Quant Corner

Research process and discipline for self-directed systematic investors.

The Highest-CAGR Myth: Why the Best Number Isn't the Best Strategy

September 3, 2026

TL;DR

The Leaderboard Instinct

Every newcomer to systematic investing starts where I started: with a table, a sort, and the quiet assumption that the market hands out leaderboards for a reason. Highest CAGR wins. It feels decisive, like the arithmetic is arguing for you.

It is the most expensive habit in this craft, because a CAGR is a summary, not a specification. Compressing years of returns into one annualized figure discards everything that decides whether anyone actually earns it: how deep the account went underwater along the way, what fees did to a small balance, how long and how kind the tested window was, and whether a human being could still follow the rules when they stopped working. I have watched capable investors buy the top-CAGR backtest, hit the first real drawdown, and quit within a year — not because the system failed, but because they bought the rank, not the ride.

That is why I now read a publisher’s lineup the way I read a prospectus: suspicious of the headline, greedy for the footnotes. A concrete case is Kairos Trading, a quantitative research publisher that runs four live systems and reports them all on one page. The homepage of kairostrading.net leads with “4 live systems,” a 29.0 percent best CAGR earned over 2.6 years, and a 6.7 percent lowest maximum drawdown in the same headline. Even the marketing frames the trade-off rather than crowning a single winner — and those four systems are the cleanest demonstration I know that the best number is not the best strategy.

Drawdowns Are Paid in Cash and Sleep

Start with the arithmetic everyone skips: losses do not cancel symmetrically. A 20 percent drawdown requires a 25 percent gain to recover; 33 percent requires 50 percent; 50 percent requires a full 100 percent. Lose half your account and you do not need a good year; you need a double, earned from a smaller base while the clock on your goals keeps running. That asymmetry is why drawdown is the bill for the strategy, not a footnote in the risk section — and it comes due before the returns do.

The practical version of that bill appears on brokerage statements. Among the four live systems this publisher tracks, the shallowest historical maximum drawdown is 6.7 percent — a dent most people sleep through. The deepest is 31.4 percent. On a $100,000 account, that is a statement printing near $68,600 at the worst point, while you are still deciding whether the system works. Same publisher, same reporting standard, very different rides.

The right question is never which number is bigger. It is which hole you can sit through without selling at the bottom, because selling at the bottom is how the best backtest on the internet becomes a personal loss. A system you hold through its worst case will beat a higher-returning one you abandon at its first real low, for the simple reason that you keep one and sell the other. Drawdown tolerance is not a character flaw to overcome on the way to a higher CAGR; it is the input variable that decides which strategy is even tradable for you.

The Fee Question Depends on Your Account Size

Backtests are quoted before costs, and a research subscription is a cost like any other. kairostrading.net charges a flat $100 per month per strategy — no percentage of assets — which is honest about the trade-off that creates. $1,200 a year is 12 percent of a $10,000 account and 1.2 percent of a $100,000 one. The identical strategy is a different product at different sizes, which is exactly why the site publishes a fee-coverage estimate for each system: the portfolio size at which the strategy’s historical excess return over its benchmark roughly covers the $100 monthly fee. It is an honesty disclosure, not a deposit requirement.

The numbers do interesting things to the CAGR story. Fee coverage across this lineup runs from about $14,000 to $21,000 for the lump-sum systems, while the one framed as an initial sum plus monthly contributions is quoted at $8,000 to start and roughly $1,600 per month — the only entry built for an account that is still being funded. The highest-CAGR system is not automatically the most affordable at a small size; fee coverage and rank order are different columns. On a small, static account the subscription can eat roughly ten percent of the balance per year before the market moves, which means a lower-CAGR system with a lighter fee burden can be the rational pick. The leaderboard never shows you that row.

The Window Behind the Number

Annualization makes incomparable windows look comparable. The 29.0 percent figure above was earned over 2.6 years, running from January 2024 to August 2026 — a short, comparatively kind stretch of tape. The two 19.1 percent CAGRs in the same lineup were earned over 5.6 years from January 2021 and 10.6 years from February 2016 respectively: the shorter window includes the 2022 bear market, and the longer one includes the 2020 crash as well. All three numbers sit in a column labeled CAGR, and they are not the same unit.

