October 29, 2020 – Today’s GDP release for the third quarter came in at 33.1%. Not a typo. After the disastrous second-quarter number of -31.4%, the worst quarterly number on record we now got the best quarterly reading on record. What’s going on here? What do I make of that number? Are we out of the woods now? I’m putting on my economist’s hat for today and share my thoughts in a short post. Let’s take a look…
My April 8 forecast still looks pretty good!
In a post in early April, while we were still in the middle of the recession, I had ventured a forecast for the quarterly GDP numbers in 2020: -4.6%, -35.6%, and +32.9% for the first, second and third quarter. Mind you, this was all before even the Q1 number was released. And the actual numbers were -5.0%, -31.4% and +33.1%. Not bad! Two years into retirement, I still got the forecaster mojo!
33.1% is an annualized number!
As I mentioned in my post “The Shortest Recession Ever?” a while ago, U.S. GDP numbers (both GDP and GDP growth rates) are reported at an annualized rate. So the true quarter-over-quarter growth rate is really “only” just about 7.4%. That’s still a massive growth rate. Just to put that into context, with a 2% annual trend growth rate, that’s almost 4 years worth of growth packed into one single quarter. So, we should certainly be ecstatic about the rebound, but don’t read too much into that 33.1% headline number! Just like the -31% disaster in the second quarter was an exaggeration of the true economic decline.
We’re still well below the GDP all-time-high!
So, GDP dropped by 31.4% in the second quarter and then grew by 33.1%. Does this mean that we are 33.1%-31.4%=1.7% above where we were before the pandemic? Well, that’s confounding quarterly and annualized numbers again. But even putting that aside, unfortunately, we’re still far away from a new all-time high. First, GDP had already dropped by 1.3% Q/Q (or 5.0% annualized) in the first quarter. If we add the 9.0% Q/Q drop (or 31.4% annualized) in Q2, then this puts the trough at about 10.1% below the old peak. Even the massive 7.4% jump doesn’t get us out of this hole yet.
Also, notice that growth rates are not exactly additive. If the trough was at -10.1% and we now grew by 7.4%, then we can’t calculate -10.1%+7.4%=-2.7%, so 2.7% below the old peak. The correct calculation would be (1-.101)x(1.074)-1=-0.035, i.e., 3.5% below the old peak. Growth rates don’t add up linearly, they compound. Most of the time the compounding generates “better” outcomes, think of the Rule of 72 where you need “only” 7.2% annualized growth over 10 years to grow your portfolio by 100%. But sometimes the compounding works against you, i.e., if your portfolio drops by 50%, you’ll need a 100% gain to come back to the old level. Bummer!
We’re still well below the old trend growth path!
I always make this case in the context of equity bear markets vs. bull markets: Just making it back to the old all-time-high, as wonderful and as blissful as that may be, is still only a pretty lousy outcome. It means you had an extended period, potentially several years, of only zero-percent growth. So, let’s do the following thought experiment: what if the economy had simply chugged along at a modest, conservative trend growth rate of 2% instead of falling into a recession? We’re now still 5% below that trend, see the chart below. But that’s still not all that bad. Because if we compare the 2020 recession with the 2008/9 Global Financial Crisis and recession, we’re looking slightly better today. Back then in 2011, when we finally reached a new GDP peak, 14 quarters after the recession started, we were still 7% below that 2% trend growth path. In fact, by that measure, we hadn’t even recovered any ground since the recession trough (Q2 2009) when we had a similar gap between actual GDP and the 2% trend path.
So, if we could eke out another one or two quarters of above-trend growth and close the gap some more we should be in really good shape. Yeah, there might still be a 3% or 4% gap, and that’s the permanent loss from the crisis because bars, restaurants, movie theaters, and many other businesses will potentially never be the same again in a post-pandemic world. But it’s still better than the 7% haircut after the Global Financial Crisis!
Actually, the shape of the 2020 recession, how short it was and how quickly we recovered and how we are likely going to recover is probably the main reason why the stock market a) didn’t fall as much as in 2009 and b) recovered to a new all-time-high after only a few months (and yes, I know, only the S&P 500 and Nasdaq did, the Dow 30 didn’t). We’re going through the recession at about 5x or even 10x speed, so the stock market simply replicates that fast-forward speed, too!
