August 25, 2020
A lot of people – online and offline – have been asking about my opinion on the economy. When will the recession be over? Is the recession over already? How does the recession stack up with other events in the past? What are the risks going forward?
So many questions! Let’s take a look…
Oh, a few more things before we get started…
First of all, I want to give a big, big “Thank You” to all the faithful readers who nominated my blog for the 11th Annual Plutus Awards. I’m a finalists in the exact same two categories as last year: Best Content Series and and Best FIRE blog. Well, it’s stiff competition. Even if I don’t win I can rest assured that some of my good blogging/podcast buddies will get the price. But it doesn’t hurt to dream of taking home a trophy this year! 🙂
Second, I’ve appeared on two recent podcasts and I just wanted to give a quick shout out to both of them:
- I talked to Taylor at Passive Wealth Strategies for Busy Professionals. Most of the discussion was about my approach to Real Estate investing through private equity funds. I’ve never written a blog post about this topic and it’s on my to-do list, so people who are interested in this topic might want to check it out until I actually write something more detailed about the topic!
- I did a short 25-minute interview on “Retirement Tips Radio” talking about FIRE in general and the challenges of retiring in a low-interest-rate and high equity multiples environment.
Bet let’s get back to today’s topic, the economy. Here are my thoughts:
The recession is bad but not as bad as the headline numbers may suggest
I’ve made this point many times before, but just to get this on the record again: The 32.9% GDP plunge in the second quarter overstates the severity of the recession quite a bit. The quarter-over-quarter drop was about 9.5% but when you annualize this decline, effectively assuming that we’ll get three more quarters from hell like this, then you’ll reach the -32.9%. But quarter over quarter we saw “only” a 9.5% drop, which is actually surprisingly benign considering how bleak the economy looked in March and April. Add that to the Q1 GDP drop and you’ll get a 10.6% drop relative to the peak GDP figure in Q4 2019. By historical comparison, that’s a massive decline. Much worse than the Global Financial Crisis in 2007-2009, but not even close to the Great Depression or the recession around WWII.
In any case, the annualization is just the way GDP numbers are reported in the U.S. For example, in my “other home country” Germany the quarterly growth rates are reported as the simple Q/Q growth rate, not annualized. So, some people lamented that the U.S. with a 32.9% drop was hit much harder than Germany with a 10.5% quarterly decline. But when you annualize Germany’s decline it’s slightly worse than the one in the U.S.
Following the longest expansion on record, the 2020 recession was likely the shortest on record: only 2 months!
Even though the 2020 recession spanned two quarters, Q1 and Q2, if we look at the economy at a monthly frequency, the drop in economic activity was concentrated in March and April only. January and February still saw solid growth and thus were still considered part of the longest economic expansion on record. Likewise, May, June and July already saw very robust growth. Below are the time series for Payroll Employment, Industrial Production and Retail Sales. (Side note: quite intriguingly, GDP is never reported at a monthly frequency, though some private economic data providers have their proprietary monthly estimates. The NBER business cycle dating committee thus doesn’t use GDP but rather some alternative monthly data series, including payroll employment and industrial production)
So, it’s indeed possible that the recession really lasted for only 2 short months. By far the shortest ever on record! In comparison, the last five recessions were:
- 1980: 6 months long and the previously shortest recession in recorded history since the 1854.
- 1981-1982: 16 months,
- 1990-1991: 8 months,
- 2001: 8 months (and quite intriguingly, this recession spanned three quarters and we saw positive growth in the middle quarter, so the 2001 recession did not satisfy the naive recession definition of at least 2 consecutive quarters with negative GDP growth)
- and 18 months (2007-2009).
… but we don’t know for sure until more data come in!
So, why hasn’t the National Bureau of Economic Research (NBER) called the end end of the recession, effective 4/30/2020? Very simple: They are facing the exact same real-time data problem I outlined in a post a few months ago dealing with the timing of Bull vs. Bear markets. For a while after the trough you find yourself in a “limbo” state. If the economy continues to grow you’ll declare the end of the recession and, specifically, you’ll have to back-date(!) it at that past trough. But if the economy were to derail again after only a months of recovery, the NBER would likely call the entire event a recession, despite the temporary reprieve as in 2001, when GDP grew again half-way through the recession only to go down again the following quarter!
