Guess what, sports fans?
Nomura’s Charlie McElligott has “recalibrated” his (in)famous CTA model, which should be of enormous interest considering the recent fanfare around his calls. He dubs today’s update “HUGELY IMPORTANT”.
When last we checked in on Charlie – a hands-down reader favorite in these pages – he was calling for a “re-correction” in US equities following the YTD surge off the overnight lows hit just prior to the December 26 upside explosion.
That was on Tuesday and indeed, stocks have stumbled since then, thanks in large part to global growth jitters (heightened materially by the IMF’s latest outlook), persistent skepticism about the feasibility of the US and China striking a comprehensive trade deal by the March deadline and, of course, the government shutdown.
One point Charlie made on Monday was that CTAs, having covered some of the “Max short” position, are now in “no man’s land”, with trigger levels for additional covering some ways off.
Ok, so that brings us to Thursday, and the above-mentioned “recalibration” of the model.
“Given the strong performance last week in Equities, we have recalibrated our CTA model and it implies EVEN LARGER ‘BUY TO COVER’ $FLOWS than we had initially anticipated last week”, McElligott writes, all-caps and all.
In light of this “recalibration” Charlie notes that the “signal across our Equities futures positions is now just ‘-56% Short’ rather than ‘-91% Short’ previously measured.”
So what does that mean, exactly? Well, Charlie is (super) happy you asked. It means that on January 15, there was actually model buying of “closer to $16 billion of SPX” versus the previously estimated $8 billion and on January 18 (so, on Friday), there was an additional $5 billion worth of incremental buying. So that would be on these two days:
But more critical than that, according to McElligott, is that the model “is now showing a VERY outsized (and growing since 2H18) importance of the 1Y window across the various model time weightings, with ~78% of the overall ‘Equities’ model impact being driven by the 1Y window.”
If you think back to what was going on a year ago, you can probably imagine why he brings this up. Here is Charlie to explain “why this is critical” (and we’re going to use a block quote here because somehow, we imagine the potential exists for something to get lost in translation if we try to paraphrase):
1 year ago we saw the escalation of the US Equities / US Rates sell-off against the enormous cross-asset “short Vol” squeeze–so the “halcyon days” of early-to-mid January ‘18 have dropped-out of the 1Y window, and instead now we “pick-up” the very volatile days surrounding the Feb 5th 2018 “Vol Event” and larger “QE to QT” macro regime change era.
Points of interest then to be on the lookout for: 8-10 weekdays from now, there will be a number of “choppy” points that roll out of the 1 year window, all of which have outlier SPX returns (-2.3%, -5.4%, +3.3%) off the back of the VIX ETN market-event price action last year.
This would potentially cause “NOISY” signals from our model in 1-2 weeks’ time–assuming that our regression still generates significant weighting to the 1 year lookback window.
That said, IF we simply hover around flat in terms of performance the rest of this week, we project no significant changes to our model’s signals these next couple of days.
AT ZERO CHANGE PROJECTED-FORWARD BETWEEN NOW AND THAT NEXT WEEK WINDOW (a big “if” of course), the next “buy to cover” would be at ~2805 while the next “re-leveraging of the short” would be at ~2317–both far OOTM.
So there’s that. In case “recalibrated McElligott CTA model” wasn’t enough to pique reader interest (and I can assure you it was), Charlie also goes ahead and addresses comments delivered in Davos by Bridgewater’s Greg Jensen, who on Wednesday said the following about the outlook for global growth:
While people have certainly diminished their growth expectations and you’re hearing all about that at Davos, we don’t think they’ve done it enough. Earnings expectations particularly in the U.S. are too high, and generally the Fed and other policy makers are still expecting stronger growth than we see.
That underscored comments he made last month (when he predicted “significantly weaker, near-recession-level growth” in 2019) and those soundbites from Davos served as fodder for the doomsday blogger crowd, which seized on the occasion to generate God only knows how many clicks, thus generating God only knows how much ad revenue.
Anyway, as noted above, McElligott touches on this in his Thursday missive. “Media commentary coming out of the firm most known as ‘the face’ of Risk-Parity in recent days confirm the observations I have been making about our own Nomura QIS Risk-Parity model estimated positioning changes over the past few months likely corresponding with a ‘downshift’ view”, he writes, before getting specific for those who have missed his updates on this.
Charlie notes that risk parity strats “with a macro overlay are clearly positioning for a ‘slower growth, lower inflation’ future” and that’s apparent in “the massive estimated buying of government bonds in [Nomura’s] model vs selling of developed market credit and DM equities.” Here’s a visual:
There’s a ton more in the full note, and maybe we’ll work that into something later, but for the time being, that’s probably a sufficient dose of highly-potent McElligott macro musings, which have to be administered carefully (“use only as directed”).