The Power of Compounding: Trading’s Least Obvious Math

Two traders finish the year with the same average monthly return. One ends up with meaningfully more money than the other.

Not because of fees, not because of position sizing luck, not because one of them was better at picking entries. Because of a term in the compounding formula that most trading education never mentions, and that quietly decides more account outcomes than win rate does.

The Formula Nobody Shows Traders

Your compound growth rate — the number that actually determines what your account is worth in five years — is approximately your average return minus half your variance.

Written out: G ≈ μ − ½σ², where μ is your arithmetic average return and σ is your volatility. That second term is the cost of your equity curve being bumpy rather than smooth, and it comes out of your account whether you notice it or not.

Two things about that term deserve more attention than they get.

First, it’s subtracted, always. Volatility never adds. The compounded return is never as high as the arithmetic average, and the gap widens with the size of your swings.

Second, it’s quadratic. This is the part that should change how you trade. Because volatility is squared in that formula, doubling your volatility doesn’t double the cost — it quadruples it.

Which means the relationship between how wild your account swings and how much that costs you is not linear. It’s a curve that gets steep fast, and most traders are operating somewhere on the steep part without knowing the curve exists.

What This Means in Plain Terms

Take a portfolio that returns +60% one year and −40% the next. The arithmetic average is a healthy +10%. The actual result: $100,000 becomes $96,000 — a 4% loss over two years.

Average return of +10%. Real outcome negative. The 14-point gap is the variance tax.

The everyday version of this is the recovery asymmetry every trader knows and most underweight: a 50% loss needs a 100% gain to break even. A 33% loss needs 50%. A 25% loss needs 33%. The reason is simple and brutal — after a loss there’s less capital present to participate in the recovery.

Traders generally know these numbers. What they don’t do is draw the conclusion.

The Conclusion Most Traders Don’t Draw

Here’s my position, stated plainly: for a trader, reducing volatility is mathematically identical to increasing returns. Not similar to. Identical to, in terms of what your account is worth later.

If cutting your monthly standard deviation in half removes three-quarters of your variance tax, that’s the same effect on your final balance as finding additional edge worth the same amount. And for most traders, cutting volatility is dramatically easier than finding new edge — it requires no new strategy, no new market insight, no new information. Just smaller size and fewer marginal trades.

Yet almost the entire trading-education industry is oriented toward the harder path: better setups, better indicators, higher win rate. The easier lever sits untouched because it looks like doing less.

This is also why the trader who grinds out consistent small months frequently ends up ahead of the one posting spectacular ones — a pattern I’ve written about in consistent compounding returns. It isn’t a personality preference. It’s the formula.

Where Leverage Fits — And Why It’s Worse Than It Looks

Leverage multiplies both your returns and your volatility. But since volatility enters the growth formula squared while return enters it linearly, the two don’t scale together.

Double your leverage and you double μ, but you quadruple the σ² term. There’s a point where additional leverage starts subtracting more through variance than it adds through return, and past that point you’re taking more risk for less compound growth — the strictly worse deal in both directions.

This is exactly why leveraged products often underperform expectations over long holding periods, with the drag increasing disproportionately as leverage rises. The instrument isn’t broken. The math is doing what the math does.

The optimal leverage point falls directly out of this formula — that’s what the Kelly criterion is. Worth knowing about it, though: half-Kelly sizing sacrifices roughly a quarter of theoretical growth but dramatically reduces drawdown risk from estimation error. Since your estimate of your own edge is uncertain, and full Kelly assumes it isn’t, sizing below the theoretical optimum is generally the correct practical choice. Almost nobody trades above full Kelly deliberately — they arrive there by drifting up after good months, which is precisely the post-streak overconfidence pattern.

Where I’d Be Honest About the Limits

This concept isn’t universally accepted in the form it’s usually presented. Some critics argue volatility drag is partly a semantic issue rather than a distinct phenomenon — the geometric mean is simply the correct way to average compounded returns, and calling the difference a “drag” implies volatility is actively removing something rather than the arithmetic mean simply being the wrong tool for the question.

