What it measures
The Mania Index is the average historical-percentile rank of two long-history valuation metrics for the US stock market: Shiller CAPE (the cyclically adjusted price-to-earnings ratio) and the Buffett Indicator (Wilshire 5000 total market cap divided by US GDP). When both metrics sit near their all-time highs, the index reads near 100. When both sit near their all-time lows, the index reads near 0. The midpoint is the long-run norm.
Why these two metrics
CAPE smooths earnings over a rolling ten-year inflation-adjusted window — it cannot be gamed by a single quarter of unusually good or bad earnings. The Buffett Indicator anchors equity market value to the actual size of the underlying economy. Together they cover the two dimensions where bubbles always appear: earnings divorced from history, and price divorced from economic substance. Eight-metric versions of this exist (Bloomberg publishes one). They add forward P/E, EV/EBITDA, P/B, P/S, Q-Ratio, and trailing P/E — useful, but most have shorter histories. We start with two and add more as the methodology earns it.
How to read the bands
0–25 is the calm band — stocks are cheap relative to their own history. 25–50 is the watchful band — slightly elevated. 50–75 is the elevated band — well into expensive territory. 75–100 is the mania band — where the watch gets hot. The current reading sits in the mania band, near the 2021 peak and the 2000 dot-com peak. Past peaks: 1929 (~95), 2000 (~99), 2021 (~95), 2025 (current).
What it does not say
Valuation is not a timing signal. The 2000 dot-com peak sat in the mania band for almost two years before it broke. The 1929 peak the same. A high reading does not predict a crash next month or next quarter — it tells you the conditions for one are present, and the margin for error has compressed. Cycle-aware stewards read this index against credit conditions, sentiment, and the policy stance. We will publish those reads alongside, and the Custom Index catalog grows from there.
Revision — 24 September 2026
Two changes. First, the Buffett Indicator component had frozen at September 2025 when its source, the Wilshire 5000, left FRED; it is now the Federal Reserve's own measure (corporate equities at market value ÷ GDP), current to the latest quarter and rolled forward monthly. Second, with both components now roughly two standard deviations above their ten-year averages, the index's straight-line 0–100 scale reached its ceiling and read exactly 100 every day, unable to show whether conditions grow more extreme. It now uses the same S-shaped scale as most Palanor indices: 50 is average, and readings approach but never reach 100. Around 89 today means valuations at historic extremes. History was recomputed on the new scale.