"A man should look for what is and not for what he thinks should be."
–Albert Einstein
In literature and film, a deus ex machina is a plot device wherein an unexpected and unlikely event resolves a seemingly hopeless or impossible situation. The term "deus ex machina" is derived from the ancient Greek theatre, where it was used to describe the practice of lowering actors playing gods onto the stage using a crane-like machine to resolve conflicts in the plot.
An example of a deus ex machina in contemporary media could be a scene in a movie where a character is stuck in a dangerous situation and, just as all hope seems lost, they are suddenly rescued by a stranger with a helicopter who appears out of nowhere.
In other words, it’s a magical quick fix of sorts that is ultimately an insult to the audience’s intelligence. Or, as Howard Mittelmark wrote in How Not to Write a Novel, the rough French translation of deus ex machina is, “Are you f****** kidding me?”.
Does such a plot device sound familiar in markets? Can you think of an unelected government body that descends onto the risk asset stage – acting like a God – at the moment when all hope is lost to save the day?
For most of the last 40 years, since the passing of the torch from Paul Volcker, the Fed has served as a deus ex machina for capital markets, but with the caveat that over that period the God-like arrivals have morphed from unexpected and unlikely to fully expected and all but guaranteed.
But what happens if there is no deus ex machina moment? What happens then?
Alternatively, what happens if that moment happens, but happens much later than expected?
Back to the Global Macro Grind…
Yesterday, we hosted our Q1 2023 Quarterly Macro Themes call and presented 129 slides of what I regard as relevant considerations and probabilistic conclusions, but what my colleague, Christian Drake, calls pure, unadulterated analytical fuego-ness (or alternatively, the quantum of macro zen and concentrated dose of epiphanic brain candy). The rough French translation of that is, the opposite of actively pursuing macro unawareness.
One of the focal transition points from 2022 to 2023 is the passing of the torch from inflation to employment. If the Fed’s hand was forced (much) higher in 2022 by inflation, it will likely be stayed from moving lower by labor’s resilience, at least in the aggregate, and at least for long enough to make the markets realize no deus ex machina is happening in 1H23 and potentially not in 2H23 either.
Consider the following data point from our presentation yesterday.
- Available workers per job opening as of November 2022 were 1.11. That is only very narrowly above its ATL of 1.0 set back in March/April of 2022 and well below its December 2019 reading of 1.6.
- But now consider the levels for this metric at which the Fed intervened in the last two pre-Covid downturns. For context, in January 2001 – the date of the first rate cut following the Dot-Com bust – available workers per job opening were ~2.0, and in August, 2007 – the date of the first rate cut ahead of the GFC – this gauge was at ~2.6. Looked at through roughly the reciprocal (the lens of Job Openings), the comparable figures at those two points in time were 5.2M (Jan ‘01) and 4.6M (Aug ’07).
- Currently, job openings are at 10.5M (unchanged M/M), and have fallen 1.4M in the last 8 months. While this may sound like a large and rapid decline, consider that at that pace it would take another ~30 months for openings to fall to 5.2M. In other words, don’t hold your breath on that rate cut.
Of course, as per Einstein’s quote this morning, a critical element of what we do here is avoiding being sucked in by confirmation bias. As such, I am an active seeker of cogent avenues of pushback/disagreement.
To that end, I was intrigued by an article I read yesterday about an Economist named Campbell Harvey who is notable because he was the first person to write about using the shape of the yield curve to predict recessions. He did so in his Doctoral Thesis at the University of Chicago in 1986. He’s now a Professor at Duke and isn’t shy about pointing out that inversions of the curve have accurately predicted 8 of the last 8 recessions, exactly as his early writings had suggested it would.
What was notable was that he is now questioning the accuracy of the yield curve’s inversion as an indicator of forthcoming recession. What? That seems notable. When the inventor of something (or, at least the first person to write about it) tells you to ignore it, it’s probably worth at least asking why.