That’s analogous to the stock market world where you call the entire 2000-2002 market move a bear market instead of three bear markets straddling two bull markets in between, if you remember the chart from that May 11 post.
So, it all boils down to the two central questions…
Will the recovery continue? And will it continue long enough to qualify as a new economic expansion?
So far, things look pretty good for a continued recovery. And again, with recovery I’m talking in terms of growth rates, not levels of economic activity. For a while, the level of economic activity will likely remain below the peak levels in late 2019/early 2020. But as long as growth rates stay positive we should be in good shape. Likewise, we’ll have an elevated unemployment rate for a while. But as long the trend in the UER is downward, we’re in good shape. This is the part that a lot of non-economists (and maybe even economists) have a hard time to wrap their head around; it may still feel like a recession to a lot of people. But the level of the unemployment rate doesn’t matter, only the direction.
And there are some indicators that tell me the economy is on the right path:
1: The stock market has obviously recovered. The broad S&P 500 reached a new all-time-high and is now 50+% above the March 23 low. I could not find any past recession where the market recovered that strongly and the recession eventually continued. Normally, the recession ends within a few months after the stock market bottom. For example, in 2009, we saw the stock market trough in March and the economic trough in June. Actually, the 2020 recession was so short and abrupt, it looks like the stock market had its trough only about three weeks into the recession and a little over a month before the end of the recession.
2: I always like to look at high-yield bond spreads as a leading indicator. In the global financial crisis, the spread peaked in December 2008, even earlier than the stock market bottom. And spreads are also not just past their peak but even just about back to normal again, please see the chart below:
3: The TED spread (spread between the 3M LIBOR and the 3M T-Bill) is also not just below its peak, not just back to normal, but now even below the 2019 levels. See the St. Louis Fed FRED database for the time series chart.
4: Business climate indicators: I like to look at the Institute of Supply Management’s (ISM) Purchasing Managers Index (PMI) as a leading economic indicator. If you’ve read all of my blog posts (who hasn’t???) I’ve mentioned this indicator previously in other posts related to the connection between macroeconomics and finance. The ISM-PMI is normally a good indicator for timing the end of a recession. When it starts bouncing up from its recession low, you’re just at or just past the economic trough. When you’re back above 50 again as we have been since June, the economy is normally way past the economic trough.
And a whole slew of other indicators that all look pretty solid. Not all of them with their own chart for the sake of brevity:
- The ISM-PMI sub index “New Orders” (something of a leading indicator within the PMI leading indicator) is even stronger than the overall index.
- The PMI index for the services industry is above 50 again.
- Housing starts and housing permits (both very popular leading indicators) bounced back and are now above their year 2019 averages again (though still slightly below the early 2020 peak).
- Unemployment claims (again, one of my favorite indicators, as mentioned in several blog posts here) are still high, but declining at a rapid pace. And here again, it’s the direction, not so much the level, that correlates mostly with the direction of the economy.
- The Atlanta Fed real-time GDP tracking estimate: It stands at a whopping +25.6% for the third quarter as of August 18. And again, that’s an annualized number consistent with a Q/Q growth of about 6%, which means we’d recover more than half the drop during the recession in one single quarter. The fastest recovery out of a recession ever. Not bad!
Related to the last point, how is it possible for the economy to grow at such a fast pace again? I made that point in a post in April already and now’s a good time to rub in how accurate my guesstimate from back then has been. Mind you, this was my forecast from before the Q1 GDP number was released and I came to within a fraction of a percentage point of Q1 and about 2.6 percentage points of the the Q2 number. Not bad, ey?
In any case, if you assume that the drop in economic activity was concentrated in March and April and after that GDP slowly creeps back, you get a tremendous bump in Q3. Even though the growth rate in July through September is not even that large. You get this large base effect when comparing the average quarterly levels (not the quarter-end figures!) between Q3 and Q2. So, despite staying well below the 2019 GDP levels you get some really large GDP growth numbers in the second half.
Also, today’s post wouldn’t be complete without addressing the issue that started the whole economic mess. Which brings me to the next point…
The “you know what” situation is getting better as well!
I don’t even dare to use the exact phrase for fear of offending the search engine gods. So, instead of using the actual word for the “you know what” situation, I use a code word: “beer” – see what I did here?!
