Turning an idea into a rules-based system

Most people who 'trade' don't have a system; they have a collection of moods. They buy because a stock 'looks good', sell because they're nervous, hold because they're hopeful, and afterwards construct a story about why. The single biggest leap from gambling toward something defensible is converting that mush into a rules-based system — a set of conditions specific enough that two different people, given the same chart, would do the same thing. This chapter is about making that leap, and about the one equation that decides whether your system is worth running at all.

The four rules every system needs

A complete trading system, however simple, must answer four questions explicitly. Vagueness in any of them is where discipline leaks out and emotion floods in.

1. 1Entry — under exactly what conditions do you buy (or short)? Not 'when it looks strong', but a precise, checkable rule: a moving-average cross, a breakout above a defined level, a fundamental screen passing certain thresholds.
2. 2Exit (profit) — under what conditions do you take a winning trade off? A target, a trailing stop, a trend-break — but decided before you're in, not improvised while you're up and greedy.
3. 3Exit (loss) — where do you admit you're wrong and get out? This is your stop, and it must exist before entry. A system without a defined loss exit isn't a system; it's a slow accident.
4. 4Sizing & risk — how much do you put on each trade, and what's the most you can lose on it? This is the rule that keeps any single trade from being fatal, and it's the one beginners skip.

Notice that three of the four rules are about getting out and limiting damage, and only one is about getting in. Beginners spend ninety percent of their energy on entries — the 'secret setup' that supposedly prints money — and almost none on exits and sizing. This is exactly backwards. Entries determine which trades you take; exits and sizing determine whether you survive the ones that go wrong. A mediocre entry with disciplined exits and sizing beats a brilliant entry with none.

Mechanical versus discretionary

Systems sit on a spectrum from fully mechanical to fully discretionary. A mechanical system is one a computer could run: every condition is quantified, every decision automatic, no human judgement in the loop. A discretionary system uses rules as a framework but leaves the final call to a trained human who weighs context the rules can't capture. Both can work; both can fail; and most real traders sit somewhere in between.

• Mechanical — fully testable, immune to in-the-moment emotion, consistent. But brittle: it does exactly what it's told even when conditions have shifted in ways the rules never anticipated, and it can be over-optimised to the past.
• Discretionary — adaptive, can incorporate judgement and context a formula misses. But far harder to test, and dangerously easy for 'judgement' to become a polite word for 'I felt like it'.
• The honest middle: mechanical rules for the things that must never be negotiable — especially the loss exit and the position size — and discretion only where you genuinely have an edge the rules can't encode.

Expectancy: the one equation that matters

Whether a system makes money over many trades comes down to a single expression called expectancy — the average rupees you expect to make (or lose) per trade: Expectancy = (Win% × Average Win) − (Loss% × Average Loss). Read it slowly, because it quietly destroys the thing most beginners chase. Your win rate — how often you're right — is only one of four terms. A system can be right 70% of the time and still lose money, if the 30% of losses are each far larger than the 70% of wins. A system can be right just 35% of the time and mint money, if its wins dwarf its losses. Being right often is not the same as making money. Expectancy is what makes money, and it's the product of how often and how much.

An edge must be defined before it can exist

An edge is a specific, statable reason your system has positive expectancy — a reason the math should work in your favour over many trades. And here's the demand the market makes of you: if you cannot articulate your edge in a plain sentence, you almost certainly don't have one. 'I buy stocks that look like they'll go up' is not an edge; it's a hope. 'I buy breakouts above multi-month consolidation in liquid largecaps, cut losses at a defined level, and let winners run, which historically captures a few big trends that pay for many small losses' is at least a candidate edge — falsifiable, testable, statable. Why insist on defining it? Because an undefined edge cannot be tested, improved, or trusted when it's losing — and every real edge loses for stretches. When you've defined your edge precisely, a losing run is information: is the edge decaying, or is this a normal drawdown the edge always had? When your edge is a vague feeling, a losing run is just fear, and fear makes you abandon the strategy at exactly the wrong moment. Definition is what lets you tell a temporary slump from a broken system.

From idea to system, in order

Putting it together: start with a hypothesis about why some pattern should be profitable (your candidate edge), express it as the four explicit rules, decide how mechanical each rule will be (and keep loss-exit and sizing mechanical), and then — before risking a rupee — estimate its expectancy and stress-test it honestly. That last step, testing, is where most systems quietly fall apart, and it's so full of traps that it gets its own chapter next. For now the discipline is this: no system without four written rules, no rule you can't state, no edge you can't name, and no faith in a win rate divorced from the size of wins and losses. A simple system you execute flawlessly will always beat a sophisticated one you abandon under stress — so begin with a single defensible rule set, even something as plain as a trend-following rule with a hard stop and fixed risk per trade, and prove you can follow it through a losing streak without flinching. You can always add nuance later; you can never add discipline retroactively to trades you've already fumbled.