His rationale is twofold. First, he argues that inversions of the yield curve have become so widely followed that they have lost some or much of their signaling value. He claims that from markets to businesses to well-heeled individuals alike, everyone is watching the shape of the yield curve. Consequently, he goes on to suggest, those same actors – to an extent – proactively prepare for the fallout from the curve’s inversion, thus reducing/mitigating its eventual or ultimate impact. Second, he suggests that the labor market this time around is extremely tight, so tight in fact that it is likely to make for only a mild recession or possibly even a soft landing.
What I like to do when I come across variant viewpoints like these is take a step back and ruminate on them for a bit, pushing the arguments back and forth in my mind to consider their merits. After doing so, a few things occurred to me.
First, with respect to the broader awareness of the yield curve’s impact, I think the most likely effect of that is simply a smoothing of impact – not a significant reduction and certainly not an elimination. The other side of the coin here is the reflexivity component. In other words, markets and businesses and individuals all making choices consistent with expectations of a forthcoming recession is likely to precipitate said recession through the channels of reduced investment and consumption. It may make the hill more gradual and less cliff-like (see my Wile E. Coyote Early Look for relevant cliff considerations) but the trip down the hill will happen nevertheless.
Second, regarding the labor market dynamic, this too has an obverse side of the coin. While it’s ostensibly true that a more resilient labor market should help buttress parts of the economy against worst-case outcomes, it’s also true that that same resilience will likely keep wage inflation running hot, and, by the transitive property, the Fed engaged on its hawkish quest. Interestingly, this is the exact caveat Campbell Harvey gave for why his This Time Is Different argument might not work out. He said that if the Fed tightens for too long or goes too high that a recession would become more likely.
Well, that’s kind of the whole point we’ve been making from very early on.
There is also a subtle, but not insignificant point about the composition of the labor market. In cycles past, it has been the base layer of the employment pyramid (lower skilled / lower wage) that has borne the brunt of the fallout in the form of significantly elevated rates of unemployment, particularly relative to the higher layers of the pyramid. This time around, however, this seems less likely to be the case. The newsworthy, large-cap layoff announcements that are happening with increasing frequency are not base-layer cullings, but those squarely in the middle to upper quintiles of earners. Facebook (Meta) announced 11,000 layoffs in early December last year (13% of its workforce) and its median-paid employee earns over $280,000/year. Yes, 11k people on a labor force base of >150M is small, but on a wage-adjusted basis, it gets larger.
That’s the interesting dynamic here – aggregate labor statistics are likely to be supported by the resilience of the pyramid’s base layer, but the earnings-adjusted pyramid looks more like an inverted pyramid and that is where the pain is and likely will continue being concentrated.
Zooming out a bit, the yield curve is but one measure/proxy for the economic outlook, and that is why we don’t look at just one measure.
To those who would take comfort in the potential resilience – for a time – of the labor market in the aggregate, I would ask – how has the market responded to strong labor readings of late? It’s always a bit risky to write about labor on the morning of the NFP report because any given print can move around significantly, but regardless of what this morning’s number brings it is my hope that you now have a better framework for considering the interplay between markets, inflation, labor and policy.
As I wrote in my Wile E. Coyote note, we would (continue to) advise sticking to the process and fading the hopium/narratives. Dispassionately mapping the RoC through our probabilistic GIP framework while overlaying the price-volume-volatility risk-adjusted signaling process – we would argue – has served our clients well thus far.
Immediate-term Risk Range™ Signal with @Hedgeye TREND signal in brackets
UST 30yr Yield 3.67-4.05% (bullish)
UST 10yr Yield 3.60-3.95% (bullish)
UST 2yr Yield 4.23-4.51% (bullish)
High Yield (HYG) 72.61-74.99 (bearish)
SPX 3 (bearish)
NASDAQ 10,120-10,607 (bearish)
RUT 1 (bearish)
Tech (XLK) 119-125 (bearish)
VIX 20.75-25.12 (bullish)
USD 103.30-106.51 (bullish)
CAD/USD 0.730-0.744 (bearish)
Oil (WTI) 72.14-77.94 (bearish)
Gold 1 (bullish)
Silver 23.25-24.68 (bullish)
MSFT 219-239 (bearish)
Bitcoin 16,309-17,021 (bearish)
To your continued success,
Josh Steiner
Managing Director
