April Income And Consumption: Near Bottom

Summary

April PCE was the worst economic report I have ever seen. Nominal PCE plunged by 14% on top of -7% in March. It is down 20% in 2 months. Unprecedented.
The big surprise in the income report was a huge jump in personal income driven by government assistance, offset by wage losses.
A third of after-tax income was saved. This is by far the highest reading on that. This is the key macro indicator going forward.
Even optimistic scenarios paint an ugly picture for the rest of 2020. The GFC took 19 months to return consumption to its previous peak.
Markets are underpricing the chance of a C19 second wave. I believe we will have one in a few weeks, centered around southern California, the upper Midwest, and the southeast.

Almost There

With BEA’s consumption and income reports at the end of last week, we now have April in full focus from the household end of things. I expect large revisions to the data, which we have already seen with Q1 and March.

The April jobs report, surveyed in the middle of the month, came in at 14.7% unemployment, but this likely understates the true number by the end of the month by about 5 percentage points.
Consumption took a further nosedive in April with some recreational services off 100% since February. There was some upward revision for March, but not substantial. Consumption is down 20% since February.
There was little sign of a recovery in April consumption, except in funeral services. I wish that was a joke.
Incomes made an astounding rebound, now way up since February on the back of unemployment insurance and the special pandemic programs.
But that income was largely saved, with the savings rate at an unprecedented 33% in April. This tells us quite a bit about the current bull market.
Housing, a huge category, held up in April, and it looks like most everyone made rent and mortgage payments.
Prices are cratering in most places, but food prices are spiraling up in response to demand and supply chain issues.

While I don’t think we are quite at the bottom yet, we are close to it. Perhaps May was it, once we see that data, which starts coming this Friday with the jobs report, surveyed a couple of weeks ago.

But in any event, this is the largest short-term change in consumer behavior of all time, and nothing comes close. How much remains in the absence of a safe, effective vaccine is up for grabs, and we’ll look at some scenarios for recovery.

A Quick Look At Q1 Revisions

I promised larger-than-usual revisions to the Q1 report, and we got them. It netted out to -20 bps on the headline real GDP growth number, taking it from -4.8% to -5.0%. Small upward revisions in consumption were offset by large ones from destocking.

The static investment number is a little misleading, as there were upward revisions in nonresidential investment, offset by downward revisions on residential.

Incomes

This was the biggest surprise of the report. After a sharp decline in March, Personal income was up over 10% in April.

Between February and April, personal income was up $130 billion. This was driven entirely by the CARES Act Pandemic Unemployment Assistance (PUA):

Unemployment insurance (UI) and PUA lifted incomes by about a quarter trillion dollars. Did Congress overshoot? It remains to be seen. In the first place, a third of disposable income was saved, by far the all-time record. That may be May 1 rent payments.

To match the $250 billion in assistance was an almost identical deficit in consumption (PCE). Savings rose by 342% since February, or almost $400 billion.

Dollars to donuts, a lot of that savings wound up in the stock market:

Putting it all together, this is where the income went, relative to February’s levels:

BEA. “All Taxes” includes income tax, and payroll taxes paid by both employees and employers.

Giant swings. Of note, that tiny slice of negative interest payments looks like credit card balances going down. We’ll know more about that on June 5 when the Fed’s consumer credit tables come out for April.

At some point, we should expect some of that savings to go into consumption - the savings rate cannot remain at 33%. But where it lands is of great importance going forward. We’ll talk about that in a moment.

On employment, it is very difficult to figure out what is going on. The BLS reported an unemployment rate of 14.7% for April. But in a note at the end of the press release, they explained that because of the unusual nature of data collection with no face-to-face interviews, and no one working at the BLS offices, the household survey saw a historically low response rate. On top of that, there was a huge amount of reporting error. They estimate that the rate was closer to 19.5%.

Under normal circumstances, unemployment insurance claims are one of the most robust data sets we have. State governments are party to every single transaction, so the records are very precise and timely.

Usually. It is clear that the archaic state IT systems are still overloaded and behind in delivering data.

On top of that, PUA gets reported in a separate line, all the way at the bottom of the table, counted in the “All Programs” breakout, which normally contains very little. Some PUA recipients are also UI recipients, but not all, so it is unclear how additive that is.

As unemployment rates:

As of 5/9, the unemployment rate is 21.5% using the All Programs number.

So what is the actual rate? No one knows, but probably in the neighborhood of 20%.

