What systematic trading is and why rules help
Understand what makes a trading approach systematic, which parts every system needs, and why fixed rules make results measurable.
What you will learn
- The difference between discretionary, systematic and automated trading.
- The seven building blocks every complete trading system needs.
- Where a trading edge can come from, and why rules alone do not create one.
- How many trades you need before a result says something, using the standard error.
Explanation
Backtesting a single chart setup by hand was covered in the course "Technical Analysis in Depth". This course goes further: it treats a strategy as a complete, written system that can be tested on years of data, measured with standard metrics and run with little personal judgement.
Three ways to make trading decisions
- Discretionary trading: the trader decides each trade using experience and judgement. Two decisions in the same situation can differ.
- Systematic trading: every decision follows written rules that give the same answer every time for the same data. A person may still place the orders by hand.
- Automated trading: a computer program applies the rules and sends the orders. Automation is a way to execute a system, not a separate kind of strategy.
This course treats systems as fully rule-based, because only that version can be tested honestly.
The seven building blocks of a system
- Universe: which instruments the system may trade (for example, "the ten largest euro-area equity ETFs").
- Data and timeframe: which prices are used (daily closes, weekly bars) and from which source.
- Signal: the calculation that says "long", "short" or "flat".
- Entry and exit rules: when and at which price an order is placed and closed.
- Position sizing: how much capital or risk each position receives (lesson 9).
- Execution: order type, timing and cost assumptions (lesson 8).
- Schedule: when the rules are checked, for example every trading day after the close or on the last trading day of each month.
If one block is missing, the system is not complete, and a test of it will contain hidden decisions.
Why rules help
- Consistency: the same situation always gets the same decision, so emotions after a loss cannot change the method.
- Measurability: because the rules do not change, past results can be calculated and compared with live results.
- Scalability: one set of rules can be applied to 50 markets as easily as to one.
- Less decision fatigue: the daily work becomes checking signals, not debating each trade.
What rules do not do
Rules do not create an edge (a positive expected result after costs). A precise rule can be precisely bad. Edges usually come from one of three sources: a risk premium (being paid for carrying a risk others avoid), behavioural patterns (for example, investors reacting slowly to news), or structural effects (for example, forced selling by index funds on fixed dates). If you cannot name a plausible source, treat a good backtest with extra suspicion.
Measuring with the standard error
Expectancy in R (average result per trade in units of planned risk) was introduced in the course "Risk & Trade Management". A system makes measurement possible, but measurement needs enough trades. The standard error (SE) of the average result is:
SE = standard deviation of trade results ÷ √(number of trades)
The standard deviation measures how widely single results spread around their average. As a rough rule, the true expectancy probably lies within the average ± 2 × SE. If that range includes zero, you cannot yet tell the system apart from luck.
Worked example
Hypothetical example: a system shows an average of +0.2R per trade with a standard deviation of 1.5R. It produces about 4 trades per month (48 per year). Costs, slippage and taxes are excluded.
| Trades (n) | √n | SE = 1.5R ÷ √n | Rough range (±2 SE) |
|---|---|---|---|
| 30 | 5.48 | 0.27R | −0.35R to +0.75R |
| 100 | 10 | 0.15R | −0.10R to +0.50R |
| 400 | 20 | 0.075R | +0.05R to +0.35R |
Only at about 400 trades does the range stay above zero. At 48 trades per year, that is 400 ÷ 48 ≈ 8.3 years of trades. A discretionary trader cannot produce such a record from the past at all, because past decisions cannot be repeated. A systematic trader can calculate it from historical data in one afternoon, if the data are good (lesson 3).
Common mistake
Practice activity
Hypothetical data: a system has 64 trades, an average of +0.3R and a standard deviation of 2.0R. On paper, (a) calculate the SE, (b) give the rough ±2 SE range, and (c) estimate how many trades are needed for the range to stay above zero, assuming the average and standard deviation stay the same.
Show solution
(a) √64 = 8, so SE = 2.0R ÷ 8 = 0.25R.
(b) 0.3R ± 2 × 0.25R = −0.2R to +0.8R. Zero is inside the range.
(c) You need 2 × 2.0R ÷ √n < 0.3R, so √n > 13.33 and n > 177.8. About 178 trades are needed.
Limitations: the ±2 SE rule assumes trades are independent and results are roughly normally distributed; real trades often have fat tails and streaks, so treat it as a minimum.
Key takeaways
- A systematic approach gives the same decision for the same data; automation is only one way to execute it.
- A complete system defines universe, data, signal, entry and exit, sizing, execution and schedule.
- Rules make results measurable, but they do not create an edge; name the source of the edge.
- Use SE = standard deviation ÷ √n to judge whether an average result is more than luck.
Next step
In the next lesson you will turn a vague trading idea into precise rules that a computer or spreadsheet can test without guessing.
Self-check
Answer the questions, then check your answers. You can retry as often as you like.
Automatic self-check for your own revision. Your answers are checked in your browser and are not seen or assessed by us. It is not an exam and not a qualification.
Your progress
“Complete” only means you marked the lesson as done. It is not a test of your understanding. Progress and notes are saved only in this browser on this device — not in an account.