Base-rate arguments about “where we are in the cycle” smuggle the conclusion into the choice of reference class. One July 2026 week produced all three versions. Merrill’s chart book: the cyclical bull (Oct 2022, +110%, 3.5y) vs the average bull since 1949 (+192%, 5.5y — and >3y-old bulls +234%/7y) → “room to run.” Same book, next frame: the secular bull (2013–?, +1,009%) vs predecessors (+1,159%, +2,353%) → even more room. Same week, Bill Smead: the S&P will lose value over 5–10 years — the valuation reference class (Buffett Indicator at records) rather than the duration one.
All three are legitimate-looking historical comparisons; the class choice does all the work, and each chooser sells the conclusion (Merrill an equity overweight, Smead a deep-value fund). The tell that an argument is shopping rather than reasoning: it never states why this class is the right one, or what the same method said at past turning points — the average bull argument was equally available in December 2021. Cousin of Records in a growing series (choosing the baseline) and Check the stat against its own chart (choosing the window).
The vehicle-vs-manager version (asklivermore, 09/2026, https://x.com/asklivermore/status/2094227266717077717): “every legend is losing to the S&P this year” — Buffett −11 pts, Ackman −13, Icahn −22, Saylor −32 vs SPY +12.6% YTD. The table’s own footnote gives the game away: these are stock prices of listed vehicles — a reinsurer (GLRE), a distressed MLP (IEP), a leveraged bitcoin proxy (MSTR), a closed-end fund at a discount (PSH, NAV −7% vs price −0.8%) — over eight months. It compares wrappers, not managers, and the window is Equity duration‘s cherry-pick. Poster sells buy/sell signals.