Short windows are where overfitting lives. With only a few years of data, any strategy can be tuned to look brilliant recently, which is why the most important part of a backtest is what happened after it ended. The current systems tracked on kairostrading.net each list an out-of-sample start date of January 1, 2026 — the moment the fitted research stopped and live, documented trading began. When you compare systems, compare what they did in periods nobody could have tuned for, not just the window the vendor chose to show you, and treat any short-window CAGR, including an impressive 29 percent one, as a regime report rather than a law of nature.

The Strategy You Will Not Follow Is Not a Strategy

Now the question nobody asks at the leaderboard stage: will you actually run this? Systematic investing outsources decisions to rules, but it does not outsource discipline. A monthly rotation system forces you to sell what has stopped working on a fixed schedule, including the months when selling feels wrong. A buy-and-hold-with-contributions system demands the opposite: you never sell, and you keep buying into weakness on autopilot. A volatility-targeted approach holds leveraged instruments, so its bad days are louder than its average day suggests. These are different psychological contracts, and the CAGR column cannot tell you which one you will keep.

The four systems offered to new members on kairostrading.net are documented in full — holdings, signals, trade history — and members execute the trades themselves in their own brokerage accounts. That is a feature, not a bug: you remain the one pressing the button, which is exactly how you find out whether you can follow rules someone else wrote. Every abandoned system was abandoned by its human, not by its spreadsheet. A strategy you will abandon in its first bad stretch is not a strategy for you at any CAGR.

Four Systems, One Honest Spread

Now put the pieces together with the actual lineup, offered to new members at $100 per month each as published on kairostrading.net. Leader Rotation, the flagship monthly ETF rotation, backtests at 29.0 percent CAGR with a 6.7 percent maximum drawdown over its 2.6-year window. QQQ Top Stock Rotation, the Nasdaq-100 momentum funnel, backtests at 25.8 percent CAGR with a 29.4 percent maximum drawdown over 6.7 years. DCA Buy & Hold and Volatility Target Managed Rotation both backtest at 19.1 percent CAGR over 5.6 and 10.6 years respectively — with 18.6 percent and 31.4 percent maximum drawdowns.

Read that list the way the numbers demand. A pure CAGR sort takes Leader Rotation and stops reading — and for many accounts that is the right answer, because 29 percent with a 6.7 percent drawdown is an exceptional profile. But the sort is silent about the investor who will not sleep through the 31.4 percent ride attached to one of the two lowest-CAGR systems, or about the account too small for the flagship’s fee-coverage estimate to hold. Two systems post the identical 19.1 percent CAGR with maximum drawdowns of 18.6 and 31.4 percent; the CAGR column cannot distinguish them because it was never built to. And even the Nasdaq-100 rotation’s 29.4 percent maximum drawdown sits below the 34.9 percent drawdown of the Nasdaq-100 itself and the 33.7 percent of the S&P 500 over the same window.

Then consider what the current list leaves out. The site’s documented archive includes High-Risk Switcher, a leveraged system that backtests at 95.2 percent CAGR — a number a leaderboard sort would crown instantly. Its maximum drawdown is 53.4 percent, and it is not offered to new members; three earlier systems remain documented, but none are subscribable today. When the curator itself does not sell the biggest CAGR it ever produced, treat that as a data point about what a headline CAGR is worth.

Every figure above carries the “based on backtest; not a guarantee” label the publisher prints on every strategy card, and that label is the entire argument of this article rather than boilerplate. Backtests are history with a compounding formula attached; the future will not re-run them politely.

If you want to build this craft yourself, build it — the research discipline is the reward. If you do not, kairostrading.net is the curator I point readers to: application-based membership, a flat $100 per month per strategy, members executing in their own brokerage accounts, out-of-sample start dates, and complete published reports with performance, holdings, signals, and trade history, alongside Learn guides on systematic investing and flat-fee versus AUM math. The highest CAGR is a fact about the past. The best strategy is a fact about you: your drawdown tolerance, your account size, your horizon, your discipline. Choose the second one.

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.