Consumption And Prices

A slower economy means less need for employment agencies, lawyers and accountants, so professional services will likely not see a full rebound. Funeral directors will be working overtime, however.

- Me, a month ago

March was a terrible month for consumption and it got much worse in April. For completeness' sake, out of the roughly 300 lines in the full PCE table, these are all the ones that were down in March, but up in April:

BEA. Funeral services. Grim.

Nominal PCE was down almost 14% MoM in April, smashing the previous record of -7% set in March.

The Y-axis Apocalypse continues with no mercy.

Breaking out the main subcategories, we can see that services dominate the report, even more than usual.

BEA. Durables in blue. Nondurables in green. Services in red/brown. Foreign travel in yellow. Nonprofits in purple. Foreign travel in US and nonprofit sales to households are both subtracted from PCE.

You can see that big red swath in the middle is services, none bigger than healthcare. Digging a little deeper into the subcategories, these were the biggest drivers in the report:

BEA.

Aside from vehicles, almost all services at the top there. Healthcare was the single biggest chunk by far. The winners list is much shorter, and a different order of magnitude in scale

BEA

The two food categories are very interesting. In March, the decline in food service was more than offset by the increase in groceries. With revision, this increased.

That never made sense to me. When you pay for food service, you are mostly paying for the service, not the food. There should be much less spent on food if everyone is eating out of the supermarket, not restaurants.

What was happening, clearly, was that people were hoarding food in March. Grocery sales in April were only 4% above February levels, while food service spending continued its decline.

A similar pattern emerges in household goods, except they are now down from February levels, as everyone makes their way through the 96 rolls of toilet paper in their garage.

Sorting by percentage loss, the crisis in travel, recreation and transportation services comes into full focus.

BEA. Yes, -100%. I could have kept going, but stopped at -50%. Next on the list is all the food and healthcare services.

As you can see, it’s dominated by services. The ones on the left half of that list are all businesses with very high fixed costs, and rely on crowd density, and reusing the same space multiple times a day to make up for it. These are going to see the toughest road to recovery, absent a safe, effective vaccine.

For completeness’ sake, the winners:

Turning to prices, they are also on a wild ride. First, the aggregates:

What we see when we look at the individual categories are many of the same problem area from consumption — weak demand brings lower prices.

These are not annualized numbers, just the price declines over 2 months, and these are shocking numbers. It’s hard to see the travel, transportation or recreation services picking up any time soon.

On the other side of the coin, we see that housing continues its ~3% annual pace of price increases uninterrupted. We also see price increases in areas where there were supply issues due to the early C19 action in China still lingering.

Of greatest concern is food, a necessity, and a large budget item for many households. When we break it down, some eye-popping numbers emerge:

BEA. Would you like some eggs?

I have really been surprised at the inability of the food supply chain to adjust to delivering to supermarkets instead of restaurants. Empty shelves and rising prices at a time when farmers are plowing under produce and having to cull livestock is a failure of the first order.

To sum up:

This is the largest decline in consumption ever. Frankly, this is the worst economic report I have ever seen. Until next month.
There is also a huge shift in behavior going on, but the losers dominate the winners as of now.
Nominally, the largest decline is in healthcare, travel, transportation, recreation and food services, along with consumer durables. Travel, transportation and recreation services will have the hardest time recovering, whereas I expect the opposite for healthcare.
Prices are reflecting the cratering of demand, and are mostly going down.
Food price inflation is a real cause for concern going forward.
There was little sign of recovery in April, except in funeral services. Some categories up in March reverted back in April.

What Would A Recovery Even Look Like?

We are getting close to the bottom - down 100% on spectator sports and live entertainment is as far as you can go. So I’m trying to conceptualize what a realistic recovery would even look like. The PCE table is a good place to start since it provides tremendous detail, and is about 70% of GDP. It is the driver of every GDP report.

When I pull the very worst of it, it accounts for $284 billion in losses, annualized out to $1.7 trillion, or a -7.8 pp drag on GDP, compared to Q4 2019. This is where we need to look at recovery.

So we will need assumptions. Lots of them. I will be looking at three scenarios best/middle/worst case for December 2020 nominal PCE levels.

