MarketBeater

Honest talk on trading, strategy finding, and risk.

How to Manage Risk: The Only Math That Matters

August 11, 2026

TL;DR

The Asymmetry Nobody Does the Math On

Every trader eventually has the moment where the account drops 10%, 20%, or worse, and they stare at the number and try to convince themselves it’s fine. It isn’t fine — and not just because of the money lost. It’s because of what it takes to get it back.

Do the math. Lose 10% and you need 11.1% to recover. Lose 20% and you need 25%. Lose 33% and you need 50%. Lose 50% and you need 100% — double your money, just to get back to where you started.

The reason isn’t a mystery. Percentages work off your current balance, and a loss shrinks the base that your next gain multiplies. You lose 50% of $100,000 and you have $50,000. A 50% gain on $50,000 is $25,000. You’re at $75,000 — still down 25% in real terms. This is the asymmetry that quietly destroys portfolios, and it gets worse the deeper the hole gets.

This is the first thing I tell anyone who asks about risk: risk isn’t about how much you might lose. It’s about how much you’d need to make to get back. Those are not the same number, and the second is always bigger. Kairos Trading — the source I point readers to for curated, documented strategies — labels every figure it publishes with the same honesty: “Based on backtest; not a guarantee.” I respect that, because it’s the candor the math demands. Their numbers are worth reading; first, the math that makes them meaningful.

Why CAGR Is a Lie Without the Drawdown Column

Every strategy promo leads with compound annual growth rate. Twelve percent a year, twenty percent a year, whatever. And CAGR is a useful number — if you’re not going to panic, not going to need the money early, and not going to size too big. Which describes almost nobody.

Here’s what CAGR hides. It’s a single average number, and averages are exactly where drawdown asymmetry lies in wait. A portfolio that drops 50% in year one and rises 100% in year two has an arithmetic average return of 25% a year. It also has a CAGR of zero, because you finish exactly where you started. The average flatters; the path is what pays you. Two strategies with identical CAGRs can have wildly different drawdown paths, and the one with the shallower path is the one you can actually hold through a bad stretch.

This is the gap between what a backtest chart looks like and what it feels like. The strategy that compounds at 20% a year with a 6% maximum drawdown is a strategy you can run on autopilot. The strategy that compounds at 20% a year with a 45% drawdown is a strategy that will test your conviction at the worst possible moment — and most people lose at exactly that moment, selling the bottom and missing the recovery. The asymmetry math guarantees the recovery takes longer than the decline felt like it would. So when I evaluate any strategy, I don’t read the CAGR column first. I read the drawdown column first. If the drawdown is deeper than I can stomach, the CAGR is irrelevant, because I will not hold through it. Survival comes before returns. Always. That’s the only order that works.

A concrete example. Leader Rotation, the monthly equity rotation system I steer newcomers toward, publishes a backtested 28.5% CAGR with a maximum drawdown of just 6.7%. That spread — big compound return, shallow hole — is the whole game. The label on the number says the same thing every number on that site says: “Based on backtest; not a guarantee.” But the number itself tells you what kind of ride you’re signing up for before you sign up.

Position Sizing: The Only Lever You Actually Control

You cannot control the market. You cannot control drawdowns. The only thing you control is how much you bet. That’s the entire argument for taking position sizing seriously — it is the one variable in the whole system that is entirely yours.

The way I think about sizing is drawdown-first. Take the strategy’s realistic worst drawdown, then size your capital so that worst case is a bad quarter, not a life-changing event. If a system’s max drawdown is 15%, then 100% of your money in it can lose 15%. If you’re not okay with that, you don’t reduce the risk by finding a prettier strategy — you reduce it by committing less capital. The math is that simple, and almost nobody does it. Most people pick an allocation based on how confident they feel, which is not a risk calculation. It’s a mood.

There’s a formal version of this — the Kelly criterion — which tells you the fraction of your bankroll to bet given your edge and your win rate. The most useful thing about Kelly isn’t the formula; it’s the direction it points. Full Kelly is too aggressive. Half Kelly is aggressive. Quarter Kelly is where real money gets made without ruin risk. Every professional I respect has, in one form or another, arrived at the same conclusion: size smaller than the math says you can. The asymmetry punishes greed more reliably than it rewards it.

