The VIX marked 15.99 early Saturday UTC. On July 29 it was 20.66. Two sessions took index volatility 22.6% lower and back under 18.65, the level it carried on July 21 before any of last month's risk-off started.
Across the same stretch the long end sold off. DGS10 printed 4.68% on July 30 against 4.61% on July 28 and 4.44% on June 30. Crude went with it: the live WTI mark reached $84.67, up 3.16% from $82.08 a day earlier.
The gauge that measured July's damage has round-tripped. Neither of the two inputs that did the damage has.
The steepening can finally be taken apart
Last week's most irritating number was 0.45. T10Y2Y jumped there on July 29 from the 0.35 implied by the July 28 pair, 10 basis points of steepening in a single session, and neither leg published a July 29 print, so nobody could say which end moved. July 30 publishes both: DGS10 at 4.68%, DGS2 at 4.23%.
Set those against July 28's 4.61% and 4.26% and the decomposition is clean. Seven basis points of long-end selloff against three of front-end rally. Seventy per cent of the widening came from the long end.
That settles a question carried at low conviction for a week. The test for a market pricing cuts was written down in advance and it was specific: DGS2 below its 4.26% July 28 level with DGS10 holding near 4.61%. First leg fired. Second missed by 7bp. What the curve did was bear steepen with a dovish tint at the front, which is the reflation branch's tape and not the front-end rally that would establish cuts are being priced. T10Y2Y widened again to 0.47 on July 31, with no legs published for that date.
One caveat cuts deep. Seven basis points of nominal yield says nothing about composition. DFII10, the 10-year real yield, last prints 2.41% on July 28; 10-year inflation compensation last prints 2.26% on July 29. A selloff driven by a higher real rate and one driven by wider compensation carry opposite messages for equities and duration, and this data cannot separate them. The next daily TIPS release can.
The calendar emptied
Volatility did not compress only in America. VSTOXX dropped 9.10% on Friday to 17.1572 from 18.8741 the day before, so whatever repriced index vol reached Europe too. Four releases sit inside the window across which the VIX travelled from 20.66 to 15.99: the Q2 GDP advance and Personal Income and Outlays on July 30, PPI and ISM Manufacturing on July 31. Event premium decays once the events clear, whatever they contained, and that is the dull explanation for most of a two-session vol collapse across two regions.
This pull carries no results for any of the four, so what they showed is not mine to assert. The price reaction is the only trace available, and it reads in one direction. A Q2 advance below 1.0%, validating the 1.3% GDPNow vintage from July 8, was the growth-crack case; that branch called for the front end to rally hard, equity vol to rise and crude to reverse toward $70. What printed was a 7bp long-end selloff, a VIX at 15.99 and WTI at $84.67. Something may still be wrong with growth. The tape of July 30 and 31 does not say so.
What Friday's single names do and do not show
Individual prices moved in a way the index gauge did not. Amazon closed at $271.58, up 15.32% on the session. Apple closed at $308.91, down 7.35%. Alphabet added 6.73% to $356.13, Meta 3.28% to $556.71, and XLY, the consumer discretionary wrapper, 3.29% to $116.09.
Running those returns into the vol print and concluding that offsetting mega-cap moves held the index still is the obvious move here, and it is not available. This data has no S&P observation later than 7,294.6 on July 29, no index weights, no single-name implied volatility and no correlation series. Four stock returns cannot be assembled into a claim about forward index volatility, which is what the VIX prices; they cannot even establish what the index did on Friday, and nothing here explains why any of the four moved as it did.
What the tape does establish is that investors had company-specific prices to argue about in a month when one variable, a real yield that added 21bp between June 30 and July 28, had been setting almost everything. That is useful for anyone picking stocks. It is not evidence about the discount rate.
One leg of a two-leg condition
A VIX below 18.65 appears as a component in two standing invalidation conditions on this book, and 15.99 completes neither.
Equities are neutral here, and the bullish flip requires the S&P above 7,482 sustained three or more days with VIX back below 18.65. Volatility satisfies its leg with room to spare. The index leg cannot be checked at all, because the last observation available is 7,294.6 on July 29, 2.5% under the trigger. Bitcoin has the same shape. Its bearish view invalidates on BTC above $68,000 held three sessions, or on high-yield spreads back inside 2.75% with VIX below 18.65. Vol clears half of that second condition; credit does not, since HY OAS last printed 2.84% on July 28 and has not been observed since. Bitcoin's own last mark is $63,715.9 on July 29.
Conditions get written with two legs because one leg is a mood.
The conditions index has a related problem running the other way. NFCI eased to -0.554 on July 24, before credit widened to 2.84% and before the VIX reached 20.66. Last week that looked like a reading four days behind the deterioration. This week the deterioration has partly unwound, so an index that never caught the tightening may not need to. August 5 is the first clean look.
Crude reclaimed the line it had been missing
On the live marks WTI sat at $83.43 on July 29, 1.3% below its $84.54 July 21 level, while Brent had added 9.4% from $81.62 to $89.27. All of the fresh energy leg was in the seaborne grade, and that asymmetry was the basis for reading it as a waterborne supply premium traceable to the July 7 Strait of Hormuz tanker strike rather than as a signal about American demand. At $84.67, the live WTI mark is back above its July 21 level for the first time in that sequence. The FRED spot series is a separate vintage and reached $84.25 on July 27, so the two should not be blended, but they now lean the same way.
That matters for the one position in this book with a dated catalyst. June's CPI cooling to 3.53% year over year from 4.25% was recorded with WTI at $70.56 on June 30. Ten-year inflation compensation reads 2.26%, two basis points above its June 30 level, and the breakeven implied by the July 28 nominal and real pair narrowed 4bp to 2.20% over the matched window. The crude input feeding July and August headline prints has risen 19.4% on the FRED series since June 30. Compensation priced against it has moved two. A WTI that has stopped lagging Brent makes that gap harder to defend, and August 12 CPI is where it gets settled.
The strongest case against all of this is that 15.99 is simply the right price. Two vol gauges in two regions compressed together, the front end gave back 3bp, and if July 31 PPI came in cool with no energy pass-through then the inflation impulse this whole argument rests on never arrives and the oil call is a chase of a move that already happened. A market that stopped worrying would have been right to stop. That case is live and it is not weak.
It still has to explain a 10-year at 4.68%, 24bp above where it sat on June 30, printing on the same day as the growth data that were meant to settle the argument.
Explore these indicators together: Chart VIX (Intraday), 10Y Treasury Yield, and 3 more on the Indicators Dashboard