Why Is the VIX So Low? Its Stocks Cancel Each Other Out

A calm index can sit on top of loud stocks. On Friday 25 September the VIX, the index sold as Wall Street's fear gauge, closed at 14.87, near its low for the year. It sat there because the stocks inside the S&P 500 were expected to move against each other rather than sit still. If you hold an index fund and one stock you chose yourself, you have probably watched this in a single afternoon. The stock lurches while the fund drifts. The VIX barely moves, and you are left wondering why the VIX is so low when your stock is anything but calm.

The options market saw the noise. Options on the members, averaged by size, priced them to swing about 2.6 times as hard as the index. Big moves were expected. They were expected in different directions, though, and inside an index those moves cancel before they can register.

Lucia's specialty is the crowd behind a chart. She treats a spike, a stall or a sudden swing as testimony from the people who caused it. Then she asks what they were afraid of. On Friday she brings that habit to the best-known fear number in finance, and her first read of it will not survive the afternoon.

It is a few minutes before the close in New York. The VIX is settling below where it opened. Her first read is the one the number invites: calm, she notes, maybe complacent. The week had given the crowd reasons to flinch. On Wednesday the ten-year Treasury yield climbed as high as 5.12%, its highest level since 2007 (Yahoo Finance). Yet the VIX spent all five sessions within a point and a half of the year's lowest close. A crowd that shrugs at a bond market like that, she decides, is either confident or not looking.

What does the VIX actually measure?

Before letting that word stand in her notes, Lucia checks what the number is built from. The VIX comes from options on the index itself. Those are the contracts bought for protection against a big move in the whole basket, or for exposure to one, over the next month. Cboe turns their prices into a single yearly figure. That makes it testimony Lucia trusts, because what a crowd pays for protection is harder to fake than what it says.

It also makes the VIX forward-looking. It is not a record of the past month but a price for the next one. Divide it by 16 for a rough daily figure. Friday's reading then priced the index to move about 0.9% on a typical day, which is low by the VIX's own history: its long-run average close is 19.42.

What the figure cannot show is where the quiet comes from. An index is a basket. Its swings depend on two things at once: how much each stock moves, and how much the stocks move together. Take two stocks of equal size on the same day, one up 3% and one down 3%. Each had a big day. Held together, the pair went nowhere. The rules for how an index adds up its members decide whose moves count for more, but they leave that arithmetic alone. Hundreds of stocks moving hard in different directions can leave the basket almost where it started. Nothing on a price chart labels the second ingredient.

Lucia pictures a station concourse at rush hour. Every person in it is moving fast, yet the crowd, taken as one body, barely shifts, because each of them is heading somewhere different. Film only the crowd's average position and you would swear the station was asleep.

The VIX only ever films the average.

So her question changes shape. A quiet index could mean the stocks are sitting still. It could also mean they are moving hard in different directions, and the VIX on its own cannot tell her which. She puts a question mark beside "complacent" and pulls up the two numbers Cboe publishes for exactly this problem.

Why is the VIX so low while single stocks swing?

Lucia starts with the Cboe S&P 500 Constituent Volatility Index (VIXEQ). It borrows the VIX's method but reads options on the individual members instead of the index, weighting each stock by its market value. She calls it the members' number; the ticker tells strangers nothing. On Friday it closed at 38.46. Weighted by size, the stocks inside the index were priced to swing roughly 2.4% on a typical day. The index built from them was priced for 0.9%.

She reads that twice, because nothing about it is calm. It is the price of a crowd in which every stock is braced for news that belongs to it alone. On a day like that, one large stock can jump on a contract while another sinks on a forecast, and the index closes nearly flat.

What the index lacks is not fear but agreement about what to fear.

That agreement has its own number. Cboe calls implied correlation "a gauge of herd behavior". The index, it says, measures "the average expected correlation between the top 50 stocks in the SPX index" (Cboe). Lucia shortens COR3M, the three-month version, to the correlation index. On its scale, 100 would mean the largest stocks move in lockstep, and zero would mean they ignore each other. Its long-run average is 41, and on Friday it read 10.32, lower than on 97% of trading days in the past twelve years.

A herd-behavior gauge near ten describes a herd that has scattered. For a fear gauge, the VIX turns out to spend a remarkable share of its life measuring agreement. Lucia looks up the day the herd scattered furthest. On 10 July the correlation index set its record low. That day the members' number stood above anything it printed during the shock this March, and the VIX closed at 15. The stocks were priced wilder than at the height of a crisis, while the index read like a slow week in August.

So she crosses the word out. The crowd behind Friday's close was split, and its splits cancel inside the index before they reach her screen. That is diversification doing its job. It is easy to mistake for a property the index owns. It is closer to a condition the members meet one session at a time, by disagreeing, and the index keeps its calm only while they keep meeting it.

