Was the extreme market turbulence witnessed over the past several sessions the product of a growth scare or was it indicative of crowded trades unwinding in unison?
Both. The answer’s both. The critical nuance — and I went into more detail on this in the new Weekly — is that the growth scare emanated from the US, and the attendant recession vibes and front-end rates repricing was fuel on the fire for a multi-week yen rally that was already in the process of stopping out shorts and forcing unwinds across the structural global carry trade.
If you’re curious as to the scope of the growth scare component, look no further than the figure below, from Goldman’s David Kostin.
That’s a pretty big blast of cyclical underperformance, where “pretty big” means on par with the COVID shock and the financial crisis. I think it’s an overshoot, but… well, what do I know, right?
During growth scares, particularly acute ones, correlation tends to increase for obvious reasons: Markets in thrall to growth jitters are, by definition, macro-driven markets. Outright macro shocks are classic “Corr 1” events.
As Kostin noted in his latest, a macro-driven market “stands in stark contrast” to the environment that typified 2024. Until last week, 2024’s market was micro-driven, defined by falling correlation (part and parcel of the crowded dispersion trade, which surely came under pressure in recent days).
The figure above gives you a sense of just how extreme the “crush” was headed into the growth scare: Correlations had collapsed almost entirely.
In the first half of the year, micro factors explained nearly 90% of returns for the typical S&P 500 stock, on Goldman’s math. For reference, the two-decade average is just 57%. By contrast, nearly every stock in the S&P fell on August 5.
“As investor focus swerved sharply away from micro back to macro news, realized 1-month S&P 500 average stock correlation rose from its lowest level in more than two decades [and] forward-looking market-implied correlations similarly jumped from record lows,” Kostin went on, editorializing around the dramatic spike illustrated above.
If you’re wondering what usually happens following growth scares as measured by a dozen historical instances of comparable underperformance for cyclicals, Goldman’s analysis shows realized correlation stays elevated as does the VIX, with both receding “only gradually” and staying “well above pre-scare levels even three months later.”
Goldman is right. The system went through a forced deleveraging. Fortunately the unwind was quick. Had it lasted longer, the Fed would have been forced to step in to prevent a system wide collapse. Credit markets reopened quickly, thankfully. But fed needs to ease and end qt. Otherwise this one will be a dress rehearsal.
That is quite an image!
I didn’t see anything about credit markets seizing up, repo rates going nuts, etc last week. Also no chatter about big shops teetering or other destabilization. Granted I’m not super plugged in, being on vacation, but I wouldn’t like to think I’d missed that.
My impression is that a whole lot of forced selling got done very quickly, but there was apparently a whole lot of bid too, and I don’t see why there won’t be a bid for the straggling forced sellers next week. SP50 broke 100D but didn’t even test 200D, and neither estimates nor macro data are breaking down, as opposed to weakening.
The Fed very likely will cut in Sept, I’d guess FOMC doesn’t have a strong lean on 25 or 50 yet, Treasury’s shift from coupon to bills has probably partly neutralized the remaining QT, and I’d guess Fed’s decision on QT is driven more by its assessment of reserve levels than by 2% and 3% swings in the S&P 500.