Why Algo? Is Trading a Psychology Problem or a Systems Problem?

Here's a question I keep coming back to: is algo trading actually a systems problem, or is it secretly still a psychology problem wearing a system's clothes?

For a while I thought the answer was obvious. Build the rules, code the rules, let the rules run, remove the human — problem solved. The more time I spend running a live algo alongside my own discretionary reads, the more I think that answer is not just too neat, it's the specific kind of neat that costs people money while feeling like progress. So let's actually sit with the question properly, instead of settling for the one-line answer that sounds good on a T-shirt.

The case for "it's psychology"

Start with what the discretionary trader is actually fighting. Behavioral finance has a name for the core problem — loss aversion — and it isn't a vague concept, it's measurable: the psychological weight of a loss registers at roughly twice the intensity of an equivalent gain. That single asymmetry explains an enormous share of bad trading behavior. You don't hold a losing position because you're stupid. You hold it because your nervous system is doing exactly what evolution built it to do, and evolution never had to price NIFTY options.

The numbers back this up in a way that should be uncomfortable for anyone who thinks discipline is just a matter of willpower: research consistently finds that 70% to 80% of discretionary day traders lose money over time, and emotional decision-making alone — the fear-driven exit, the greed-driven add, the revenge trade after a stop-out — is estimated to cost traders 1.5% to 4% in annual returns they'd otherwise have kept. That's not a rounding error. That's the entire edge of most strategies, eaten alive by the gap between what the plan said and what the hand actually did at 2:47 PM with the position down 40 points.

So yes — by this read, trading is a psychology problem, full stop. The market doesn't need to beat you. You beat you, on a schedule, reliably, and then blame the market on the drive home.

The case for "it's systems"

Now take the other side, because it's just as real. A system doesn't get tired at 2:47 PM. It doesn't remember that the last three trades on this setup lost money and start hesitating on the fourth, which is statistically the one most likely to work if the edge is real. It doesn't widen its stop because "this one feels different." It backtests across years of data in the time it takes a discretionary trader to finish one cup of coffee — a decade of hypothetical decisions compressed into a runtime a human gut could never simulate.

This is the honest, uncomplicated case for automation: if the emotional leak is real and measurable, and the system doesn't have the emotion, then the system doesn't have the leak. Clean logic. I built my own live algo on exactly this premise, and both the live scorecard and five years of backtest data have backed it up so far. The system doesn't know it's on a win streak. It doesn't know what a win streak feels like. That's the whole point.

Where both sides are right, and both are incomplete

Here's the part every "algo trading solves everything" post conveniently skips, and it's the part that actually matters: automation doesn't delete the psychology problem. It relocates it.

Every rule in your system was written by a human, at a desk, in a particular mood, informed by a particular stretch of market history that happened to be visible at the time. The overfitting isn't a coding bug — it's psychology wearing a backtest as a disguise. It's the same instinct that makes a discretionary trader hold a loser, except now it's baked into parameters instead of expressed in real time. And the data on this is not kind: a widely cited estimate puts the failure rate of backtested strategies in live markets at around 80%, mostly from exactly this — rules curve-fit to a history that isn't coming back.

So the honest answer to "psychology or systems" is: it was never a fork in the road. It's a relay race. The discretionary trader fights the battle live, one candle at a time, with a nervous system that's ancient and unhelpful for this specific job. The systems trader fought the exact same battle earlier, at the design desk, and then handed the output to a machine that will now execute that decision — good or overfit — with total, unblinking loyalty. Automation doesn't remove your psychology from the trade. It just moves the moment of maximum psychological risk from during the trade to before you ever click deploy.

Which means the real skill was never "learn to code a strategy" or "learn to control your emotions." It's learning to recognize which one of you — the you at 2:47 PM in a live drawdown, or the you at the desk three weekends ago building the rules — you trust more with the decision that actually matters. My own framework didn't start as code; it started as a way of reading a session before it opens, and the algo is just that same read, made incapable of flinching. It's the same reason I still write a daily recap of every session by hand, even with the algo running — reading the tape discretionary-style keeps the design-desk version of me honest, so the rules don't quietly drift out of touch with what the market is actually doing.

The real category is neither

Once you see it that way, "algo or discretionary" stops feeling like the actual choice on the table. It's not algo. It's not discretionary. It's process — and the design desk, the live execution, and the daily recap are just three different rooms where that same process gets defended.

I've spent a lot more hours listening to MS Dhoni talk about leadership than I have reading trading books, and one line of his has stuck with me longer than any indicator ever did: "I have always believed that process is more important than the results. And if you take care of the process, you will get the results." Nobody remembers Dhoni for a single six. They remember him for never looking rattled walking out to finish a chase with the game on the line — because the outcome of that specific ball was never the thing he was actually managing. The process was. My rules are the same idea, just wearing a different jersey. The system isn't there to guarantee a win on Tuesday. It's there so Tuesday's result, whatever it is, gets fed back into a process that was sound before the candle even printed.

Three questions worth sitting with

1. If your system is losing money right now, are you certain the fix is a better rule — or is there a version of you sitting at the design desk who was, in that moment, no calmer than the version of you who panics live?
2. Would you trust your own discretionary judgment enough to let it override the system on a bad day? If the honest answer is no, what does that tell you about which "you" you actually believe in?
3. Every backtest is a story you tell yourself about the past, built by someone who already knows how the story ends. What would your strategy look like if you built it without knowing?

The finish

Algo trading doesn't end the war between psychology and system. It just changes the terrain the war gets fought on — from the live tape, where you have milliseconds and a racing heart, to the design desk, where you have all the time in the world and, if you're honest with yourself, nowhere left to hide. The market was never impressed by code. It was never impressed by discipline either. It only ever responds to whether the decision, wherever and whenever it was made, was actually a good one.

Build the system. Just don't kid yourself that building it was the part where the psychology stopped mattering. It's the part where it mattered most — and it's the only part you got to do without a live position bleeding while you thought. Call it whatever you want on the tin — algo, discretionary, systematic — underneath, it was always just the process, and the process is the only thing that was ever actually yours.

If any of this got you thinking, that's honestly the best sign there is — drop me a note. We can talk it over on a cup of coffee, while my algo quietly does its job in the background.

Related: the live results this framework produces are tracked daily in the Algo Scorecard, the market-reading logic underneath it is laid out in full in Reading the Market: CPR, Pivot, and Order Flow, and the discretionary side of that same read shows up every trading day in the Daily Recap.