“And I think as those green shoots begin to appear in different markets and as some confidence begins to come back, that will begin the positive dynamic that brings our economy back.”
- Ben Bernanke, former Federal Reserve chairman in his interview with "60 Minutes" in March 2009
“And there’s been some signs from different partners or different companies that ad spending whilst down considerably has actually improved a little bit, maybe some green shoots from the declines that you might have been referencing at the end of March.”
- Heather Bellini, Goldman Sachs analyst asking a question on the earnings call for the company built around Google, April 2020
“I know it’s early and don’t want to over extrapolate, but just any more detail on the types of behavior you’re seeing, which verticals or which categories or which geographies were you seeing sort of the first sign of green shoots on the search side.”
- Brian Nowak, Morgan Stanley analyst asking a question on the same Google earnings call.
Is it green shoots season already? When then-Federal Reserve Chairman Ben Bernanke famously used the phrase, he ended up calling the bottom pretty well. But that was well after the carnage of 2007 and 2008 was behind us, after Lehman Brothers had filed for bankruptcy, after the AIG and Fannie Mae, Freddie Mac bailouts, after we’d already suffered through three quarters of negative growth over the prior year.
I know we’re all looking for some good news, but maybe it’s just a little premature for green shoots. Green shoots are all about changes in the rate of change. Yes things are bad, but they might be getting less bad. Things aren’t good, but they might be better than they were. This has led your authors to return to a favorite internal argument: Which is a more valuable approach to economic data, level or rate of change?
Levels are important. Levels make headlines. After Bernanke uttered those words in March of 2009, the unemployment rate would go on to rise from 8.3% to 10%, with nearly 3 million new job losses in the subsequent 11 months. Those numbers are levels, and those levels look bad. Bernanke looked wrong, and people told him so.
But green shoots aren’t about levels, as we’ve said. What Bernanke got correct was the rate of change of those numbers was slowing. Yes, almost 3 million jobs were lost after the “green shoots interview” but nearly 6 million had already been lost. The rate of job loss was slowing.
“Green shoots” became a derisive term, referring to over-optimism when things were at their worst, desperate for any silver lining. We are at that stage right now, but unlike March 2009, which was a year into the recession, we are only two months in. We’ve only gone through one month of the COVID-19 rent payments era. But who knows. Let’s let the data help us figure it out.
Returning again to our high-frequency data, it appears the rate of decline in activity has slowed across major metropolitan areas, now down 25% from the first week of January in Dallas, 65% in New York City, compared to 50% and 75% declines a month ago, respectively.
This is a perfect version of the rate of change versus level argument: The level of activity is down quite a bit from where it was, but things are on the upswing and the worst is behind us. Indeed, even OpenTable restaurant booking data is no longer down 100% across the board! Atlanta and Houston are now only down 98%.
OK, so maybe being optimistic about that chart is a stretch. But it’s something. That’s the thing about being down 100%: You can’t go any lower.
Below is an example of a level that looks pretty bad. Here we estimate expected total payrolls for this week’s April jobs report. Total payroll represents total employment times hours worked times hourly wages. Utilizing biweekly survey data from a special study , we expect employment, average hours and average wages to fall by 17%, 29% and 5%, respectively.
Total wages paid in April have likely returned to mid-2013 levels. This is only being somewhat shored up by actual benefits paid so far in the CARES Act expansion, according to draws from Treasury , which brings us back to mid-2015 payroll levels. This is the important level baseline for any green shoots discussion.
The rate of job loss appears to have slowed, with “only” 3.8 million jobless claims filed last week, but it is slowing slowly and remains at an absurd, unprecedented rate. Google trends data expects further multimillion job losses this week, as the pace is slow to fall.
We are likely to see an unemployment rate north of 20%, which is far and away a record since the Great Depression. That is not a level we thought we’d ever see. While it’s nice to think green shoots, and see the level of unemployment rise more slowly, we think when it comes to the labor market indicators, these awful levels still overwhelm improvements in rate of change.
How quickly might the recovery begin once we are able to move beyond the stay-at-home orders? How quickly could the rate of change improve? One of our favorite COVID-19 surveys comes from PricewaterhouseCoopers. Every two weeks it asks chief financial officers, “If COVID-19 were to end today, how long would you estimate it would take for your company to get back to business as usual?” You can see their responses in the chart below. While their answers are interesting, what is most useful is how their answers have trended over time. Yes, this is a level versus change thing as well.
The trend of responses is the picture of entrenchment. In mid-March, a strong majority said business would return to normal in less than a month. Now, less than 1 in 7 say the return would be that quick. More now say it will take over six months to get back to normal, with 8% even indicating over a year.
In short, the longer the level of activity is this low, the more difficult it will be for the rate of change to rise quickly. The sentiment was echoed in the Fed’s statement this week.
“The ongoing public health crisis will weigh heavily on economic activity, employment, and inflation in the near term, and poses considerable risks to the economic outlook over the medium term,” the statement read. The Fed conceding to weakness for the medium term, defined by Powell in the press conference as “a year or so,” is unusual. The Fed see hopes fading a V-shaped recovery, much like the CFOs.
The green shoots will come, but it might be early yet.
The Week Ahead …
The jobs report will likely trump all news next week, becoming historic in every sense of the word. We have indicated our expectations above, but the range of outcomes is also historically wide. Beyond the unprecedented scale of change in the coming report, surveyors working from home likely had more difficulty collecting data than usual. Perhaps the best guess so far from anyone with inside information has come from President Trump’s economic adviser Kevin Hassett. He's predicting 20% unemployment.