I think that critique is technically fair and practically irrelevant. Whether you call it a drag or call it “the arithmetic mean overstates compound growth,” the actionable consequence is identical: the swings cost you, the cost is quadratic in the size of the swings, and reducing them raises your terminal balance. The naming argument doesn’t change what you should do on Monday.

The genuine limitation worth stating is different: this analysis assumes your edge is stable and your returns are roughly independent. If reducing size also reduces your edge — because you’re passing on your best setups rather than your marginal ones — then you’re not buying growth by cutting volatility, you’re just trading less. The gain only materializes if the trades you cut are the low-conviction ones.

The Practical Version

Measure your monthly standard deviation, not just your return. Most traders can tell you their win rate and their monthly P&L and have never once computed the variability of those monthly numbers. That variability is a term in your growth equation, and you’re currently managing it by accident.

Treat a reduction in swing size as a return improvement. When you cut position size and your monthly range narrows, you didn’t sacrifice performance — you converted volatility into compound growth. Framing it as a sacrifice is why most traders won’t do it.

Recognize that your worst months matter more than your best ones. Because of the recovery asymmetry, eliminating a bad month is worth more than adding an equivalent good one. Which reorders your priorities: the effort is better spent on what causes your losing months than on finding more winning setups. Usually the answer is overtrading or size drift, not a strategy flaw.

Stop drifting size upward after good runs. Every unrecorded size increase raises σ, and the cost of that increase is quadratic. This is the specific mechanism by which a good quarter turns into a bad year, and it’s the reason profitable traders blow accounts — they were right about direction and wrong about variance.

Understand that this compounds with the psychological problem. A volatile equity curve doesn’t just cost you mathematically; it degrades decision quality, because loss aversion makes large drawdowns push you into exactly the behavior that makes them worse. The financial cost and the behavioral cost point the same direction, which is unusual and worth exploiting.

The Uncomfortable Trade-Off

Lower volatility means slower visible progress, and slower visible progress is genuinely hard to sit with — particularly during the long flat stretch that any compounding process produces before the curve turns up. I’ve written about that stretch as the lag phase, and it’s where most traders abandon the approach that would have worked.

There’s also a real opportunity cost. If you’re trading a small account, the variance tax on a small balance is small in absolute terms, and there’s a legitimate argument for accepting more volatility early when the downside is a recoverable amount and the upside is reaching a size where the math starts mattering. I wouldn’t call that irrational — I’d call it a deliberate bet with an understood cost, which is completely different from drifting into high variance without knowing the term exists.

What isn’t defensible is running high volatility while believing your average return is what you’ll compound at. That’s not a risk preference. That’s an arithmetic error, and it’s the one most people’s intuition about compounding is built to make. Patience and risk management aren’t temperament advice — they’re what the growth formula looks like when you translate it into behavior.

Questions Worth Asking

If I cut my position size, don’t I just make less money?

Less arithmetic return, yes — but the relevant question is what happens to compound growth, and that depends on which trades you cut. Dropping your lowest-conviction setups reduces volatility more than it reduces edge, which raises the compounded number. Cutting uniformly across all trades reduces both proportionally and doesn’t help.

How much volatility reduction is actually worth pursuing?

There’s no universal figure, but the quadratic relationship means reductions are worth most when your volatility is already high. Going from very volatile to moderately volatile buys far more than going from moderate to smooth, so the first cut is the valuable one.

Does this mean high-volatility strategies can never work?

No — a strategy with a large enough edge can absorb substantial variance drag and still compound well. The point isn’t that volatility disqualifies a strategy; it’s that the edge required to justify high volatility is much larger than most traders assume, because the cost rises with the square.

Should I be using the Kelly criterion to size positions?

Kelly gives the theoretical growth-maximizing size, but it assumes you know your edge precisely, which you don’t. Sizing meaningfully below the Kelly figure is the standard practical adjustment, trading some theoretical growth for protection against your own estimation error.

My account is small. Does any of this matter yet?

The absolute cost is small, so there’s a genuine argument for accepting more variance early. What matters even at small size is knowing the term exists, since the habits you build on a small account — particularly size drift after good runs — carry forward to a balance where the same behavior costs materially more.