Understand the workflow
Pulsix helps you turn a market idea into rules you can review, test and compare. The objective is to understand why a result appears under the selected conditions, rather than stop at a favourable chart.
Start with one market and one idea. A simple project makes signals easier to inspect and differences between versions easier to explain. Add a filter when you can describe the question it addresses.
| Stage | What you prepare | What you should be able to explain |
|---|---|---|
| Build | Entries, direction and protections | Why a position opens or closes |
| Test | Data, dates and cost assumptions | What the test actually represents |
| Understand | Trades, metrics and comparisons | Where gains and losses come from |
| Export | Selected version and available file | What needs checking on the target platform |
Describe a first idea
Describe an observable behaviour rather than a profitability target. “Buy after price returns above a moving average” is a starting point. The average period, confirmation, exit and risk still need defining.
I want to study mean reversion on EURUSD M15.
First clarify the signal, exits and missing information.
Suggest a simple hypothesis without adding many filters.
Do not create a strategy yet: I want to review the rules first.The instrument and values in examples are illustrative. Adapt them to your research. When a rule remains subjective, ask for a measurable definition before building it.
Build and protect
- Open Studio and create or select the relevant strategy.
- Check its name, market and timeframe.
- Choose a first condition in Flow and configure its values.
- Review whether it expresses the intended buy or sell behaviour.
- Set position size, stop and target, then save protections.
- Review the complete strategy before testing.
A chart drawing helps explain a rule. It does not establish that every entry condition has been satisfied. Check the saved rules as well as the visual example.
Prepare a useful first test
Select data for the strategy’s market. Check the available dates, timeframe and coverage information. Review the capital and cost fields offered in the test settings. After the test finishes, check the dates in its report; changing a chart selection does not by itself prove a new test was run.
Save an identifiable baseline before changing parameters. A useful name might be Mean reversion · baseline · M15, but keep the report’s version and settings as well.
Read before changing
Read net result, drawdown and trade count together. A gain based on very few observations leaves substantial uncertainty. A large trade count also needs inspection if a few exceptional trades explain most of the result.
Inspect a winning trade, a losing trade and an event close to a time boundary or exit. Compare direction, timing, levels and costs with your intention.
Plan the next experiment
Change one hypothesis at a time. Keep the other settings comparable and write down what would make you retain or reject the variant before reading its result.
Question: does a session filter reduce losses without removing too many signals?
Baseline: saved version without the filter.
Variant: same strategy with a defined time window.
Compare: net result, drawdown, trade count and distribution of losses.
Next check: another period and an explanation of the trade-off.Continue with Build in Flow, Read a backtest and Understand metrics. Use the export guide after selecting and reviewing a version.