The calm is borrowed.

What happens to the VIX when stocks start moving together?

Borrowed calm can be called in, and Lucia has already watched it happen once this year. Late in February the options market was priced for the same kind of calm: a market where winners and losers "dissociate", as Traders Magazine put it, with moves that offset inside the index. Then the conflict around Iran escalated over the last weekend of the month. The "Epic fury" operation "got into gear", in the magazine's words, and oil jumped 8% in a morning with tanker traffic in the Strait of Hormuz "essentially halted". Nobody at a desk decides when an index's stocks agree. A weekend can decide it for them.

Stocks that had spent the winter moving on their own news began moving on the same news.

She remembers how those weeks looked from her chair, every spike seeming to happen everywhere at once. For the first time in months, the crowd agreed on what to fear. The damage shows in where the VIX's jump came from. Between 25 February and 30 March the members' number rose about 10%, so the stocks were priced only a little wilder than before. The VIX rose 71% over the same weeks, mostly because the correlation index nearly tripled.

Lucia wants that split in one figure. So she divides the VIX by the members' number on each day that matters to her. What comes out is the share of the stocks' own expected volatility that survives into the index. It rises and falls with the correlation index.

The share is the moving part. On 16 March 2020, US trading halted market-wide at the open. That day the index carried 87% of its members' volatility, because almost everything was falling at once. On Friday it carried 39%. Apply this March's 68% to Friday's members' number and the VIX would read about 26. No single stock would have to get any wilder. That 26 is not a forecast, but it does measure how much of Friday's low reading rests on the stocks continuing to disagree.

If you trade crypto, you have seen that agreement in its extreme form. Coins with nothing in common turn red within the same hour. A portfolio of ten coins turns out to have been one position wearing ten names. That is correlation running toward its maximum. The index's calm rests on the opposite condition. On Friday the market priced that condition close to a record, with the correlation index about three points above its all-time low.

How should a low VIX change the size of a stock position?

All of which brings Lucia back to your two positions. The VIX describes the index fund reasonably well. It says very little about the single stock, because your stock gets none of the benefit of the cancelling. In Lucia's station, the fund is the crowd's average position and your stock is a single walker. It moves on its own news, at its own size, whatever the rest of the index does.

Say you hold $2,000 of each. At Friday's pricing, a typical day would move the average member about two and a half times as far as the fund. With a restless stock, the gap is wider. Equal dollars feel balanced, because the amounts match on your screen.

The risk does not match.

Lucia's rule is to size each position off its own range, so that a typical day costs a similar amount in each. You do not need the options market for that. You cannot use the members' number either, because your stock is not the average. A stock's own average range does the job: the average of its daily high-to-low swings, overnight gaps included, written as a percentage of its price. Say the fund's range is about 0.9% and the stock's 2.4%. Then the stock position that risks as much on a typical day as $2,000 of the fund is about $750.

March adds a second reason to keep the VIX out of this arithmetic. When the stocks fell into step, the average member's priced swing rose by about a tenth, while the fund's rose by more than two-thirds. So a low VIX misleads you twice. It understates the risk in your stock today, and it hides how quickly the risk in your fund can grow once the stocks start agreeing.

Lucia is not trading volatility here, and she is not asking you to. She is keeping two ledgers apart: the crowd's number for the crowd's basket, and your stock's own range for your stock.

A quiet index built from loud stocks

Lucia closes Friday's notes with "complacent" crossed out and "split" written above it. The VIX did its job. It priced the basket, and the basket was calm because its members disagreed. What it could not tell her was whether any one stock was calm, or how long the disagreement would hold.

She opens the notes again on Monday morning. Shortly before half past eleven in New York, the VIX is up about 10%. The members' number has risen less than 3%, and the correlation index has climbed with the VIX. Even that jump comes mostly from stocks leaning the same way.

Her habit now runs to three numbers instead of one. Before calling a market calm, she reads the VIX beside the members' number and the correlation index. A low VIX over a low members' number is a quiet market. A low VIX over a high members' number is a loud market whose noise is cancelling out, and the correlation index tells her how much of the quiet depends on that.

Is a low VIX a warning, then? "It isn't a clock," Lucia says. "It's a description of a crowd that, for now, is moving in different directions." The VIX is built from options on the whole index, the protection a large portfolio buys against everything falling together. For that buyer it is the right number. If you own one stock, you were never the customer it was built to quote. An index built from loud stocks stays quiet only while they disagree.

In March, agreement took one weekend.

Friday's ratio belongs to the average stock, so find the one for yours. Run the ATR tool in the simulator on the S&P 500 token and on Nvidia or Tesla, then split the simulated $5,000 balance so a typical day costs both positions the same dollars. Check back each evening for a week, and let a balance that was never real absorb the surprise if the stock's slice still turns out too big.

Size both positions free →