Sentiment and Positioning Indicators

What “sentiment” measures, and from where

Sentiment indicators try to measure the same quantity — how crowded one side of the market is — from different places. The logic behind reading them is simple and sound: a market where nearly everyone is already bullish has fewer buyers left to push it higher, and a market where nearly everyone is bearish has fewer sellers left. What is not simple is turning that into timing. They come in four families. **Surveys** ask people what they think: the AAII Sentiment Survey asks individual investors weekly whether they are bullish, neutral or bearish. **Options-market measures** read what people pay for: the put/call ratio compares put volume with call volume, and the VIX is the implied volatility in S&P 500 options. **Positioning** reads what people hold: the CFTC’s weekly Commitments of Traders report breaks down futures positions by type of trader, and short interest counts shares sold short. **Flows and leverage** read what people do with money: fund flows and margin debt. The families disagree often, and that is useful. A survey says what people feel; positioning says what they have done about it. When a survey is bearish but positioning is still long, people are worried and have not acted — a different situation from one where they have already sold. • Surveys — what investors say (AAII weekly survey). • Options market — what protection costs (put/call ratio, VIX). • Positioning — what investors hold (Commitments of Traders, short interest). • Flows and leverage — what investors do with money (fund flows, margin debt).

The put/call ratio and the VIX: fear you can price

The **put/call ratio** divides the volume of puts traded by the volume of calls. Cboe publishes it for equities and for indexes separately, and the difference matters. Index puts are dominated by institutions hedging portfolios every day, regardless of mood, so the index ratio runs high as a matter of course. The equity-only ratio is closer to speculative behaviour and is the one contrarians usually watch: unusually high readings mean many traders are buying protection or betting on falls; unusually low readings mean many are chasing calls. The **VIX** is the market’s expectation of S&P 500 volatility over the next 30 days, computed by Cboe from a strip of index option prices. It rises in sell-offs because demand for put protection rises, and it tends to mean-revert: extreme readings rarely last. That is the source of the contrarian reading — a very high VIX has historically been followed by above-average equity returns on average over the following year. But the average hides the path. In the autumn of 2008 the VIX stayed above 40 for months while stocks kept falling, and its record close came in March 2020, a week before that crash’s low. The honest summary: both measures describe the price of fear right now. That price is a useful input to sizing and to option strategy — protection is dear when everyone wants it — and a weak input to timing. Two VIX extremes, read afterwards — October 24, 2008: VIX trades as high as 89.53 intraday — its record · What followed: Stocks fell roughly another fifth into March 2009 · November 20, 2008: VIX closes at 80.86, the crisis’s highest close · March 2009: The low comes with the VIX well below its peak ←

Positioning: what people hold, not what they say

The CFTC’s **Commitments of Traders** report, published each Friday for positions held the previous Tuesday, splits open interest in U.S. futures by type of trader. In the legacy format the split is commercials (hedgers with a business exposure), non-commercials (large speculators such as funds) and small traders below the reporting threshold. The contrarian reading watches the speculators: when large speculators are more net long or net short than they have been in years, the trade is crowded, and the unwinding of a crowded position can be violent in either direction. **Short interest** — shares sold short, often expressed as a percentage of the float or as days to cover at average volume — is positioning in single stocks. High short interest means many traders expect a fall, and it is also potential buying: every short has to be bought back eventually. That is the fuel of a short squeeze, which the Markets subject’s GameStop case (M21) shows from the market-structure side. **Margin debt**, published monthly by FINRA, measures borrowing against securities. It rises with bull markets and has historically peaked near some market tops, but the peak is only visible afterwards, and margin debt can rise for years before it matters. The pattern across all three is the same: positioning tells you where the fuel for a sharp move is stored, not when it ignites. • COT: weekly, Tuesday positions released Friday; watch speculators at multi-year extremes. • Short interest: crowded bearish bets — and future buying, the fuel of squeezes. • Margin debt: leverage that builds in bull markets; a context for fragility, not a timer.

