Authored by Simon White, Bloomberg macro strategist,

US large-cap indices are currently diverging from recessionary leading economic data. However, a decisive steepening in the yield curve leaves growth stocks and therefore the overall index facing lower prices.

Leading economic data has been signalling a recession for several months. Typically stocks closely follow the ratio between leading and coincident economic data.

As the chart below shows, equities have recently emphatically diverged from the ratio, indicating they are supremely indifferent to very high US recession risk.

What gives? Much of the recent outperformance of the S&P has been driven by a tiny number of tech stocks. The top five S&P stocks’ mean return this year is over 60% versus 0% for the average return of the remaining 498 stocks.

The belief that generative AI is imminently about to radically change the economy and that Nvidia especially is positioned to benefit from this has been behind much of this narrow leadership.

Regardless on your views whether this is overdone or not, it has re-established growth’s dominance over value. Energy had been spearheading the value trade up until around March, but since then tech –- the vessel for many of the largest growth stocks –- has been leading the S&P higher.

The yield curve’s behaviour will be key to watch for a reversion of this trend, and therefore a heightened risk of S&P 500 underperformance. Growth stocks tend to outperform value stocks when the curve flattens. This is because growth companies often have a relative advantage over typically smaller value firms by being able to borrow for longer terms. And vice-versa when the curve steepens, growth firms lose this relative advantage and tend to underperform.

The chart below shows the relationship, which was disrupted through the pandemic. Nonetheless, if it re-establishes itself then the curve beginning to durably re-steepen would be a sign growth stocks will start to underperform again, taking the index lower in the process.

Equivalently, a re-acceleration in US inflation (whose timing depends on China’s halting recovery) is more likely to put steepening pressure on the curve as the Fed has to balance economic growth more with inflation risks. Given the growth segment’s outperformance is an indication of the market’s intensely relaxed attitude to inflation, its resurgence would be a high risk for sending growth stocks lower.

For those following the recent sharp drop in job openings, or perhaps merely fascinated by the narrative that AI will cause a margin-busting corporate revolution as millions of mid-level employees are replaced by a cheap "bullshitting" AI algorithm, then today's latest bizarro JOLTS report will come as a shock. That's because after three months of sharp declines, the BLS reported that in April the number of job openings soared by 358K from an upward revised 9.7 million to 10.1 million, the biggest increase since Dec 2022...

.... and printing not only above the median consensus which expected the trend to continue with 9.4 million job openings this month, but came higher than the highest Wall Street estimate! As shown in the chart below, the delta to median consensus print was a whopping 703K.

According to the BLS, the biggest increase in job openings was in retail trade (+209,000); health care and social assistance (+185,000); and transportation, warehousing, and utilities (+154,000)

The sudden, bizarre reversal in the job openings trend, meant that after falling to the lowest level since Sept 2021, in April the number of job openings was 4.446 million more than the number of unemployed workers, the highest since January.

Said otherwise, after dropping to just 1.64 job openings for every unemployed worker, the lowest since Nov 2021, in April there were 1.79 openings for every worker, a sharp spike back to levels that the Fed does not want to see.

To be sure, none of the above data are credible for reasons we have discussed before but the simplest one is because the response rate of the JOLTS survey is stuck at a record low 31%. Which means that only those who actually have job openings to report do so, while two-thirds of employers are either non-responsive or their mail is quietly lost in the mail.

Another reason why today's data is meaningless is that even as employers allegedly put up many more job wanted signs, the number of workers actually quitting their jobs - a proxy for those who believe they can get a better-paying job elsewhere, and thus strength of the overall job market - tumbled by 129K to 3.8 million, the lowest number since May 2021.

Even the Fed's WSJ mouthpiece Nick Timiraos ignored the stellar headline print, and instead focused on the plunge in quits, writing that the "rate of workers who are voluntarily leaving their jobs (including leisure and hospitality) is returning closer to pre-pandemic levels, a possible sign of less tight labor markets. Quits tend to rise when workers think they can receive better pay by changing jobs."

The rate of workers who are voluntarily leaving their jobs (including leisure and hospitality) is returning closer to pre-pandemic levels, a possible sign of less tight labor markets

Quits tend to rise when workers think they can receive better pay by changing jobs. pic.twitter.com/hAEAMy1I81

— Nick Timiraos (@NickTimiraos) May 31, 2023

And the biggest paradox: as pointed out by Peter Tchir of Academy Securities, the seasonally adjusted JOLTS quits rate was 2.4 (we reached a "peak" of 2.4 in July 2019), while the Hires rate (also seasonally adjusted) was 3.9% just like it was 3.9 in July 2019. So allegedly there are 3,000,000 more jobs available now than then.

So what to make of this bizarro, conflicting report?