Are we in recession already?

Setup

We are set for yet another down day in equity markets, with S&P and Nasdaq futures showing 3% losses as I write this. The real moves are in the bond market though because we’ve seen 20 basis point moves at the long-end of the curve as everyone is rushing into safe haven assets.

My contention yesterday was that the US economy was in a full blown recovery from a mid-cycle slowdown when the coronavirus pandemic hit. And that put it in the best position among large Western economies to withstand the shock that this outbreak presents.

The question is whether the shock is large enough to cause a recession and an L-shaped outcome because that’s where we see big job losses, a steep fall in equity indices and the potential for a financial crisis. In fact, if the coronavirus is so big a hit it precipitates a recession, then we are likely in a recession right now.

So, I’m looking at the data that’s being released now to gauge what it’s telling us about the economy. What follows below is my assessment. To spare you going through everything, I will say upfront, I don’t think the data are saying recession quite yet. Here’s why.

The yield curve

I want to start with US government bonds right now because that’s been getting a ton of play in the media. The fall in yields is absolutely frightening, with the 10-year bond yield falling at one point this morning to 0.695% and the 30-year well under 1.50% at 1.278.

How should we think about this?

I am looking at the yields as a signal of anticipated future fed funds rates, with the shape of the yield curve giving us a sense of when we will see interest rate cuts or increases and when to expect those changes. And what I see is what’s called a bull flattening, where long-term interest rates are decreasing faster than short-term rates. In the short term, normally, a bull flattener is ‘bullish’ in that it usually followed by higher stock prices and good times. But, we know that the market is really not discounting good times right now. And, what’s more is that, over the long-term, lower yields mean lower returns. So, let’s not take this bull flattening 100% at face value.

There’s also the shape of the curve. Last time I wrote you about this, I mentioned that the curve was inverted from 3 months to 2 years, whereas it had been inverted from 3 months to 3 years before the Fed’s 50 basis point rate cut. And last summer, when the 2-year – 10-year spread was inverted the front inversion was to 5 years. So, the inversion is shortening now, with the back end of the curve upward-sloping.

When I first thought of writing this piece yesterday, I thought of the curve shape potentially telling us something sinister about the potential for recession because the front-end curve inversion was shortening. The best interpretation of that shortening is that rate cut anticipation is moving forward. And, given back end steepening, that could either mean, cuts are moving so far forward that we get a recession or that the aggressive cuts mean we miss recession and normalize policy sooner.

But, having pored over the data, I realize, it’s more of a bullish sign than a bearish one. If you look at the spread between 3-month or 2-year Treasury yields and 10-year yields, the traditional metrics for recession, the numbers are getting more positive – 23 basis points from the 3-month and 28 from the 2-year. Now, the bears will tell you that’s because we had the inversion and now we are getting the steepening before recession. Nevertheless, I would hold open the possibility that it simply means that all of the economic stress is being moved forward and the back-end policy normalization with it.

The bottom line is this: the coronavirus pandemic means we move aggressively lower on base rates in the US. And so, the US curve is going to look increasingly like curves around the world. I would look at the UK curve as a baseline for where this is headed, with the 10-year Gilt now trading at 21.7 basis points.

So, concerns about inflation are completely out the window now. And there is plenty of upside left for Treasury bulls here. Remember that some bond market participants trying to keep the duration of their portfolios as long as possible may be chasing yield down because of the convexity of the curve. That convexity increases dramatically at these yield levels. And so the move to near zero rates may be aggressively swift in the US, just as it was in Europe. Be prepared.

Markit Composite PMI

What precipitated this post is actually not yields but the Markit services PMI numbers, which I think are more reflective of the US domestic economy than the ISM numbers that I talked about yesterday. The reality is that these numbers were bad.

The seasonally adjusted final IHS Markit US Services Business Activity Index registered 49.4 in February, unchanged from the ‘flash’ figure, but notably down from 53.4 seen at the start of the year. The contraction in output was only marginal overall, but was nonetheless the fastest in over six years. Firms attributed the decline to less robust domestic demand conditions and a further fall in export sales

Any number below 50 means recession.

