- July retail sales were up 1.7% over February levels, likely due to the involuntary shutdown-to-reopening shift; however, 90+ day credit card delinquencies hit a 7-year high in Q2, and August spending has retrenched, suggesting consumer finances are becoming more fragile
- For those with six-figure incomes, Bloomberg’s Consumer Comfort Index relapsed hard in August, while headline renter confidence fell for a third straight week; those with the lowest and highest incomes are becoming increasingly uneasy over the recovery
- Per the University of Michigan, unemployment expectations are diverging at the lowest level on record; with pandemic uncertainty slowly coming off mid-July highs, per Gallup, and job market convictions at all-time lows, recovery cannot take hold until COVID fears recede
Was it the Dutch circa 1680, the Germans and British a century on, or Tabitha Babbitt of Harvard, Massachusetts in 1813? Bragging rights technically go to the U.K.’s Samuel Miller whose Patent #1152 in the year 1777 put the circular saw into existence. Serrating was not a huge leap for the man already acclaimed for having invented the sewing machine. But word traveled slowly back then. And women have been given short shrift of late with the coronavirus disproportionately hindering their job security as they (we) struggle to moonlight as home schoolers. So we’re going with Babbitt who crafted a notched tin disk and rigged it to spin with the pedal of her spinning wheel after watching two fellow Shakers struggling to rip logs with a pit saw. We’d like to think American lumbermen thank her to this day for replacing tediousness with efficiency.
That said, the circularity of economic data has become tedious. Markets still key off whip-sawing virus data. It was just a matter of weeks ago that the R-naught – a gauge of how many people will catch COVID-19 from one infected person – was above ‘1’ in all but two states. As of midday Sunday, two had improved to 29 states, including Washington D.C. That’s great news. But then you come to find that testing has dropped by 12% vs. a 19% fall in daily case counts. In Florida and Texas, testing rates had crashed by a third and more than 50%, respectively, by the middle of last week. Good news. Good news tempered.
Retail sales tell a similar tale. After adjusting for seasonal factors, sales were up by 1.7% in July over February levels. QI’s Dr. Gates postulates that the last three months’ retail sales data have been more technical than fundamental: “Discretionary spending reflects the path of involuntary shutdown-to-reopening. It’s like what happens when you turn a switch off, then turn it back on. What else is supposed to happen?!” Real-time data from Opportunity Insights, GlobalData, Earnest Research and Open Table show spending has retrenched anew in August.
As for household finances, we can’t yet determine the effects of declining unemployment benefits. We do know that statewide evictions have expired in 30 states and that up to 40 million are at risk of losing their homes. Fresh second quarter data from the New York Fed also revealed overall household credit fell for the first time in six years, posting the largest decline since 2013’s second quarter, the same three-month period 90-plus day credit card delinquencies were as high as the current 9.8% reading. Unlike other forms of debt, credit cards have not been shielded by the CARES Act; they provide a cleaner read on consumer balance sheets.
You might detect an emergent pattern when we weave into this narrative consumers’ views of their current finances via the University of Michigan backsliding in early August to its post-pandemic April lows and before that, a July 2014 low. Add to this Bloomberg’s Consumer Comfort Index among those earning $100,000 or more (yellow line) which made a hard about face in August. Last Thursday’s weekly Comfort Index also marked the headline’s first setback in three weeks. Weakness was particularly pronounced among renters whose confidence fell for a third straight week. There’s a lot less comfort among those with the highest and lowest incomes.
In one word, the data add up to confusion, or if you prefer, uncertainty. To arrive at Income Uncertainty (red line), UMich aggregates three consumer assessments: if times are good, if times are bad, and if the future is uncertain. As you see, the year-over-year trend in income uncertainty tracked the plunge in Consumer Discretionary Spending (blue line) which followed the COVID shock.
Since then, though, uncertainty has diverged from spending as households factor current economic conditions and future economic uncertainty into their decision frameworks for big ticket items like autos, homes and large household durables. We now see downside risks to consumption past July’s retail sales report. If spending momentum dissipates through the third quarter, we could remain on the saw’s edge, contemplating contracting GDP in the fourth quarter.
Did we mention confusion? Divvy up the percentage of households who see the job market as improving, staying the same and worsening. You get to the lowest divergence on record for households’ expectations for unemployment (orange line) in UMich records back to January 1978. As per Dr. Gates, “In ‘Ask the Audience’ terms, there has never been such low conviction about the future path of unemployment. At a minimum, this will create unevenness in the outlook for spending as some households will be more cautious than others.”
As we’ve mentioned in the past, we have great confidence in Gallup polling, which dates back to 1935, and is rated Least Biased by Media Bias/Fact Check, a resource on which we increasingly rely as the media is politically charged. In Gallup polling for the week ended July 19, a record 73% of Americans said they believed the pandemic was getting worse. By August 9, that level had dropped to 60%, the third-highest reading on record, but a marked improvement. If the future of confidence rests with the virus, we can only hope fear continues to recede. That’s the only thing that will extinguish the economic circular saw buzzing in the background.