Forecasting where the U.S. economy is heading is currently impossible. So rather than building a predictive model, here is a collage of perspectives to build a composite view of what may happen.
There are two recessionary parts. First, a recession due to the Covid pandemic. Second a financial recession due to the imbalances introduced by government.
We start with a description of the components of this collage so that readers can judge for themselves what they agree with and what they disagree with. For those not interested in the behind the scenes, skip to the “U.S. Economic Outlook” section further down.
The collage has three parts:
Each component will be updated as we learn more.
Perhaps surprisingly, the most difficult component to get a grip on is “where does the economy stand now?” Fortunately, there is a solid statistical relationship between GDP growth and growth in unemployment claims and in the civilian workforce.
ΔGDP = f(ΔClaims, ΔWorkforce)
The Stata graph below shows the fit of the model:
We then applied this model to the months of March (actual claims) and April (half actual, half estimated claims). The new observations are not far out-of-sample, so we are somewhat confident in the results.
The resulting estimates are shown below.
This establishes the starting point in the collage.
Next, how does an economy rebound after a short, intense economic shock? There are few examples of this. The closest we could think of (with solid data) is the aftermath of Hurricane Katrina in Louisiana, 2005.
Below is the economic activity level in Louisiana before, during, and after the hurricane. Note that the economic contraction only lasted 2 months but the full recovery took 21 months from the trough.
Thus, the rebound was not instantaneous. Is the prolonged rebound instructive for the current Covid recession? It is hard to tell. On the one hand, Louisiana suffered physical damage beyond the human toll. On the other hand, it represents only 1% of the U.S. economy and should be more agile in the aftermath of an instantaneous shock.
As it happens, the rebound was quite quick compared to regular recessions (more on this below). We therefore feel comfortable using this as the second component in the collage. The shape of the Louisiana curve thus represents our Covid recession.
The U.S. government is trying to stimulate the economy to save jobs and to avoid a total economic collapse. The same is happening around the world. Never before has the world undertaken an economic experiment of this scale.
Many observers believe this will lead to a financial recession over and beyond the pandemic-induced recession. How should this be built into this collage?
Using ideas from a paper by Reinhart and Rogoff (2104): “Recovery from Financial Crises: Evidence from 100 Episodes” below are the U.S. recessions and GDP growth since 1948.
The key question is how quickly an economy bounces back after reaching the trough during the recession. We will turn to physics for this.
In physics, how an object bounces is described by the Coefficient of Restitution (COR). If a ball bounces perfectly with no energy loss, COR = 1. If it stops without any bounce and thus has 100% energy loss, COR = 0. Golf, basket and table tennis balls have high CORs: >0.8.
COR = velocity after collision / velocity before collision.
In economics, velocity is growth. We define economic COR as:
COR = GDP growth after recession / GDP growth before recession
We also transform the data so that 0 < COR < 2. COR = 1 means that the economy grows at the same pace before and after the recession. If COR > 1, there is a chance that the rebounding economy will make up for the GDP losses during the recession. COR < 1 means that the losses will not be recuperated. Many economists have noted that usually COR < 1 (except they do not express it with COR).
The graph shows COR for U.S. recessions since 1948 and the Katrina recession in Louisiana. Median COR = 0.9. (Note that the Katrina recession has a high COR, as mentioned in the previous section.)
Many observers see the current recession as equal to or worse than the 2008/9 Great Recession. This suggest a COR around 0.9-1. However, we remain hopeful that the bounce will be stronger because the initiating event is not a financial meltdown. We thus use a COR in-between the Great Recession and Katrina, around 1.2.
This establishes the third component in the collage.
Combining the three components allows us to create two scenarios.
In the Covid recession it is assumed that the interventions by the Federal Reserve, the Treasury Department, and other actors are benign and have no repercussions once the Covid recession is over.
With this assumption we see a short recession but the recovery still takes almost 2 years. This is the Katrina logic overlaid on the initial shock. There will not be a “quick V”.
The decline in 2020 of -4.7% is a bit lower than the IMF forecast of -5.9%. We do not know why since the IMF does not provide any details. We hypothesize that they assume a deeper trough than we have modeled. Or maybe they see the start of the rebound a bit later than us (May).
What does this outlook imply for people, companies, and government?
In all, an unpleasant yet manageable scenario.
This recession assumes that the $2 trillion printed by the Federal Reserve and the $2.3 trillion of stimulus decided on will lead to financial havoc soon. It may be purely inflationary effects, but it is more likely complex mechanisms we cannot fathom yet. In addition, household wealth is shrinking.
It took 69 months after the trough of Great Recession for the economy to re-climb the peak before the recession. This collage has the same time frame and the recapture of lost GDP happens in October 2025. We will experience growth already in a few months, but it is growth from a low base.
The decline in 2020 is -6.1%, slightly higher than the IMF’s -5.9%. The big difference is 2021 when our outlook has a tepid 1.5% growth and the IMF has 4.7%. The IMF states that they view government interventions as benign, which explains the difference.
How should the government’s monetary interventions be evaluated? There are two views:
If the interpretation holds, we see the following implications:
Where do we come out on this? We do not. It is impossible for us to judge which way things are blowing.
What we will keep my eyes open for are signs of financial turmoil. If we do not see them, we will lean toward the Covid recession. If we see them, the Covid + financial recession.
These signs, if they arrive, should be visible fairly soon. Sometime within 6 months is our guess.
We certainly hope for the Covid scenario. Our experienced management has lived through the incredible recession in Mexico in the 1980s, was in the middle of the Gulf war recession that hit Sweden hard in the early 1990s, and was part of the Great Recession in 2008/9 waking in the middle of the night to check what Asian and European markets could indicate for the US. No fun.
Let us work hard to avoid such an outcome. We remain optimistic that in all its messiness the government is doing the right thing and that companies and individuals find creative ways to overcome the difficulties.