it.
The factory measures the past. Analytics follows the present.
Every approved strategy leaves with its backtest package, the “AlgoLens backtest package”: what it promised in the test. In AlgoLens Analytics that package becomes the ruler: you see, trade by trade, whether the account is doing what the test did.
What every trader asks first
Do I need to know how to code?
No. You choose the pieces and the approval rules on screens. The factory generates the strategy as an .mq5 EA, ready to compile in MetaTrader 5.
Does it run on a Mac? On MetaTrader 4?
No. The factory runs on Windows and exports for MetaTrader 5. There is no MT4 version.
Do I need a broker account?
To download history through MetaTrader 5, yes, and a demo account works. If you already have the history, you can import it from CSV.
Does it come with market data?
No. The factory uses your broker's history, which is what makes the cost of the test the cost of your account. The spread is measured asset by asset.
Is there a parameter optimiser? Does it mine on ticks?
No and no. Mining on ticks would cost 1,092 times more, and a broker keeps about two months of ticks. The factory works from M1, resampled to your timeframe.
Does the EA manage risk on its own?
It executes stop, target, trailing, break-even, time exit, Friday close and risk-based lot size with a cap. It does not police your account: no daily loss limit, no prop-firm rules.
Is the exported EA the same as the tested strategy?
The factory compares the backtest against the exported .mq5, trade by trade. That is how six kinds of divergence were found and fixed, such as an “at least 50%” condition that became “all of them”, and pending orders placed on the wrong side of price, which the broker refuses without saying so.
How long does a mining run take?
It depends on the size of the search and on the machine. The run quoted on this page took about a minute on the machine it ran on, using the processor’s cores. A different search takes a different time.
Will an approved strategy make money?
There is no such guarantee. Approval says it passed the tests you chose, on the history you chose, with your broker's costs. That is all it means.
What is AlgoLens Analytics?
The sister product, which follows your live account. Every approved strategy leaves with its backtest package, what it promised in the test, and Analytics shows, trade by trade, whether the account is delivering the same.
How do I update the Factory?
When a new version is out, the Factory shows it on the Home screen and updates with one click. Your factories, EAs and license stay.
The factory does not promise profit
It tests strategies against what has already happened. No test measures what the market will do tomorrow. An approval is not a forecast of profit and not investment advice.
The strategies you create are yours, and so is the responsibility for how you use them. Trading involves the risk of loss, including the loss of all your capital. Past results, in a test or a live account, do not guarantee future results. AlgoLens does not manage money, does not place orders and does not give investment advice.
Start from the idea you already have
The factory takes care of finding out whether it holds.
Build my strategies →Windows + MetaTrader 5. No code.What comes with AlgoLens Factory
- The factory app, running on your machine
- Trade setups that combine indicators at every level, price action, candles, time of day, session and another asset, with 12 ready base-idea palettes and everything customizable
- Ten robustness stages, with the reason for every rejection
- Strategies exported as .mq5 EAs for MetaTrader 5
- The backtest package AlgoLens Analytics reads