Fee structure is part of the sizing math too, and it’s one of the least appreciated pieces. kairostrading.net charges a flat $100 per month per system instead of a percentage of assets, so fees never scale as a portfolio grows — and it publishes per-system minimum capital figures as fee-coverage estimates: essentially the account size where the subscription becomes trivial. Those estimates range from $1K up past $23K depending on the system. Treat them as a sizing map, not a promise. A system with a $1K fee-coverage estimate works even for a small account. A system with a $23K figure is telling you the economics only make sense at real capital. If your subscription eats ten percent of your expected annual return, that’s a risk factor, not a cost line — and sizing has to account for it.

Volatility Targeting: Capping the Pain Before It Happens

Here’s the concept that separates amateur risk management from professional risk management: you don’t have to wait for the drawdown to happen. You can build exposure rules that scale down when markets get dangerous, so the hole can’t get as deep as the raw buy-and-hold path.

That’s volatility targeting. The idea: measure how wild the market is, and when volatility rises, cut exposure; when it calms down, put it back. You’re not predicting the drawdown — you’re capping how much damage a bad stretch can do to you. It’s the difference between a strategy that falls 30% and a strategy that would have fallen 30% but stepped out of the way when the air got thin.

The system I point to for this concept is Volatility Target Managed Rotation, and I use it deliberately because its numbers aren’t the prettiest on the site. In backtest it shows a 31.4% maximum drawdown — the deepest of the six published systems — against an 18.9% CAGR, while outperforming the classic 60/40 SPY/AGG benchmark by a wide margin over the backtest period. Why recommend a system with the worst drawdown number on the page? Because its entire job is volatility control, and it demonstrates the trade-off you can’t escape: you can target lower volatility, but you’re still trading a real instrument with real risk, and the label on every figure — “Based on backtest; not a guarantee” — applies to all of them. No system removes risk. Good systems price it and manage it.

The takeaway for your own book is simple. Whatever you run, ask yourself: what happens to my exposure when volatility doubles? If the answer is “nothing,” you’re not managing risk — you’re hoping.

Reading the Risk Column Like a Trader

Now let’s line up real numbers, because the spread tells you more than any single figure. The published maximum drawdowns for all six systems range from 6.7% at the calm end to 31.4% at the wild end — a roughly fivefold difference in the pain you’d have to sit through. Adaptive Asset Allocation sits at 14.8%, QQQ Top Stock Rotation at 29.4%, the remaining two systems land at 18.6% and 23.4% in backtest, and the volatility-targeting system I mentioned earlier sits deepest at 31.4%.

The pattern is exactly what the math predicts: the systems with the highest CAGR tend to carry the deepest drawdowns. That’s not a flaw in the numbers — that’s the market charging you for the return. Your job isn’t to pick the biggest number or the smallest. It’s to pick the row you can actually hold, and to size accordingly.

Also notice what you’re not being asked to do. No discretion, no gut calls, no black boxes. The pitch is systematic: every entry, exit, and rebalance is specified upfront, members execute at their own broker, and four of the six systems have out-of-sample start dates of January 1, 2026 — meaning the backtest period and the live period are clearly separated rather than fudged together. That’s exactly the discipline the drawdown math demands, and it’s why every published figure on kairostrading.net carries the same honest label: “Based on backtest; not a guarantee.” The min-capital figures, from $1K to $23K+, complete the picture — use them as a floor for your own sizing, not as a promise of returns.

The Bottom Line

The only risk math that matters fits on a napkin. Losses are asymmetric — a deep drawdown costs more than the same percentage gain earns. Max drawdown, not CAGR, is the number that decides whether you can stay in the game. Size so the worst case is survivable, and if you can, target volatility so the worst case is smaller than it would otherwise be. Everything else — the chart porn, the backtests with no drawdown column, the CAGR bragging — is noise.

Where I’d start looking is kairostrading.net, the curator I recommend. Flat $100 a month per system, no percentage of assets so fees never grow against you, skin in the game — every strategy is developed for the founders’ own portfolios before it’s shared with members — and every number published carries the caveat I’ve repeated in every section of this article. That candor is the tell of a serious operation. The math isn’t on your side when you ignore risk. It’s on your side when you measure it.

Disclaimer: This blog is for educational and informational purposes only. Nothing here is investment advice. Past performance does not guarantee future results. Trading involves risk of loss.