Market Temperature
The thermometer on this page comes from our sister site Still Water: several US market gauges, each compared with its own long-term norm, averaged on a 0–100 scale. It describes the present; it is not a forecast or a trading signal.
The US Market Is at 89° Now
Most indicators are far above their long-term norms; the market is hotter than at most times in history.
Each Indicator Right Now
- Price ÷ 10-yr average94
- Distance from the high79
- Market cap ÷ GDP, vs. 10-yr average91
- Margin debt ÷ 10-yr average89
- Financial stress (inverted)92
Each number is that indicator’s percentile in its own history (0–100); financial stress is inverted: the lower the stress, the higher the temperature.
The Last Ten Years
The Giants at Each Temperature
The 110 quarters since 1999, grouped by the US temperature at quarter end: did the giants as a whole buy more than they sold?
| US temperature at quarter end | Quarters | Runs | Share of giants buying on balance (median) | Median net deployment |
|---|---|---|---|---|
| Very cold (deep pessimism) 0–20 too few to say | 10 | 5 | 53% | +0.1% |
| Cool (pessimism) 20–40 | 18 | 5 | 56% | +1.4% |
| Normal (neutral) 40–60 | 36 | 9 | 50% | +0.1% |
| Warm (optimism) 60–80 | 30 | 7 | 49% | +0.0% |
| Very hot (exuberance) 80–100 now | 16 | 4 | 38% | −0.8% |
| All quarters | 110 | — | 48% | 0.0% |
- “Share of giants buying on balance” = giants whose buying exceeded their selling that quarter, as a share of the giants counted that quarter; “net deployment” = each giant’s estimated net buying ÷ its own previous-quarter portfolio, median. Quarters of money moving between managers are left out.
- A “run”: quarters in the same group less than a year apart, with no market low in between, count as one run — neighbouring quarters share the same rise or fall and are not independent. Groups with fewer than 12 quarters or fewer than 3 runs are too small to say anything. The rule is the same as Still Water’s.
- In 1999–2000 only a dozen or so giants were counted, and the roster was picked looking back from today (survivorship bias).
- These are after-the-fact statistics, not a forecast and not advice; 13F filings appear up to 45 days after the quarter, so no reader could have followed along.
Since 1881: Price Coloured by Temperature
The Gauges
US Stocks: Price ÷ 10-Year Average
Data for 2026-09; average since 1880: 128%; 94th percentile
The further the price runs above its long-term average, the hotter it is. High does not mean it will fall right away (see Chapter V, Section 6).
US Stocks: Distance from the High
Data for 2026-09; median since 1871: -13%; 79th percentile (the closer to the high, the hotter)
A fall of more than 20% from the high is commonly called a bear market.
US Buffett Indicator (Corporate Equity ÷ GDP)
Data for the quarter of 2026-04; 10-year average 177%, now 1.44× that; 91st percentile since 1960
The bigger the stock market is relative to the economy, the hotter. This ratio has trended up over the decades, so it is compared with its own 10-year average rather than its whole history.
US Margin Loans
Data for the quarter of 2026-04; a record high; 10-year average US$0.40 trillion, now 1.87× that; +47% over a year; 89th percentile since 1960
How much is borrowed to buy stocks, compared with its own 10-year average: the higher the multiple, the more people are buying on borrowed money. We don’t divide by total market value, because market value keeps growing and the ratio would always look low. Data: Federal Reserve Z.1.
Financial Stress Index (St. Louis Fed, Monthly Average)
Data for 2026-09; 8th percentile since 1993 (lower stress = hotter); 0 = normal
The St. Louis Fed combines 18 weekly market readings into one number — for example, how wildly stock prices swing and how much more companies must pay than the government to borrow. When investors fear trouble, price swings grow and borrowing gets more expensive for companies, pushing the number above 0; when they are not worried, it falls below 0 (the average since 1993 is 0). It reflects how afraid investors are, not how high interest rates are: rates can be high while investors stay calm, and the number can still be below 0. The lower the stress, the hotter it counts on the thermometer (people fear risk less).
US Federal Funds Rate
Data for 2026-10; the Fed’s estimate of the longer-run normal rate is 3.20% (median projection, 2026-09), so it is +0.68 points away; a year ago 3.86%
The cost of money. Compared with the longer-run rate Fed officials project (often called the neutral rate): 0.5 points or more above counts as tight, 1.5 or more as very tight, and the same distances below as loose and very loose (bands set by this site).
US 10-Year Minus 2-Year Treasury Spread
Data for 2026-10; average since 1976: +0.84%
A negative spread (an inverted curve) has often come before recessions, though the lag varies.
The Long Run
What Followed Each Temperature?
Still Water groups every month since 1881 by its temperature and lists, run by run, the actual returns one and three years later, how many times they lost money, and the crashes along the way. See every group and run on Still Water →
Data and Method
- The thermometer’s data and method come from our sister site Still Water’s market thermometer (checked for updates weekly); this site syncs once a month. This page uses the data of 2026-10-14; the latest US month is 2026-09.
- Sources: Federal Reserve Economic Data (FRED) — federal funds rate and its longer-run projection, Treasury yields, corporate equity market value, GDP, the St. Louis Fed Financial Stress Index and broker-dealer margin loans (Z.1) — and Prof. Robert Shiller’s long-term S&P 500 data (S&P 500 is a trademark of S&P Dow Jones Indices LLC). The temperature and the gauges are Still Water’s calculations; “the giants at each temperature” is this site’s, from 13F filings.
- Data are delayed and may contain errors; this page only states historical data and is not investment advice.