This article compares two approaches to trading decisions. It is educational, not personalised investment advice, and not a claim that either approach is profitable for any individual. Nothing here is a recommendation to trade in any particular way.
Strip away the jargon and there are two ways to decide what to trade. You can apply a fixed rule to data and do what it says, or you can look at the situation and use judgment. The first is quantitative, or systematic; the second is discretionary. The debate between them is usually framed as a fight, with each camp certain the other is foolish. It isn’t a fight, they win in different places, and the useful question is not which is better but which is better for what you are actually trying to do.
What each one actually is
A discretionary trader takes in a situation , the chart, the news, the feel of the tape, and decides. The edge, when it exists, lives in pattern recognition built from experience: a sense for context that is hard to write down. The same trader can weigh things differently on two similar days, and that flexibility is both the strength and the weakness.
A quantitative trader writes the decision down as a rule and lets it run. The edge lives in the rule’s consistency and in having tested it across history before risking anything. The same inputs always produce the same decision, which removes the in-the-moment flexibility, and with it, the in-the-moment mistakes.
Where discretionary genuinely wins
It would be dishonest to pretend systematic always wins. Discretion has real advantages, and they are worth naming:
- Novel situations. A rule only knows what it was built to know. Faced with something genuinely new, a structural break, an event with no historical analogue, a good discretionary trader can adapt in the moment, where a rule keeps applying yesterday’s logic to a changed world.
- Context a model can’t see. Much of what a skilled trader knows is hard to encode, the difference between two superficially similar setups, the read on whether a move has conviction behind it. A rule sees the numbers; it does not see everything the numbers leave out.
- Small scale, deep focus. Watching a handful of names closely, a discretionary trader can extract things a broad systematic screen would never bother to look for.
Where systematic genuinely wins
The systematic advantages are different in kind, less about any single decision and more about the aggregate:
- Consistency. The rule does not get tired, bored, greedy, or scared. It applies the same logic on the day after a big loss as on the day after a big win. Most discretionary edge is destroyed not by bad analysis but by emotional deviation from it, the system simply can’t deviate.
- Breadth and scale. A rule can examine an entire universe every day and rank it, attention that no human can sustain. The edge per name can be small because it is applied across hundreds of them, the statistical view of the boring middle, where the result comes from the distribution, not from being right on any one call.
- Testability. A written rule can be run over history and stress-tested before it risks money. A discretionary process can’t be backtested honestly, because you can’t reconstruct what you would have decided without hindsight contaminating it.
- Auditability. When a systematic decision is wrong, you can inspect exactly why and fix the rule. When a discretionary call is wrong, the lesson is often “I’ll know it next time”, which is not a lesson that compounds.
The real divide: consistency vs. adaptability
Underneath the comparison is a single trade-off. Systematic buys consistency at the cost of adaptability; discretionary buys adaptability at the cost of consistency. Which one you want depends on where your failures come from. If your problem is that you analyse well but act emotionally, the common case , a system removes the failure mode. If your problem is that rigid rules keep misfiring in conditions they weren’t built for, discretion is the relief. Most retail trading struggles with the first, which is why so much of it would be improved by becoming more systematic, not less.
Why Shishin is systematic
Shishin is fully systematic by deliberate choice, for the reasons above: consistency across hundreds of names and hundreds of days, a process that can be tested before it is trusted, and a track that can be verified in the open rather than asserted. Every signal is a rule applied to data, not a judgment call, the mechanics are in how a stock signal is made, the scoring in composite scoring.
The discipline cuts both ways, and we hold ourselves to it. When the live bot once ran a slightly different configuration than the tested one, the fix was to make the live system match the rules that were validated, not to let judgment override them (running what was tested). And the whole track is paper-traded in public, because a systematic claim is only worth what it can be checked against.
None of which makes discretion wrong. It makes it a different tool, with a different failure mode, suited to a different operator. We chose the systematic side because its weaknesses , rigidity, blindness to the genuinely novel, are ones we can manage with regime-awareness and stress-testing, while its strengths, consistency, scale, testability , are exactly the ones that compound.