If a rule cannot be written clearly, it cannot be tested honestly.

A trading idea becomes testable when another person can reproduce the same market, timeframe, entry, exit, sizing, and evidence rules without asking what you meant. The work before a backtest makes the result interpretable and the process repeatable.

The short answer: define six decisions

Write the strategy as a specification, not a story. A complete specification answers where the strategy operates, when it evaluates information, what opens a position, what closes it, how much capital it can use, and what evidence would make the result trustworthy or inconclusive.

1. What market are you testing?

Name the instrument or universe. “Large-cap stocks” still leaves questions about membership, survivorship, rebalancing, and liquidity. A specific symbol is simpler, but it still needs a timeframe and a defined data window.

2. What creates an entry?

Replace subjective language with observable conditions. Instead of “buy when momentum improves,” define the indicator, calculation period, comparison, and crossing behavior. State whether the signal is evaluated intrabar or after a bar closes.

3. What closes the position?

An exit can be a reversed signal, stop, target, time limit, or a combination. When several exits exist, define how they interact. A strategy without an exit is not a complete specification.

4. How large is each position?

Record whether size is fixed shares, fixed dollars, a portfolio percentage, or a risk-based calculation. Also define the maximum number of simultaneous positions and whether leverage is allowed. Our guide to separating the signal from the sizing system explains why this decision deserves its own audit.

5. What assumptions make the test possible?

Every simulation relies on assumptions about starting capital, data frequency, available history, fills, fees, slippage, and execution timing. These details belong beside the result, not in a forgotten settings panel. Review the execution-cost assumptions before treating a frictionless result as decision-ready.

6. What would make the result untrustworthy?

Decide in advance what counts as insufficient evidence. Examples include too few trades, inadequate indicator warm-up, one unusual period driving most returns, or a small parameter change reversing the conclusion.

A strategy becomes testable when another person can apply the same rules without asking what you meant.

A reusable strategy-specification template

Fill in every line before running the first test. If a field is not relevant, record why instead of silently omitting it.

1. Market: instrument or universe, eligibility rules, and data source.
2. Clock: timeframe, session, signal evaluation time, and test dates.
3. Entry: observable conditions, indicator definitions, and order timing.
4. Exit: invalidation, stop, target, time limit, and exit priority.
5. Size: allocation rule, maximum positions, exposure cap, and leverage policy.
6. Evidence: warm-up, costs, minimum sample, robustness checks, and rejection criteria.

Then use the eight-point backtest validation checklist to confirm the implementation and evidence match the written specification.

From conversation to configuration

TradeLab is designed to help translate plain-English strategy ideas into explicit rules and adjustable parameters. The goal is not to make every idea profitable. It is to make the idea precise enough to challenge with evidence.

Choose the next test deliberately

Once the rules are frozen, a historical backtest can examine how the specification behaves on past data. A simulated paper-trading stage asks a different question about forward operation. Compare those purposes in our guide to backtesting versus paper trading before deciding what evidence the strategy needs next.