Last week a trading-platform newsletter landed in my inbox with a strategy of the week.
The headline: Hypothetical Annual Return (Compounded) +90.2%. Somewhere above it, in smaller type: "All results are hypothetical."
I do not know whether the strategy works. I do know that a monthly percentage table is enough to check the arithmetic, so I checked it.
The table
Two and a bit years of monthly cells, printed right there in the email:
| Year | months (%) |
|---|---|
| 2024 | -3.8, +8.3, -0.5, -7.9, +5.3, +3.9, +4.5 |
| 2025 | +8.5, +0.1, +0.6, +50.7, +7.6, +4.5, +3.3, +10.9, -, +6.2, +12.4, +10.7 |
| 2026 | +14.1, +23.4, -20.9, +20.3, -6.1, -8.9, -3.9, +2.8, +9.3, +18.5, +2.2, +50.1 |
Compound them:
- 2024 (7 months): +9.16%
- 2025 (11 months, September missing): +180.27% — the email prints +180.4%
- 2026 (11 months): +49.9% — the email prints +50.1%
- Full span: roughly +350%, which annualises to ~90.3%
So the +90.2% is not a typo and not a fabrication. The arithmetic holds. That is the part almost nobody checks, and it passed.
What the arithmetic hides
Three questions turn "consistent" into "informative".
1. How much rests on a few months?
The three largest cells are +50.7%, +23.4%, +50.1%. Remove those three months and the annualised figure falls to about 34.8%. The strategy is not a smooth 90% engine; it is a handful of good months carrying a decade of averages.
2. What is the worst peak-to-trough on monthly closes?
-21.8%. And 2026 contains -20.9%, then -6.1%, -8.9%, -3.9% in a row. Anyone sizing this needs to know that sequence exists before they read the annualised number.
3. What exactly is being measured?
The vendor's own disclaimer: "there is no single real-life trading account that looks exactly like the track record shown here." Fees are estimated, not taken from a broker statement. The track record is a model account.
None of that is fraud. It is a hypothetical number, clearly labelled as such, in 8-point type.
The three-question check
For any seller quoting an annualised return:
- Ask for the monthly cells, not the annual figure.
- Compound them yourself. If it does not match, stop and ask why. If it matches, continue.
- Drop the three best months and recompute; compute the worst monthly peak-to-trough.
Two minutes, a calculator, and you learn whether the number is durable or decorative.
I keep the working notes and a small standard-library toolkit for this kind of check in backtest-audit-kit. The write-up for this case is case 15.
If you sell a strategy and your numbers survive these three questions, say so — that is rarer than a 90% headline.
Top comments (3)
The check I'd add to your three lives inside the table itself: the "-" cell. 2025 is 11 printed months plus a missing September, and compounding has no representation for missing - the product silently converts that cell to 1.0, a flat month that never happened. That is the general failure: every aggregate has an identity element (sum turns gaps into 0, product into 1, count into skip), so a missing value never errors, it just becomes the most convenient possible value for the arithmetic. The honest report format is coverage next to the number: "23 of 24 monthly cells." A 90% annualised figure over 96% coverage is a different claim from the same figure over 100%, and the table already contains the information needed to say which one it is.
The second addition is above the table rather than inside it. Even a perfectly honest monthly table arrives pre-selected: this is the strategy the newsletter chose to feature this week, out of every strategy they track. The three-question check measures whether the number is durable; it cannot measure the presentation filter, because the reader never sees the strategies that were not featured. That asymmetry is why your step 2 (compound it yourself) catches arithmetic problems but not selection ones - for those, the only fix is asking how many strategies were eligible for the slot and what their median looked like. A vendor who will answer that question is rarer than one whose cells compound correctly.
Both points land, and the second is the harder one.
On coverage: agreed on printing "23 of 24". Working through monthly tables this week, I would add one more count next to it - whether the page is live at all. One strategy on a public leaderboard had 89 monthly cells, all zero: a complete table, 100% coverage, and a number that means "inactive", not "flat". Coverage plus a live/dead marker catches both.
On selection: this is why the eligible set matters more than the arithmetic check. I screened 37 public strategies this week instead of one, which is the closest a reader gets to a denominator from outside. 19 of 36 carry more than 70% of their log return in the best three months, 14 carry no flag at all, and on the most quotable one the best three months are 84% of the log return. The featured example and the median are different claims, exactly as you put it.
So the order I would put the questions to a vendor: how many were eligible, what did the median look like, and how many monthly cells am I actually looking at.
The all-zero leaderboard row is the cleanest possible demonstration that coverage is necessary but not sufficient - a table can be complete and still say nothing.
Your 19-of-36 number is the part of this exchange I would keep. It converts the concentration check from a single-table audit into a population baseline: once a reader knows the median public strategy carries most of its log return in its best three months, the featured 90% stops being a claim about a strategy and becomes a claim about which strategy got picked. That is the selection effect measured rather than asserted, and measured is the only version a skeptical reader can reuse.
Merged order, then: eligible set, median, coverage, liveness, concentration, drawdown. A vendor table that survives all six is rare enough to deserve its own write-up.