What the strategy builder does
Most trading ideas start as a sentence: "buy when the trend is up and RSI pulls back". That is enough to trade by feel, but not enough to test. Which trend measure? Which RSI length and level? Is the condition checked on the bar's close or as soon as it touches? Where is the stop, and how big is the position?
The builder turns your description into an explicit rule set with every number stated: market and timeframe, entry rules, exit rules, stop-loss, target, position sizing, filters and a backtest plan. Whatever your description leaves open is filled with a common default, marked "(assumed)" and listed as a question for you to settle.
It does not write code and it does not judge whether the idea is profitable. It is the step between an idea and a backtest.
How to use it
- Describe the idea in plain English, up to 1,500 characters. Say what you trade, on which timeframe, what triggers an entry, when you get out and how much you risk, as far as you know.
- Click Build my rules.
- Read Ambiguities to clarify first. Each question shows the default that was assumed; if one is wrong, add the right value to your description and run it again.
- Copy the rule set into your trading plan, or build it as an indicator or a bot so you can backtest it.
You get 3 free builds a month without an account.
What a testable rule looks like
A rule is testable when two people applying it to the same chart would take exactly the same trades:
Vague: buy when the trend is up and RSI is oversold
Testable: long when, at the close of a 1-hour bar,
EMA(50) is above EMA(200)
and RSI(14) crosses back above 30;
enter at the open of the next bar
The answer follows the order in which you would code or test the strategy:
- Market and timeframe: instruments, chart timeframe, sessions, and the data a test needs.
- Entry and exit rules: numbered conditions with exact settings, long and short separately.
- Stop-loss and target: a price level or a distance (pips, points, ATR multiple), never "where it feels right".
- Position sizing: as a formula, size = account × risk % ÷ stop distance. The position size calculator turns it into lots.
- Filters: trend, volatility, session, spread or news conditions.
- What to backtest: period, costs, the metrics to read and the parameter variations that test robustness.
- Ambiguities to clarify: the open questions, each with the assumed default.
Worked example
Try an example loads this description: "Buy EUR/USD when the 50 EMA crosses above the 200 EMA on the 1-hour chart and RSI is above 50. Get out when price closes back below the 50 EMA. I want to risk 1% per trade and stay out around big news."
A typical answer pins down what the sentence implies. The crossover is checked at the close of a 1-hour bar and the trade opens at the next bar's open. RSI uses 14 periods (assumed) and must be above 50 on the crossover bar. The exit is the first 1-hour close below EMA(50). The description gives no stop, so one is assumed, for example below the latest swing low or at a multiple of the ATR, and the answer asks which you prefer. Sizing is 1% of the account divided by the stop distance. "Big news" becomes an objective filter: no new entries in a window around high-impact USD and EUR releases, such as 30 or 60 minutes either side, with the window marked as assumed.
The questions then cover what the text leaves open: must RSI be above 50 on the crossover bar only, or at any time while the averages stay crossed? Are shorts traded with mirrored rules? Is there a time limit on a trade? None of this says whether the idea works. That is the backtest's job.
From rules to a backtest
- Cover different conditions. Test over enough history to include trending and ranging periods, and keep the most recent stretch out of any tuning for an out-of-sample check.
- Include costs. Spread, commission and slippage. On short timeframes they can turn a small average win into a loss.
- Read more than net profit. Number of trades, profit factor, maximum drawdown, expectancy and the longest losing streak. The expectancy calculator and the drawdown calculator put them in context.
- Nudge every parameter. Try EMA 40 and 60 instead of 50, RSI 45 and 55 instead of 50. If the results collapse, the edge was probably fitted to the past. More in backtesting mistakes that inflate results.
What it will not do
- Write code. The rule set is the input for that step: Edgoro's builders turn it into a Pine Script v6 indicator or strategy, or an MT5 Expert Advisor.
- Give advice. It does not say whether an idea is good, predict a win rate or recommend a market, leverage or risk level. Every number it adds is a placeholder for you to decide.
- Know the market today. It works only from your description, not from live prices or news.