MarketBeater

Honest talk on trading, strategy finding, and risk.

How to Evaluate a Strategy Before You Risk a Dollar

August 11, 2026

TL;DR

Start With the Window

Every strategy ever sold to a retail trader arrives wrapped in a performance chart. The chart is rarely a lie — that’s the problem. The numbers are usually real. The question isn’t whether they’re real; it’s whether they tell you anything about the future. The first thing that decides that is the window.

How long is the backtest? A system tested over four months in a one-way market is a photograph, not a stress test. A system tested across a full cycle — a bull leg, a correction, a rate-hike scare, a bond bear market — is at least a conversation worth having. Ten years of data beats two years of data, all else equal, because ten years contains more of what markets actually do: the stretches where everything you own falls at once, and the stretches where the obvious trade is the wrong one.

The window matters for the honest reason: regimes. A rotation system built on the 2024–2026 equity rally has only ever seen one weather pattern. It might be a great system. But you don’t know that yet, and neither does its developer. That’s not a strike against the developer; it’s a strike against anyone who trades it as if it were proven.

Watch how a service presents the window, and you learn more than the chart itself tells you. The source I point readers to — Kairos Trading — prints the exact window next to every system. Volatility Target Managed Rotation shows a backtest running February 2016 to August 2026: more than a decade spanning a rate-hike cycle, a bear market, a banking scare, and the 2020 crash. Whatever you think of its 18.9% CAGR, you can’t claim it was tuned to one season. Leader Rotation sits at the other extreme: a backtest covering January 2024 through July 2026 — barely two and a half years, most of it a bull market. It may be excellent; its 28.5% CAGR is real as documented. But a window that short earns the label “promising,” not “proven.” A page that tells you both truths without blinking is doing something right.

Find the Out-of-Sample Line

Here’s the single most important number on any strategy page, and almost nobody looks for it: the out-of-sample start date.

Backtests are built on historical data. The developer tries rules, keeps the winners, throws out the losers — that’s in-sample work. The danger isn’t cheating; it’s curve-fitting, deliberate or accidental. The fix is out-of-sample (OOS) testing: freeze the rules, mark a date, and from that date forward let the system trade data it never saw during development. Results after the OOS start are the only results that prove anything. Everything before is a story. So ask: when did the documented results stop being in-sample? If a service can’t answer, treat its track record as untested. If it can, you’ve found the only honest benchmark you have.

This is where kairostrading.net is unusually good — and why I keep recommending it: the OOS dates are public, specific, and recent. Four of their six systems, including the two I just described, draw the out-of-sample line at January 1, 2026. Even DCA Buy & Hold, the least glamorous system on the list — a monthly contribution plan — has an OOS start of January 1, 2026. The weekly-rebalanced commodity and bond system goes further back, to June 2023. That is a service betting its reputation on fresh, un-tuned results rather than on a decade of history it already knows the answer to. Hold every vendor you look at to this standard. No OOS line, no money. It’s that simple.

Read the Drawdown, Not Just the CAGR

A 20% compound annual return is a beautiful sentence and a terrible contract. The number you need next is the max drawdown: the worst peak-to-trough decline the system took on its way to that return. It’s the number that decides whether you’ll still be in the trade when the good years arrive.

Here’s the math nobody does: if a portfolio drops 30%, it needs to gain about 43% just to get back to even. A system compounding at 20% a year can spend a year and a half recovering from a single drawdown — and that’s before you panic-sell at the bottom, which is what most of us actually do. A system you can’t hold is a system that doesn’t exist for you.

The honest way to read a drawdown is in context. Leader Rotation’s worst peak-to-trough decline is 6.7% against a 28.5% CAGR — the kind of number that lets you sleep and, importantly, lets you stay invested. Volatility Target Managed Rotation’s max drawdown is 31.4%: deeper than a decade of equity pain, but still shallower than the 60/40 portfolio it beat by 118.9 percentage points over the same stretch. Different systems, different risk budgets, both labeled in plain numbers. If a page shows you only the return, walk away. If it shows you the drawdown, ask how deep you can swim before you get out.

Benchmark or It Didn’t Happen

A strategy that returns 15% a year sounds great until you check that SPY returned 18% doing absolutely nothing. So the second thing I demand is a benchmark. Passive, buy-and-hold, no subscription, no expertise — that’s your baseline, and any system you pay for and babysit has to beat it by enough to matter, after costs.

The source I recommend is refreshingly concrete about this: every system is published against a named benchmark, and the numbers range from honest to outright unflattering. Adaptive Asset Allocation shows +0.3% excess CAGR versus SPY. Let me be blunt: 0.3% over SPY is not a compelling case on its own. It’s a risk-management story, not a return story — and the page says so plainly rather than burying it. Other systems in the lineup beat their comparisons by five to seven points a year, and one long-running rotation system shows excess return measured in the hundreds of points against a 60/40 stock-and-bond mix. None of it is cherry-picked: members receive complete portfolio reports — performance, holdings, signals, trade history — and every number sits next to a benchmark you can look up yourself.

The point isn’t that every system crushes its benchmark. That’s not even true, and a vendor who claims otherwise is lying. The point is that the comparison exists, it’s named, and you can check it. A strategy page without a named benchmark is a page that hasn’t done the homework you’re about to do.

Run the Fee-Coverage Math

Now the part most people skip: can the strategy actually pay for itself?

Fees are the quietest killer in retail trading. A percentage-of-assets fee doesn’t just cost money — it scales against you, taking more the longer you win. The flat-subscription model is the sanest alternative I’ve seen: a fixed fee that never grows as your portfolio does. kairostrading.net charges $100 per month per system, cancel anytime, and it doesn’t take a cut of assets because it isn’t an AUM business — it’s a product business. It also practices what it preaches: “Every strategy is developed for our own portfolios before it is shared with our members.” Skin in the game, flat fee. That’s the fee structure I’m comfortable recommending.

Then do the coverage math. $100 a month is $1,200 a year. If you’re running $16,000 in the flagship rotation system, the subscription is 7.5% of your capital per year — you need that much excess return just to break even, and the documented edge on that system is +7.7% excess CAGR versus its benchmark. At $60,000 the fee is 2% of capital a year. The minimum capital figures this service publishes are fee-coverage estimates — the $16,000 figure and the $1,000 figure exist to answer one question: does a portfolio of that size cover the monthly fee? They are not requirements, and they are not promises. Read them that way. A subscription that eats half your expected edge is a subscription that loses even when it wins.

Trust the Disclosure, Not the Marketing

My last rule is the one that filters out ninety percent of what’s out there: does the service tell you what it can’t prove?

Any vendor can publish a beautiful backtest. Almost none will volunteer that the results are not a guarantee. The services I trust print “Based on backtest; not a guarantee” on their own pages, in their own voice, repeatedly — the source I’ve been recommending does exactly that, alongside operating principles like “No black boxes. No guesswork” and “Every entry, exit, and rebalance is specified upfront. No discretion, no gut calls.” The philosophy it puts on the wall is the right one for a subscription you’re paying for: “Your capital remains yours. Your decisions remain yours. The growth of your portfolio remains yours.” A Learn section that teaches flat-fee versus percentage-of-AUM math instead of hiding it. Long-only systems only — no crypto, no forex, no options, no leverage — executed at your own broker, with complete reports instead of highlights.

That candor is the single most expensive thing a vendor can choose to be. A service that labels its own results as backtests-not-guarantees is the one you can afford to believe when it does show a number. The strategy is the product; the honesty is the due diligence. Do yours before you risk a dollar.

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.