May 9, 2020
Every time is different, not just this time. Like with people. We are all different, but we are mostly the same. It’s the differences that provide unique insights into any person, place or in this case, economic times.
As expected the employment report released yesterday was horrible. It was actually worse than I suggested. Two weeks ago I thought non-farm payrolls would drop to about 135 million people (black arrow on diagram below). It was much worse, dropping to about 131 million people (the red arrow on diagram below).
Meanwhile many other economic metrics are starting to show up in the data, more next week than this week. This post is to try to give context to some of the numbers, to put things into the perspective of history. In the last 20 years, employment has dropped down to this level twice. Below are some key indicators for those periods.
There is a lot of data here, and the data highlights why every time truly is different, but the overall theme could be a very simple take away. Every time there is a sustained decline in GDP, (aka a recession), the effects take many, many months to work through the economy. Every single time.
Here are the key differences that may be worth highlighting:
1 – While the population of the US has grown over the past 20 years, labor force participation near the bottom of a cycle was declining. This time, that participation rate has fallen through the floor. Most economists expect many of the employees who are not working during the COVID crash to be rehired. If 80% of those who have been laid off are rehired, then over four million Americans will still be ‘newly’ unemployed. During the 2008/2009 downturn, the process of laying off four million employees took 11 months and the S&P 500 dropped 36% during that period alone. (Yes, there was a three month bounce from 1260 to 1440 in the S&P index during that period.)
2 – One area that is starting to get attention is the level of the S&P related to earnings. Historically the price earnings ratio of the S&P averages around 20. During times of stress it goes much higher (earnings decline sharply and prices decline less). In this cycle the earnings estimates are just slowly starting to ramp down. The estimate that I have used here is from Yardeni Research last week, but I suspect companies will continue to make downward revenue and earnings revisions in the coming weeks and months. As that happens, the equity prices will likely fall as uncertainty increases.
3 – One common theme from my blog is the increases in government debts. They are rapidly increasing everywhere and much of this increase is being undertaken to support asset prices, primarily bonds and equities. As well, grants and loans are being made to individuals and businesses to sustain them, to help them make fixed payments. This significant increase in debt levels has been undertaken while taxes have been (recently) cut substantially. This will need to change in the not too distant future. When taxation does rise, it will further reduce employment, investment, earnings and profits, perhaps not uniformly, but it will happen.
4 – Government programs are supporting assets. As the central bank and other government programs use their grant and lending programs to remove the immediate risk of default for thousands of businesses, the ‘cash’ on the sidelines is being deployed in riskier assets (equities and high yield debt) forcing prices higher.
5 – Time is required. The significant and dramatic upset to the lives of individuals cannot be overstated and some of that altered behavior will be maintained for months or years. Other downturns did not have a health/safety component associated with them and they moved slowly through the system while causing dramatic effects on revenue, earnings and asset prices. To some degree this should be expected here as well.
This downturn is substantially different from any other downturn and many are predicting that the outcome will be different too. While it is unlikely that the outcome will be different, the timelines for events has reached the level of bizarre.
The pace of this shut down has never been experienced before. Similar events would have occurred in wartime, but the metrics are not around to examine them in detail. The restart is being pushed to reverse this and it may work well, but the changes in human behavior are difficult to predict. Unlike wartime, there is almost no destruction of capital property and no decimation of the human capital making that restart easier.
Asset prices, including stocks, bonds and commodities dropped rapidly and have risen almost as sharply. The justification is that earnings in the long term will not be badly affected. There are many economists on both sides of that argument, but it is highly unlikely that fundamentals will be able to return to their pre-COVID levels. The destruction to the economy will present itself in new ways and require time to repair. Specifically, the COVID crash is related to an exogenous event, and it affects the health and safety of individuals. This is very difficult to measure but it would be reasonable to assume that there will be a modest drop in demand for services (think travel, entertainment, etc.) and this will transfer to purchases of all manner of other goods and supporting services. Until a vaccine is developed this effect will likely persist at unusually high levels.
Further consider that periods of recession usually result in a 2-3% drop in GDP over 2-4 quarters. The last recession saw a decline of 6% occur over two years. The current downturn has barely started and cannot yet be accurately measured, but estimates for a 5% decline in second quarter (April-June) are on the low end and some guesses as high as a decline of 15% for the current quarter seem plausible. The Atlanta Fed has an unofficial ‘forecast’ for current GDP published yesterday. It is currently estimating a drop in GDP of 12.1 in the second quarter. That’s massive.
The duration of the declines is also difficult to guess, but high unemployment is a result of many things other than temporary layoffs. With many small companies shutting down and large companies suffering from significant drops in demand, the effects will be moving through the system for many months. A ‘V’ bottom in equity markets may be possible but a similar bounce is unlikely in commercial transactions.
There is another major disconnect with other periods. There are economies around the world that are beginning to open back up, but others (Russia, Brazil, Chile, Mexico) are still experiencing growth in their COVID cases. The movement of people and perhaps goods between these locations will be slow to return to normal. With so much of GDP related to foreign trade (about 12% in the US and 32% in Canada), some impact may be felt for a long time. The converse of that is that the production of some goods may be re-homed, or brought back into the country of demand. This will almost certainly raise prices, crimp profits and increase needs for capital at home, and hurt recovery in offshore manufacturing. (Imagine the effect of low oil prices on the budget of Saudi Arabia, or the freeze in clothing purchases effects on Bangladesh’s sewing shops.) While these issues may not appear to be significant on their own, the combination of these effects are likely to result in profound changes to economic models.
The economic environment will repair itself, but it may take many quarters and perhaps even years. The path to normal will not likely begin until AFTER a vaccine is created and widely available. With current thinking that a vaccine may not be widely available for 12-18 months, the path to normal is a long way off.
The path of prudence is the wisest course. Protect your capital and don’t worry about the fear of missing out (FOMO).