Healthcare. This will see a robust recovery, as people who put off non-emergency procedures will flock back to doctors and dentists. Healthcare facilities and workers are already well-versed in sanitary practices, so trust can be established quickly. This is good news, and is by far the largest hole in PCE. Best/middle/worst recovery is 120%/110%/100%. Pharma and medical devices match the healthcare recovery.
Food services. This is a tougher one to judge. Absent a vaccine, restaurants will still not be able to pack them in on weekend nights, which many require to remain profitable. But a new model is emerging with limited seating, takeout and delivery, though it is unclear if that is a viable model for a lot of non-chain restaurants. Best/middle/worst recovery is 60%/50%/40%.
Gasoline. This depends a lot on how the global economy and demand recover. Best/middle/worst recovery are 100%/90%/80%.
Travel, transportation and recreation services. These will be the toughest areas, as they all rely heavily on density and volume to make up for high fixed costs. You will not be able to fill an airplane, train, theater, theme park or arena to capacity for some time. Best/middle/worst recovery are 50%/40%/30%. The exception is live sports, already down 100%, and zeroed out in every scenario.
Consumer durables ex-vehicles: I think this is an area were we could see a very robust recovery, as people decide to take some of that savings and buy a new TV or couch. Best/middle/worst recovery are 110%/100%/90%.
Vehicles. Vehicles came into this with a host of issues, primary among them is that US consumers have begun to favor used trucks and SUVs over every other category. The Hertz (HTZ) bankruptcy will flood the market with used inventory, and further drive the market down for both used and new vehicles. Best/middle/worst recovery are 90%/85%/80%.
Household, education and child care services. 100% recovery in all scenarios.
Clothing. This is a tough one. If more people will be working from home, there is much less need for new clothes, and clothing services as well. They also have massive inventory issues that go all the way back to the fiber producers, and this will keep prices depressed from the supply side. Best/middle/worst recovery are 95%/90%/85%.
Personal care services will not be able to operate at full capacity, but can probably work around that with extended hours and other modifications in the best case. Best/middle/worst recovery are 75%/67%/50%.

Adding that all up, even the best-case scenario is not pretty for December 2020:

This is not a V-shaped recovery; we will still be in a recession at the end of the year. I think we’ll be lucky to get back to those February levels by the end of 2021. A comparison with the previous crisis is grounding, to say the least:

The key thing I will be watching going forward is the savings rate. The GFC was a huge shock to people’s certainty about their retirement, and boosted the savings rate very high throughout the entire cycle. I believe that this will reinforce that, and keep the savings rate in the double digits for the foreseeable future. There will be no rapid recovery in consumption.

Literally, as I was typing that last paragraph, this headline popped up in my notifications:

I will be watching that savings rate like a hawk.

The Market Is Not Pricing In a Second Wave

That much is clear. What is less clear is whether there will be one. I believe there will be one, centered around southern California, the upper midwest, and the south.

I’ve been keeping track of the state data on C19 cases and deaths, and updating which states are going up, which are coming down, and which are going sideways. Here is the most recent chart from last Friday’s data:

COVID Tracking Project. Green circle = daily cases coming down. Red circles = daily cases going up. No circle = sideways, or inconclusive.

The states that are doing well are in two distinct groups to the left and right. To the left are states, mostly in the northeast, that had large early outbreaks that they are beginning to get their arms around. The other group to the right are mostly tiny and isolated states. Oregon and West Virginia stand out as non-tiny states that have done well, and they are close to hotspots as well.

The vast majority of cases until now come from those states on the left, whether circled or not. But the states that are the worst right now are near the median infection rate. They have a long way to go before they get as bad as the northeastern states. To take the starkest example, were southern California reach New York’s rate of infection, that would be about 350,000 new cases

After seeing the number of directions requests in Apple Maps decline dramatically in March, Apple began releasing daily data on that number, all over the world. It is far from perfect, but I think it provides a good rough measure of the extent to which people are coming into to contact, and having an opportunity to transmit the virus.

We’ve all seen the videos from Minneapolis. No matter what else you think of it, it looks like mobility has returned there, and I do not see a lot of social distancing. Apple’s data goes through Sunday, and we see walking direction especially pick up on May 28, last Thursday.

Apple. Number of directions requests, indexed to 3/9/20 = 100

Beyond everything else, the problem is that Minnesota is one of the worst states right now when it comes to their C19 epidemic.

COVID Tracking Project; Apple. The green line is pushed to the right 11 days to account for the 1-2 week incubation period of the virus, so 3/20 on the chart is actually 3/9 in the data. Data through 5/26.

Minnesota had just started to flatten out their curve at around 650 new cases a day, but I don’t think that will last. Neighboring Wisconsin also looks like it’s headed towards a larger outbreak: