Is there a perfect trading strategy?

There isn't, and that is not an opinion: we took one simple rule, ran it across ten pairs and all twenty-four hours of the day, and out of 240 combinations only nine survive the cost of trading. Strategy decides when you enter; money management, discipline and patience decide whether you are still there to collect.

The search that never ends

Almost every beginner starts the same way: convinced that somewhere out there is the signal, the indicator, or the system that solves everything. So begins a hunt that can last years — switching strategies every week, buying courses, following other people's signals, jumping from one indicator to the next.

The problem isn't that the strategies are bad. It's that none of them works the same way everywhere, and almost nobody checks where theirs stops working.

What if we measured instead of arguing?

We took a deliberately simple rule, the kind that fits in two lines, and applied it to ten currency pairs in each of the twenty-four hours of the day.

The rule is this: at a given hour, if price closed above the close from three hours earlier, buy; if it closed below, sell. The position closes six hours later. No filters, no indicators, and no parameter tuning.

The data is our own: 5.7 years of hourly candles, from January 2021 to September 2026, with Sundays excluded because they are half a session. Every pair-and-hour combination holds between 1,139 and 1,461 trades, and the hours are in Coordinated Universal Time (UTC). When we say "10:00" we mean the candle whose close fires the signal: the position runs from the end of that hour to six hours later.

What came out? That it depends on where and when

Results are median pips per trade, before costs. One warning before you read it: each pair's best and worst hour are picked after the fact from twenty-four candidates, so the gap column overstates by construction. Read it for the spread, not to trust any single cell.

Three things jump out. The first is that every pair has its own best hour: it takes seven different hours to cover the ten pairs, so there is no single good time that works for everyone.

The second comes from USD/CAD, and it is the uncomfortable one. Its best hour is 11:00 and its worst hour is 09:00: two hours apart, and the sign flips. Nobody who had "validated" that rule at nine would have suspected it behaves the opposite way at eleven.

The third is that four of the ten pairs have their worst hour at 21:00, around the New York close — which lands at 21:00 UTC in summer and 22:00 in winter, because the real close is 17:00 New York time. That is the stretch where the market runs out of anyone to move it.

How many of those 240 combinations actually work?

Out of 240, eighty have a positive median, eight land at exactly zero and 152 are negative. The rule loses in 63.3% of contexts before paying for anything.

Then comes the second filter. The spread on a liquid pair runs about one pip, so subtracting it is the least we can honestly do. With that pip taken out, the combinations still in the black are nine out of 240. That is 3.75%.

And that nine is a ceiling, not a floor: five of the nine are yen crosses, where the real spread usually runs above one pip. With an assumption tuned pair by pair, fewer would survive.

This is the answer to the question in the title. The rule is neither good nor bad: its result is decided by context, and the context is 240 cells of which 231 are worthless.

Why does no strategy win every time?

The market isn't a machine that pays out when you crack the formula. It's a game of probability: you win if, across many trades, your winners outweigh your losers.

Even a strategy with a real edge strings together losing trades routinely. And if each of those losses risks too much of the account, the ordinary losing streak — the kind every edge passes through — is enough to blow it up. The strategy decides when you enter; money management decides whether you're still alive for that edge to play out.

What nobody sells you

What separates the trader who lasts from the one who disappears isn't a secret indicator, but three boring habits:

• Fixed risk per trade: deciding in advance how much you can lose on a single position — usually 1-2% of the account — and calculating your position size from there.
• Accepting the loss: cutting quickly when price proves you wrong, instead of waiting for it to "come back".
• Thinking in drawdown: understanding that recovering a loss costs more than it seems — a 50% loss requires a 100% gain just to break even.

What should you do instead?

Stop collecting strategies and pick one with logic you actually understand. Then do to it what we did here: look at it pair by pair and hour by hour, not in aggregate, because an average that lumps all 240 cells together hides exactly what you need to know.

And judge yourself by whether you followed your plan, not by the result of a single trade. One trade tells you nothing; a hundred trades executed with discipline tell you everything. The metrics worth using to evaluate a strategy are not the win rate either.

What this post does not give you

This measurement has limits, and stating them is part of the result:

• We did not measure the real spread. We do not store bid and ask prices, so the pip we subtracted is a stated assumption, not a figure of ours.
• No compounding, no position management, no financing cost. Fixed-hour entry and exit, nothing else.
• This is not a trading recommendation. The rule is a measuring instrument, not a strategy we are proposing.
• The night hours are not measured on the same sample. Dropping Sundays and requiring the trade to fit inside the week removes Fridays from the hours whose exit would land on Saturday, and Mondays from those whose reference would land on Friday. That is why the middle hours hold about 1,455 trades and the 18:00-02:00 stretch about 1,150: 21% fewer, and always the same weekdays.
• Most of those differences cannot be told apart from luck. With medians of a pip either way across roughly 1,300 trades, many cells are noise — and that reinforces the argument rather than weakening it: if you cannot tell your edge apart from chance, you do not have an edge, you have a streak. It is the same hole that overfitting slips through, and the reason a backtest almost always hands you the answer you wanted to hear.

The strategy is the tool, not the job

The useful question isn't "which strategy always wins?" — because you have just seen that none does — but "in which context does mine work, and how do I manage risk when I'm wrong?". The day you stop chasing the magic formula and start protecting your capital is the day trading stops costing you money and starts making sense. And if you want the hard data behind that, it has been measured: between 67% and 97% of intraday traders lose money, and their analysis is not the culprit.