The GDP number is still subject to revisions!
Not a major concern, but for completeness and full disclosure, we should keep in mind that this is only the very first estimate based on the data we have available today. There will still be revisions as new data come in. And with new data, I don’t even mean macroeconomic data for October, November, and December. That’s all impacting the Q4 numbers to be released in January 2021. New information, more complete information and revisions of the July/August/September numbers will still come in over the next months. Of course, those revisions will not change the current number from +33.1% to -33.1%. But we could easily see revisions worth +/-5 percentage points.
What about the “you know what” situation?
One of the biggest threats to the outlook is the “you know what” situation. I don’t even want to name it for fear of falling out of favor with the search engine gods, which have been very kind to my little personal finance blog recently. So, just like the last time I wrote about the topic, I’m going to invent a code name to denote the situations we’ve been facing this year: “beer,” so, I’ll be talking about “beer cases” and “beer deaths” in the U.S. and abroad.
In the U.S. we’re going through what looks like our third(!) wave of beer cases. And each consecutive wave looks worse! A peak of over 30,000 cases in the spring (measured as 7-day rolling daily average), with a short reprieve down to about 20,000. Then another rise to about 67,000 in the Summer with a short reprieve to about 35,000. And now we’re at about 75,000 cases in October. With a strong upward momentum! We can also plot this relative to the population, i.e., daily cases per 1 million residents, please see the chart below. I also include the “beer fatalities” at the bottom. When looking at cases only, the three waves got worse, measured by cases, (100, then 200, then 230 cases per million). But the good news is that the fatalities didn’t rise: a peak at 6+ deaths per million during the first wave, then 4, and currently about 2.5 per million residents per day despite the increase in cases.
But just to be sure, there’s some troubling news coming from abroad as a lot of European countries are now locking down again. But if you look at the numbers, they also tend to have significantly higher case numbers, see below the comparison chart for the 29 countries I follow on WorldoMeters.info. Belgium now has a new case average of over 1,300 per day per million residents. Translate that into a country the size of the U.S., and you’re looking at over 430,000 cases per day. Yeah, probably if we had almost half a million new cases every day in the U.S. instead of 75,000, we might be locking down again. But we’re still far from that!
Also, in the chart below is the beer fatality rate per 1m residents. A similar picture: France and Belgium, countries with some of the most stringent lockdown measures in the Spring, are now a lot worse off than the U.S. And really shockingly, the Eastern-European countries (Poland Czechia) that largely escaped the earlier waves are now going through some of the worst waves. Czechia had 12 deaths per day on average over the last week. If the U.S. had a fatality rate like that it would be over 4,000 daily deaths. We haven’t seen 1,000 deaths on a rolling 7-day average since August!
So, looking at the numbers, I’m certainly concerned about the resurgence of cases, but it doesn’t look like lockdowns are necessary again. I cross my fingers and hope the momentum reverses again soon, just like in the Summer.
We’re not completely out of the woods!
Even in the best possible case where the “beer” cases and deaths stay under control and we avoid another round of lockdowns, we may not be completely out of the woods either. Unemployment claims are still elevated, and while the direction is still downward (which is positive for the economy, of course) the pace of the decline has moderated quite a bit.
I’m also a bit worried that some components of GDP might have gotten a bit ahead of themselves. Remember, overall GDP is still 3.5% down from the Q4 2019 peak. But looking at the components of GDP, notice how goods consumption, especially durable goods consumption has significantly surpassed the Q4 2019 peak. Same with residential investment (=residential housing construction). I wouldn’t be surprised if there’s even a bit of a retrenchment in Q4 in most of those series once the low-mortgage-rate party subsides a bit. The spike in the demand for TVs, cars, RVs, and other comfort purchases thanks to people flush with stimulus money will probably normalize again, too. I would hope that some of the other series, notably business investment picks up the slack if demand for RVs goes down again!
Conclusion
There you have it. Today’s GDP release was great news. I will still maintain my prediction that the recession ended on April 30, 2020. But not all is well. There’s uncertainty about renewed lockdowns and some potential hiccups for Q4 growth rates. So, despite all the economic optimism, I don’t want to sound like I’m spiking the football here. The official announcement from the NBER economists calling the end of the recession is at least several quarters away.