Building a strategy: from an idea to written, verifiable rules

So far you've learned the tools: market, costs, orders, charts, indicators, fundamentals, risk, mind. Now we put them together. A strategy isn't a magic indicator and it isn't "I buy when it feels right": it's a set of written rules that say when to enter, where to place the stop, when to exit and how much to risk, so precisely that another person, applying them, would reach the same decision.

What we're looking for: a statistical edge

An edge is a condition that, repeated many times, produces positive expectancy: over the long run you earn more than you lose, costs included. It doesn't mean winning every time: it means having the maths on your side (expectancy lesson). An edge comes from:

• A plausible reason: why should it work? (for example: in strong trends, pullbacks toward the average tend to be resumed; or: many stops are placed above obvious highs.)
• A precise rule that captures it.
• Risk management that lets you survive the losing streaks.
• Verification on historical data and in demo (next lesson).

Be wary of "mysterious" edges with no explanation: they are often random fits to the past.

The main strategy styles

For a beginner the most sensible style is almost always the pullback in a trend: it uses structure, allows short stops and has a naturally favourable risk/reward.

The four components of every strategy

Every complete strategy answers four questions:

1. Context (filter): when do I look for setups? (uptrend on D1, time window, no imminent news).
2. Setup (area): where do I expect a reaction? (support, moving average, retracement).
3. Trigger (entry): what makes me press the button? (a confirmation candle, a breakout).
4. Management (exit and risk): where do I place stop and target, how much do I risk, when do I close or move to breakeven.

How to write unambiguous rules

A vague rule ("I enter when price looks strong") isn't testable. A precise rule ("I enter when an H1 candle closes above the previous high, with a body larger than 1.2 times the ATR") is. Always try to make conditions yes/no.

Bad: "I buy at an important support with a good pattern." Good: "I buy if: (1) D1 is in an uptrend (HH and HL, price above EMA 50); (2) price touches the previously identified support zone; (3) on H1 a bullish candle forms that closes above the previous one; (4) there is no high-impact news in the next 2 hours."

A complete example: pullback to the EMA in a trend

Here's a complete, teaching example. It's not a strategy proven profitable: it's a model to learn how to write and test rules.

See how every item is verifiable. You can try to find it yourself on a real chart:

Parameters: few and simple

Each parameter (average period, RSI threshold, ATR multiplier) is a degree of freedom with which you could fit the strategy to the past. The more you have, the more likely your system is an illusion. Use standard values (EMA 50, RSI 14, ATR 14) and don't optimise them to make the results fit.

How to build it in practice (a 6-step method)

1. Observe the market for weeks and note what repeats (trends, pullbacks, hours).
2. Form a hypothesis in one sentence: "In strong trends on H4, pullbacks to the 50 EMA on H1 tend to resume direction".
3. Write the rules in the four components.
4. Test by hand on historical data (next lesson), recording every case.
5. Analyse expectancy, drawdown, number of trades.
6. Try on demo for at least a few dozen trades, then live with minimal risk.

Typical construction mistakes

• Too many rules: the more conditions you add, the fewer signals you get and the more you fit the system to the past.
• Copying an influencer's strategy without understanding its logic or verifying it.
• Changing it after every loss.
• Ignoring costs (spread, commissions).
• Neglecting risk management and focusing only on entry.
• Looking for the strategy "with no losses".

In short

• A strategy is a set of written, verifiable rules; the goal is positive expectancy.
• To start, the pullback in a trend is the most suitable style: short stops and good R:R.
• Every strategy has context, setup, trigger and management.
• Rules must be yes/no with few standard parameters.
• Proceed hypothesis → rules → test → demo → live.

Practical exercise

1. Write your own strategy in the four components, on one page, with yes/no conditions.
2. Ask someone else to read it: would they reach the same decision on a chart? If not, it's ambiguous.
3. Find 10 historical examples on the chart where the conditions would have been met and note the result in R.