And if we already see a print below 50 on the services PMI in February data, doesn’t that mean we go lower in March because of the coronavirus? And if so, if this bout of weakness lasts, that also means that February would mark the beginning of a US recession.

That’s how I am looking at the Markit PMI data then. And so I have been looking to see if other economic datasets confirm that thesis. The yield curve doesn’t necessarily say that’s what bond markets are thinking. So what about other data points?

Jobless claims

Claims are great real-time data because you can look at these numbers and get a sense of how much lost income we’re seeing relative to six months or a year ago and get a sense of the economic shock that means. But, before I tell you what claims are telling us, remember that the initial coronavirus shock is not a ‘people not having enough money’ shock; it’s a ‘people not spending enough money’ shock. So, it’s savings rates and consumption increases that matter more right now. It’s only later, when that consumption loss turns into job losses that claims data will verify.

The latest numbers show that the pre-conditions are favorable though i.e. the 4-week average initial claims number is 213,000. And that’s lower than the 221,500 we saw a year ago.  So, we are not yet seeing any impact from coronavirus on people losing their jobs.

Payrolls

The same goes for payrolls. The jobs figure for February just came out while I was starting this post and we saw a massive 273,000 print on non-farm payrolls, with the unemployment rate at 3.5%. Those are great numbers, especially this late in the cycle. And they were accompanied by upward adjustments to prior months’ data. The upward revisions of +85,000 to December and January put the rolling 3-month average increase in non-farm payrolls at 243,000.

So, we are not yet seeing any impact from coronavirus on firms hiring fewer people.

Moving forward, it’s all about coronavirus

So, I am going to leave the backward-looking data sets there. If I had to sum up, I’d say what I’ve been saying for many weeks; they show the US in a full-blown recovery from a mid-cycle slowdown just as the coronavirus hit.

The question now is how sizable the impact of the virus is going to be. And there, you have to wade into the murky territory of epidemiology a bit. Let me use Justin Fox’s piece from yesterday as a baseline because I think he gives us some decent numbers to ballpark this. This is the part that matters most:

The 61,099 flu-related deaths in the U.S. during the severe flu season of 2017-2018 amounted to 0.14% of the estimated 44.8 million cases of influenza-like illness. There were also an estimated flu-related 808,129 hospitalizations, for a rate of 1.8%. Assume a Covid-19 outbreak of similar size in the U.S., multiply the death and hospitalization estimates by five or 10, and you get some really scary numbers: 300,000 to 600,000 deaths, and 4 million to 8 million hospitalizations in a country that has 924,107 staffed hospital beds. Multiply by 40 and, well, forget about it. Also, death rates would go higher if the hospital system is overwhelmed, as happened in the Chinese province of Hubei where Covid-19’s spread began and seems to be happening in Iran now. That’s one reason that slowing the spread is important even if it turns out the disease can’t be stopped.

Could Covid-19 really spread as widely as the flu? If allowed to, sure.

The context here is important because the juxtaposition in testing and preparedness in South Korea and the US tells you to err on the side of worst-case outcomes. The US simply hasn’t tested that many people and so, likely, the virus has spread very widely without our knowing it. California is mandating that healthcare providers cover the costs of testing. So, we should expect testing to ramp up. But, a lot of people don’t even have coverage. And remember what I was saying yesterday about the lack of sick days for US workers. That is an additional risk that most other OECD countries don’t have. The US is unique both in terms of its economic system basically incentivising people to downplay health concerns in order to receive a paycheck and in terms of the lack of universal healthcare.

I was talking to an executive at a large US-based company yesterday about the steps they’re taking for the virus. And while the conversation left me heartened by how prepared his company was, the conversation was alarming regarding the likelihood for economic disruption. He basically told me that they are preparing for waves of work-at-home status for all but essential employees, even buying large monitors staff don’t have at home to facilitate productivity. And he also told me this plan is very near execution, that it’s not a matter of if but when this virus causes the plan’s execution.

