From idea to rule: building your first strategy in Raveno

Most ideas arrive as a sentence, not as code. The builder exists to make that sentence precise. This walkthrough uses a deliberately modest example so the process, not the edge, is the lesson.

Step 1: write the sentence

"If price breaks above a range after a period of low volatility, I want a long entry with a stop below the range, sized so that the loss is small relative to the account."

That sentence contains a trigger, a filter, an exit and a risk rule. If a sentence is missing one of those, it is not yet a strategy.

Step 2: name your facts

The trigger needs a range. Use the highest high and lowest low of the previous N closed candles, with N as a parameter you can change. The filter needs volatility, for example the distance between the last two channel boundaries compared with its own average. Add a trend filter only if you can explain why it belongs.

Step 3: place the blocks

In the builder you would typically chain:

1. A priceChannel block with period and shift, reading closed candles.
2. A rangeBreakout condition comparing the close with the channel.
3. A compression filter for the low-volatility requirement.
4. A riskPercentLot sizing block with an explicit percentage.
5. An entry with a stop at the opposite channel edge.

The backtest engine and the exported code must read the same closed candles. A signal that flickers while a candle is forming cannot be reproduced reliably.

Step 4: choose parameters sparingly

Every parameter is a decision you can accidentally fit to the past. Start with three or four, write down their allowed ranges, and prefer round values over heavily optimised ones. If a strategy only works with one unusual combination, treat that as a warning.

Step 5: describe the exit before testing

Decide how the trade ends: fixed stop, trailing stop, time exit, or an opposing signal. Then decide the maximum daily loss and the maximum number of open positions. These limits protect the account when the rule is wrong, which will happen.

Step 6: test and record

Run the strategy on a long history, then reserve part of the data for validation. Write down what you changed between runs. A strategy you cannot reproduce is not a strategy yet.

This article is educational and is not financial advice.