The education page

Why fixed R:R fails

Every trading course teaches the same ritual: pick a risk-to-reward ratio, say 1:2 or 1:3, place your target, and let arithmetic make you profitable. The arithmetic is real. It just does not say what the courses claim it says.

Win rate and R:R are one dial, not two

If you risk 1R to make a fixed multiple, the win rate you need just to break even is set by a single formula: 1 / (1 + R).

The table looks like a free lunch: demand more per win, need fewer wins. But the market sets the other side of the dial. A target twice as far away is reached roughly half as often, because price has to travel further without first traveling 1R against you. Raise the target and your hit rate falls in step, which is why expectancy barely moves when traders "improve" their ratio. Win rate and reward are the same number read twice. What a fixed ratio really chooses is not profitability but the shape of your results, with expectancy decided by something the ratio never touches: whether your entries have any edge at all.

The target does not reduce your losses

Ask what controls the size of a loss and the answer is never the target. The stop does that, together with position size. A trader risking 1:1 and a trader risking 1:5 with the same stop lose exactly the same 1R when they are wrong. Yet fixed targets are sold as risk management, and traders reach for a nearer target when they want to "be safer." A nearer target does not make any single loss smaller; it makes wins smaller, so losses loom larger relative to the wins that must repay them. The one thing a fixed target reliably does is decide, in advance, the maximum a winner is allowed to become.

Truncating a fat right tail

Returns are not bell-shaped in the tidy way textbooks draw them. Multi-day price moves follow something close to a lognormal distribution with a right tail fatter than lognormal: most outcomes cluster near zero, declines are bounded by the stop you place, and advances occasionally run far beyond anything the cluster suggests. Plot a breakout system's results in R: a pile of outcomes between -1R and +1R, a thinning middle, and a long right tail (5R, 10R, once in a great while near 20R) carrying a disproportionate share of the total profit. Now draw a vertical line at +2R or +3R, which is what a fixed target does, and delete everything to the right of it. The losses are untouched. The small wins are untouched. The only outcomes removed are the rare ones that pay for everything else. That is the whole objection, and it is mechanical, not philosophical: capping wins in a distribution whose profit lives in its tail is deleting the profit.

The specific case this kills, and the one it does not

This is not a claim that a fixed target can never work anywhere. Mean reversion is a different animal: it earns from a high win rate and small, frequent gains, its winners do not run to 20R because it exits into strength by design, and a fixed target fits its distribution honestly. The trap is specific to tail-dependent systems (breakout, trend following, episodic momentum) whose entire expectancy is one or two enormous winners per hundred trades. Bolting a fixed target onto one does not smooth it into profitability; it gives it the losing streaks of a tail strategy and the capped upside of a scalper, the worst of both shapes. You can watch the amputation on identical entries in the daily report: the fixed columns have a maximum win written into their rules, and it never moves.

Sequence risk: surviving until the tail arrives

The second problem is order, not averages. At a 70% loss rate, ten losses in a row on any given run of ten is unlikely, but across a few hundred trades a streak of about ten is close to certain to appear somewhere. Expectancy does not care about order; accounts do, and prop firm accounts with daily and maximum loss limits care brutally. Risk 2% per trade and that streak leaves you down roughly 18% before the tail trade shows up. Risk 0.25% and you are down about 2.5% and still trading. Same system, same expectancy, one survivor. Tail-dependent trading is sized for the streak, not the average trade, because the 20R trade only pays traders who are still in the game when it arrives.

The variable alternative

The alternative is not recklessness; it is moving the decision from a number chosen before entry to the distribution itself. Set the stop, size for the streak, and then have no take profit at all. Protect the trade by moving the stop (to breakeven once the trade has earned the right, then trailing behind the move) and let the exit be decided by the trade failing to continue, not by it succeeding a predetermined amount. Most trades still end near -1R or with a small win; the stop sees to that. But the one trade in fifty that wants to run is allowed to become the trade the whole system exists to catch. Test it on your own history with the Tail Capture Analyzer: cap your finished wins at 2R and 3R and watch what your total does. For a tail-dependent trader, that table is the entire argument in four rows.