What’s more, he and I both agreed that the China outcome is a best case scenario for the US given the draconian measures they took after slowfooting their response. The type of lockdown and quarantine measures they took are simply not possible in a country that prides itself on civil liberties like the US. If incubation in China was December and January and they are trying to normalize policy in early March, the equivalent US timeline means  normalization in May or June.

The Spanish Flu example

There’s one other thing to remember about this pandemic. We do have two separate strains now. So, we have had virus mutation. And so far, there’s no evidence one strain is much deadlier than the other. But that’s not necessarily how it will stay. The lesson of the 1918 great influenza pandemic is that mutation can make a virus return much deadlier months later.

The year was 1918, the last year of World War I. Earlier that year, a new strain of influenza virus originated in China, spreading around the world from there. This first wave wasn’t especially lethal, and over the course of the summer small outbreaks defined the disease. At some point during this period, the virus mutated, becoming far more lethal. It began killing healthy young adults.

The new strain arrived in the U.S. in the early fall, with two outbreaks in Boston at army and navy facilities. At this point, though, few people realized the dangers it posed. And in any case, it had yet to spread throughout the country. But that would quickly change.

It was that mutated deadly strain that took its toll. And in today’s context, that would mean the impact of this epidemic not lasting just until June but well into 2020, perhaps the end of the year.

Financial crisis

I am not really concerned about the equity market or bond yields falling right now. Unless we have a recession, a market correction or short-lived bear market isn’t fatal for investors. And while lower bond yields lower long-term returns, that is also not ‘fatal’ for the financial system.

What does concern me is a recession or a financial crisis, which is why I am focusing this post on recession. A recession means a protracted and deeper bear market in equities. It means lower earnings and lower price earnings ratios over a longer period of time at the same time that yields are plummeting for the long-term. That’s a dual whammy you can’t recover from if you are a defined contribution retiree as most people in the US are. And it also crystallizes all sorts of underfunding in defined benefit pension plans that are systemic in nature. In short, it’s an unmitigated disaster financially.

In terms of a financial crisis, again I think it’s related to liquidity drying up. And I am looking at the oil patch in particular for signs that this is happening. Spreads in the corporate bond space have not gapped out tremendously yet. But, there are no bond deals getting done right now. And that means there is a lack of financial liquidity right now.

WTI is trading below $43 a barrel. And this is quickly becoming existential for a large part of the oil patch. Given what we know about the likelihood for a lockdown and quarantine approach, expect this level to get worse for oil companies. We are definitely going to see a liquidity crisis in the energy space. The question is how bad it gets and how much it spreads elsewhere.

I have talked about repo in the past as a locus for liquidity strains because of the need for US dollar funding. And so, if you look at year-to-date currency moves, the dollar is actually rising against a fair number of currencies. DXY is not a good indication of what’s happening because it is skewed by the dollar’s fall against the euro, Swiss franc and Japanese yen. Some of the other currencies like the Brazilian real are getting killed. The potential for liquidity problems has increased.

My take

So, my view here is that we have to look at the past two weeks as sort of a break point. Before that time period, the coronavirus pandemic wasn’t acute enough outside China to matter. After that point it was. And so, all of the metrics before that period are largely meaningless except to the degree they tell you how susceptible an economy is to recession or crisis.

The US was sitting good as the coronavirus went global. But the US is not well prepared for what’s about to come. And I believe the economic fallout will be severe and long-lasting. Increasingly, we have to think about recession. And if we do have a recession, we are already in it right now.

That would mean a 40% down move in equities, base rates at zero and a 10-year yield approaching UK levels under 25 basis points. Moreover, we have simply not put the safeguards in place to prevent a liquidity crisis from worsening the outlook. And to the degree we get bankruptcies, the rush for dollar funding is going to be huge.

The days for happy talk are now over. This thing is getting serious. And over the next several weeks, the coronavirus will be all that matters for health, in the economy and for markets.

disasterEconomic Datainterest ratesJobsrecession