Traders about to automate
Check a rule set on years of data before a bot trades it with real money.
Backtest a trading strategy on years of real crypto data
Backtest Lab takes a strategy from your Edgoro library (a Pine Script strategy, an MT5 Expert Advisor or a Python bot) or a plain-English description, and simulates it on up to three years of real crypto candles. You get an equity curve, the metrics that matter, every trade, an out-of-sample check on the last quarter of the period, and a plain-English verdict on whether the result looks robust or fragile.
The numbers come from a deterministic engine, not from the AI: the same candles, rules and settings always give the same result. The AI translates your strategy into rules beforehand, and suggests what to test next afterwards.
- 20 USDT pairs, 15m to daily
- Fees, slippage, no lookahead
- Add-on: $9/month
How it works
Pick a strategy
Choose a creation from your library or describe the rules. The AI turns it into a strict rule set and lists any assumption it made. If part of it can't be simulated faithfully, the run is refused rather than approximated.
Choose the market and costs
One of 20 USDT spot pairs; 15-minute, 1-hour, 4-hour or daily candles; 6 months to 3 years; long, short or both. Fees default to 0.1% per side and slippage to 0.02%, and you can match them to your exchange.
Run it
Candles load from exchange market data, then the engine simulates every trade. Signals use closed candles only, and orders fill at the next candle's open.
Read the verdict
Robust, fragile, over-fitted or not enough trades, with the reasons in plain English, AI advice on what to change, and one click to send a change back to your strategy.
Example: the rules Edgoro tests
Before anything runs, the strategy becomes a rule set, and the report shows it exactly as tested so you can check nothing was lost in translation. For the request below, the "Rules tested" section reads like this.
RequestTest on BTC/USDT 4h over 2 years: long when the 20 EMA crosses above the 50 EMA and RSI(14) is below 70, only above the daily 200 EMA. Stop at 2 × ATR(14), target 2R, risk 1% per trade.
| Rule | As tested |
|---|---|
| Indicators | fast = EMA(20) · slow = EMA(50) · rsi = RSI(14) |
| Long entry | fast crosses above slow and rsi below 70 |
| Stop | 2 × ATR(14) from entry |
| Take profit | 2R (2 × the stop distance) |
| Trend filter | Longs above, shorts below the EMA(200) on 1d |
| Position size | Risk 1% of equity per trade (never more than 1× equity) |
Around it, the report holds the verdict, the metrics, the equity curve, the out-of-sample comparison, monthly returns and every trade.
What the report shows
- A verdict (robust, fragile, over-fitted or not enough trades) with the reasons behind it.
- Metrics after costs: net return, CAGR, profit factor, win rate, maximum drawdown, expectancy, average R, Sharpe ratio, longest losing streak, exposure and fees paid, next to simply buying and holding.
- In-sample vs. out-of-sample: the first 75% of the period against the last 25% held back, where a strategy fitted to the past usually gives itself away.
- The equity curve with its drawdowns, and monthly returns.
- Every trade, with a CSV export.
- AI advice: a few honest observations and two or three changes worth testing, based on the report's numbers only.
Honest by design
The engine is built to make results harder to flatter:
- Signals are evaluated on closed candles and filled at the next candle's open: no lookahead, no repainting.
- One position at a time. Fees are charged on both sides, and slippage on market fills.
- When a stop and a target could both have been hit inside one candle, the stop is assumed first.
- Under 30 trades, the verdict says the result is mostly luck instead of praising it.
The optimizer follows the same spirit: up to 50 combinations of one to three parameters, ranked by out-of-sample profit factor, shown next to the combination you'd have picked from the first 75% alone. 8 backtesting mistakes that inflate results explains why each rule matters.
What's included, and what isn't
Included
- Backtests on 20 USDT spot pairs: 15m, 1h, 4h and 1d
- Up to 3 years of candles, with fees and slippage you can set
- Out-of-sample check and a plain-English verdict
- Optimizer with walk-forward validation, up to 50 combinations
- AI advice you can send back to your strategy in one click
- Trades CSV export and a history of your runs
- 100 backtests and 10 optimizer runs a month (fair use)
Limits, plainly
- Crypto only for now: forex, stocks and indices can't be tested in the Lab.