In any case, after the first wave of beer cases and beer deaths in the Spring, we saw a marked decline in May until early June only to see a resurgence again, of both cases and deaths starting in June. Will this derail the recovery? Well, so far it hasn’t. And it looks as though the second wave is now past its peak as well. Why wasn’t the second wave as damaging as the first? Well, most importantly, it was mostly the cases that went through the roof: about +100% relative to the first peak. But the “beer” situation only claimed about half the deaths during the second wave. Which means the fatalities divided by the cases was actually only a quarter of what it was during the first wave.
I’m not a medical expert, so I can’t say what’s the reason for the second wave to be less deadly…
- Maybe we’ve found treatments that finally work?
- Maybe the states that were hit the hardest in the second wave (FL, TX, GA, AZ) just had better procedures to protect the elderly?
- Related to the previous point, maybe younger folks are catching the beer virus now and their health is more resilient?
- Maybe all susceptible folks have already died in the first wave?
- Maybe we’re testing more and are more likely to find “light cases”?
It’s probably a combination of all. And related to the last point, see the chart below. It’s a scatter plot of tests (x-axis) vs. cases (y-axis), both in 7-day moving averages. There has been a surge in tests that coincided with the surge in positive tests. And the blue line looks like that despite the aggressive testing we’re now running out of folks testing positive. A very different picture compared to April when we probably had very limited testing capabilities and 22% positive results. I suspect that if we had tested 800,000 people every day back in April, we probably would have found close to 200,000 cases back then. So, don’t get too worked up about a resurgence in new cases now. There were a lot of unreported cases back in April!
So, in a nutshell, while I’m not happy with the resurgence in cases and deaths, and every death is a tragedy, I find the second wave of beer cases a lot less scary than the first one. And the stock market seems to feel the same!
As always, just to cover my behind so I can blame it on something or somebody if I’m wrong, here are some ways my rosy outlook could darken again:
End of the stimulus: Some people may claim that the end or at least the reduction of unemployment benefits will wreak havoc on the economy. Yeah, true, that’s a possibility. But it could also do the exact opposite. Previously, a lot of people were making more money staying at home and collecting UE benefits than working. Enticing workers to come back to work may actually help the economy. But I concede that economists are not in agreement on this one. The more “demand side” economists (old-fashioned Keynesians) will find the large transfer payments useful. The supply side economists (my personal camp) will prefer a slow phase-out of the payments that incentivize people not to work.
Inequality: This may not exactly derail the economy, but I like to point out the usual disclaimer when looking at aggregate economic and financial data; we tend to ignore that despite the aggregate recovery, a lot a people and businesses are still hurting. This is a very unequal recovery: corporations tend to do well, small/mom-and-pop businesses not so. The technology sector is doing well, brick and mortar companies not really. White collar employees do OK, blue collar workers not so. Multi-millionaire early retirees with a big stock portfolio are doing great, while single working parents are suffering. So, again, just a general disclaimer: I’m very much troubled about this recession being particularly damaging to lower-income Americans.
Politics: There is an election in November that will bring uncertainty. Obviously before the election and on election night. But even more damaging would be a contested election with lawsuits and uncertainty over who really won.
Inflation: Ever since the mid-1980s, the Fed has been blessed with mostly demand shocks: economic weakness coincides with low inflation. So, accommodative policy (low interest rates) always went together really well with low economic growth. Like milk and chocolate chip cookies.! In contrast, a stagflation scenario, high inflation and low growth, not seen since the 1970s and early 80s, would put the Fed in a real bad spot. Do you want to fight inflation and sink the economy? Or save the economy but let inflation get out of hands? So, an inflation shock could be a real spoiler here. And the inflation shock could come from many sources, including higher cost of running your business due to expensive health safety regulations, a new trade war with China, energy price spikes, etc.
More waves and more deadly “beer” strains, new waves during the 2021 flu season, etc.: Yeah, that’s a concern! I hope that doesn’t happen!
Other “black swan” (or not so “black swan”) events: Military conflicts and other unexpected stuff out of left field.
There you have it. I have a pretty optimistic outlook on the economy. But since I planned my safe withdrawal strategy well I can be assured that if the worst were to materialize again and we have another or a continued recession and bear market, I can handle that too. Hope for the best, prepare for the worst! Stay healthy and prosperous everybody!