What the evidence supports

The research on sentiment is more useful than the folklore around it, because it says where sentiment matters rather than whether it does. Baker and Wurgler built a composite sentiment index for U.S. stocks and found that when sentiment is high, the subsequent returns of the stocks most dependent on it — small, young, volatile, unprofitable, non-dividend-paying companies — are relatively low, and when sentiment is low, they are relatively high. Sentiment works through the stocks that are hardest to value and hardest to arbitrage. At the level of the whole market, the contrarian signal is weaker and noisier. Survey readings and put/call extremes have some association with later returns, but the relationship is unstable from decade to decade, extremes can persist much longer than a contrarian expects, and the signal arrives without a date. The practical use is therefore narrow and specific: as a reason to size smaller when a crowd is all on your side, as a reason to expect expensive protection when fear is high, and as a prompt to look for the disconfirming evidence the crowd is ignoring. One more caution belongs here because it is the one most often broken: sentiment indicators are easy to read in hindsight. Every famous top and bottom has a chart where sentiment was extreme. The test is whether the same reading, in real time, would have been acted on without knowing the outcome — which is exactly what the replay below asks you to do. Using a sentiment extreme, properly — Wrong use: “Fear is extreme — buy now.” · Right use, sizing: The crowd is on my side: smaller size, tighter plan · Right use, options: Protection is expensive: buying it now costs more · Right use, thinking: What is the crowd ignoring? Look for the disconfirming case ← A sentiment chart with the famous turning points marked on it proves nothing — every extreme looks like a signal once you know the ending.

The AAII survey, read against its own history

The American Association of Individual Investors has asked its members every week since 1987 whether they expect stocks to rise, fall or stay flat over the next six months. Over that history the average has been about 37.5% bullish and 31.0% bearish, and those averages are the right frame: a reading means something only as a distance from them. The two most extreme readings are famous because they came close to turning points. The highest bullish reading on record, 75.0%, was taken on 6 January 2000 — about two months before the Nasdaq’s peak. The highest bearish reading, 70.27%, came on 5 March 2009, two trading days before the S&P 500’s crisis low. It is tempting to conclude that the survey times markets. It does not: many lesser extremes came and went with no turn, and the two famous ones are remembered precisely because they worked, which is the survivorship problem of P8. Used honestly, the survey is a thermometer for the crowd you are part of. A reading far above the bullish average says you are in a crowded trade and should check your size and your disconfirming evidence; a reading far above the bearish average says protection is probably expensive and selling now means selling with the crowd. Combine it with positioning data — what people hold, not just what they say — before drawing any conclusion. • Weekly since 1987; long-run averages about 37.5% bullish and 31.0% bearish. • Record bullish: 75.0% on 6 January 2000. Record bearish: 70.27% on 5 March 2009. • The famous extremes worked; many lesser ones did not. • Read it as a distance from the average, and confirm with positioning. Two record readings and what followed — 6 January 2000 — 75.0% bullish: The Nasdaq peaked about two months later · 5 March 2009 — 70.27% bearish: The S&P 500 bottomed two trading days later ← · Most other extremes: No reliable turn — the reason it is context, not a timer

Leverage and flows: the fuel gauge

Surveys measure mood; leverage measures fragility. FINRA publishes margin debt — money investors have borrowed against their securities — every month, and it has tended to climb to records near major market peaks, including those of 2000, 2007 and 2021, before falling sharply in the declines that followed. The reason is mechanical rather than mystical: borrowed money has to be repaid when prices fall, so a market carrying a lot of it has more forced sellers waiting below. Flows tell a related story. Money moves into funds after strong performance and out after weak performance (P8), so record inflows into equity funds tend to arrive late in a rise and record outflows late in a fall. Speculative gauges — the volume of short-dated single-stock call options, the assets in leveraged and inverse funds — rise in the euphoric stage of a mania (P13) and collapse after it. None of these measures tells you when a decline will start; margin debt can rise for years before it matters. What they tell you is how violent a decline could be once it starts, because they measure how much of the market will be forced to sell rather than choosing to. That is a reason to size smaller and avoid leverage yourself when the gauge reads high — not a reason to bet on the timing. • Margin debt (FINRA, monthly) climbed to records near the peaks of 2000, 2007 and 2021. • Fund flows chase performance: record inflows arrive late in a rise. • Speculative gauges — short-dated calls, leveraged funds — swell in euphoria. • Leverage measures how violent a fall could be, not when it starts. What each gauge says, and does not — Margin debt at a record: Many forced sellers below — not a date · Record equity-fund inflows: Late-stage buying, chasing performance ← · Surging short-dated call volume: Speculative euphoria — a stage, not a top

Sources

• Investor sentiment and the cross-section of returnsBaker & Wurgler, “Investor Sentiment and the Cross-Section of Stock Returns” (Journal of Finance, 2006)
• The VIX and its methodologyCboe Global Markets, VIX White Paper
• Commitments of Traders reportsU.S. Commodity Futures Trading Commission, COT report explanatory notes
• Weekly investor sentiment surveyAmerican Association of Individual Investors (AAII) Sentiment Survey