- Spot USDT pairs, one position at a time, never more than 1× equity.
- Indicators are limited to EMA, SMA, RSI, ATR, MACD, Bollinger Bands, Stochastic, VWAP and highest/lowest. Anything the rules can't express is refused, not approximated.
- A backtest describes the past. Robust means consistent on this data, not guaranteed.
- Backtest Lab is an add-on and requires a plan.
Who it's for
Strategy tinkerers
Change one parameter, run it again, and see whether the edge survives the out-of-sample check or only fits the past.
Traders burned by pretty backtests
Fees, slippage, next-candle fills and a held-back period make it harder for an equity curve to flatter you.
Compared with doing it yourself or a generic AI chatbot
| Edgoro | Doing it yourself | A generic AI chatbot | |
|---|---|---|---|
| Getting the rules in | From your library or plain English | Recode the strategy in a backtester | Writes backtest code you then run |
| Data | Up to 3 years of exchange candles, loaded for you | Download and clean it yourself | You supply it |
| Costs and fills | Fees, slippage, next-candle open, stop-first rule | As careful as your code | Depends on the code it writes |
| Overfitting check | Out-of-sample split and a verdict | If you build one | Only if you ask |
| Markets | Crypto only | Any market you have data for | Any market you have data for |
Pricing
Backtest Lab is an add-on at $9 a month, on top of a plan (Pro Monthly at $29 a month or Lifetime at $49 once). Add it when you choose your plan or later from Settings, and stop it anytime: access runs to the end of the paid month. Strategies come from the Pine Script generator in every plan, or from MT5 and crypto bots with the bots option.
Frequently asked questions
Which markets can I backtest?
Crypto only for now: 20 major USDT spot pairs such as BTC/USDT, ETH/USDT and SOL/USDT, on 15-minute, 1-hour, 4-hour or daily candles, over 6 months to 3 years.
Can I backtest a Pine Script or MT5 strategy?
Yes, from your Edgoro library or from a description. The AI translates it into rules shown in the report. The test still runs on crypto data, and anything the rules can't express stops the run with an explanation.
Where does the data come from?
Spot candles from public exchange market data, with fallbacks between providers when one is unavailable. Each report lists the providers it used.
Does a robust verdict mean the strategy will make money?
No. Robust means the strategy was consistent on this data after costs and held up on the period it wasn't judged on. Markets change, and a backtest describes the past.
How many backtests can I run?
Up to 100 backtests and 10 optimizer runs per calendar month. Failed runs don't count.
Do I need a plan?
Yes. Backtest Lab is an add-on at $9 a month on top of Pro Monthly or Lifetime.
Related guides and free tools
Guide
8 backtesting mistakes that make a strategy look better than it is
Look-ahead bias, repainting, overfitting, missing costs and unrealistic fills: how each inflates a backtest and how to spot and fix it in TradingView and MT5.
GuidePine Script EMA crossover strategy: a complete v6 example with stops and costs
Build an EMA crossover strategy in Pine Script v6 with a trend filter, ATR stop and target, commission and slippage, then test it without fooling yourself.
GuidePine Script repainting: how to spot it, test for it and fix it
Why some TradingView indicators repaint, the usual causes, how to test any script on replay and live charts, and the Pine Script v6 fix for each one.
Free indicatorEMA crossover, with Pine Script v6 code
How the EMA is calculated, why fast/slow EMA crosses are popular, which lengths traders use, and a free EMA crossover script with a trend filter.
Free toolExpectancy & Win-Rate Calculator
Turn win rate and average win/loss into expectancy per trade and a 100-trade projection.
Free toolDrawdown Recovery Calculator
See the gain needed to recover from a drawdown and how a losing streak hits your balance.
Test it before you trust it
Answer a few questions about how you trade, choose a plan, add Backtest Lab, and run your first test.
Instant access after checkout. 7-day money-back guarantee on your first purchase.
Trading involves substantial risk of loss and is not suitable for every investor. Edgoro is a software tool, not a broker or financial adviser: nothing it produces is investment advice, and past or backtested performance does not guarantee future results.