Our framework for thinking about Covid-19
With market volatility at 2008-09 levels, we are entering a period where long-term investors are presented with a unique opportunity to earn better-than-average returns if they can keep their heads and not flinch at just the wrong time—what we’ve said would be the “Be Not Afraid” moment.
The difficulty is markets are being buffeted by fast-moving events and cross-dimensional forces that span all the way from an uncertain health crisis, the potential for both supply and demand destruction over an indeterminate period of time, monetary and fiscal responses globally, a U.S. presidential election that could determine the course of whatever recovery lies ahead, and a reflexive feedback loop between all of the above and the financial markets that could be either very positive or very negative. Overlay all this against the backdrop of a stock market that, while attractive on a long-term basis, was short-term stretched (just two weeks ago we were only a few percentage points away from our target for the year), add the negative effects of crowd psychology and you have a recipe for chaos and panic.
At Federated Hermes, we have developed a framework for thinking about the present crisis, which along with some key milestones we’re watching across all dimensions, has the potential to reinforce our base-case scenario or cause us to adjust it. The purpose of this memo is to share that framework with you.
The health-care risk: Milestones to measure progress against the global pandemic
Given the novelty of this virus, we have no perfect template to answer some very key questions that ultimately will affect the outlook for the global economy and financial markets. How far will the virus spread? How long will it last? How many people in each country will be infected and worse, die, as it progresses country by country? No one knows, and most scientists, lacking the data, are unwilling to project potential outcomes. So we need a framework for understanding the data as it comes in and for projecting the trend and, importantly, the change in the trend. Here are the key stats we are following:
- Virus progression in China: Slowing. China arguably is the key country to watch since the virus started there. Importantly, it is a “worst-case scenario” for how the epidemic will unfold because the flu got a more-than-usual head start in establishing itself, in part because the government blacked out news flow about it for weeks and because China’s health-care system is less advanced than in the developed world. So it’s good news infection rates in China already have begun to slow to just 3% per week, far below the explosive growth witnessed early on and is now occurring (off very low initial numbers) in the U.S. and Europe. If the weekly new infection rates in China continue to decline, that would portend positively for the path of the virus in the developed world. But if it reaccelerates, that would imply the virus is likely to enter a dangerous second growth phase even after the first phase fades. We’ll see.
- Virus progression in Korea, Japan, and Italy: Still expanding. These are the next countries to watch. The virus landed there sooner than the U.S. and should peak sooner as well. If that occurs in these three countries within the next two to three weeks, consistent with what we’ve seen in China, markets will breathe a sigh of relief even if the expansion in the U.S. and Europe continues. But if the virus continues to expand in these early-stage countries, markets will begin to discount the possibility of a longer-term disruption in commerce and ultimately earnings.
- Virus progression in the U.S.: Still expanding in the very early phase. For the next two to three weeks, we expect a daily drumbeat of new and ever-growing numbers of cases and cities with infections. Like China, once better and more available testing kits/protocols are deployed in the field, an alarming “sudden” jump in detectable cases is probably on the horizon. Ultimately, we doubt the U.S. markets will fully stabilize until U.S. infection rates begin to decline.
- Virus progression in the presently hot, humid climates of the Southern Hemisphere: Low. We view the progress of the virus in hotter countries where it already has been introduced as accidental experiments that will help the markets answer a central question: like most others before it, will this flu epidemic be bailed out by the warmer weather only a few weeks away in developed Northern Hemisphere countries? Or is this new Covid-19 virus a monster than can better survive warm weather? With only 83 of a total of 97,886 worldwides cases of the virus reported occurring through yesterday in Southern Hemisphere countries, our base case is the former but we acknowledge the risk of the latter. If infection rates in the warmer countries above remain low, markets will take this as good news.
Our base case on the progress of Covid-19 is that with increased attention and safety precautions being taken across the world, infection rates are likely to peak and begin declining far sooner than the current market panic implies. Again, we’ll see.
Economic risks and key data to monitor them
Our view is the underlying health risks are potentially serious but likely overstated. Nonetheless, there remains a risk the economic impact is exaggerated by the individual reaction of the millions of market-economy participants operating within a collective, fear-driven crowd psychology (see insert at bottom). We are more worried about demand destruction than supply-side disruption, but are watching both.
- Supply-side disruption measures: Limited so far. Supply disruptions can be dangerous because as shortages of key products and services needed for the global economy to operate smoothly develop, the overall economy can start to sputter and eventually stall. If iPhone components become unavailable, iPhone production inevitably will slow. With less iPhones in retail stores, iPhone sales inevitably will slow as well. Etc., etc., etc. China is the epicenter of this argument, largely because it is a crucial global supply chain source and because in its attempt to arrest the spread of Covid-19, it literally ordered key production centers to shut down temporarily. The good news is factories there are starting to come back on line, and so far, few other production centers have shut down. This makes sense. Most modern factories are highly automated, and if necessary, could be operated by workers in protective gear if there was a buck to be made. We are watching for both anecdotal reports of further supply disruptions, along with published weekly economic series such as rail and truck volumes. So far we classify this risk as “limited.”
- Demand-side disruption: Limited but growing Demand disruption is the key risk ahead. As the public panics over the flu, and/or governments and corporations take preventive measures, we likely will experience additional demand disruption already hitting airlines, cruise ships, casinos, hotels and other components of the travel and leisure industries. Key potential weak spots include restaurant traffic in large cities, mass transit ridership and mall traffic. At the moment, most of these indicators are holding up, but we expect them to begin worsening in coming days. How long this will last will depend a little on the public’s perception of the risks of being out and about, even if everyone is taking normal precautions to limit the virus spread. Our view is that within a few weeks, the public in general will begin to assess the risk of the virus spreading more realistically, and with it, commerce in the retail side of the economy will begin to stabilize and even reaccelerate. Key indicators we are following include weekly series such as hotel occupancy and movie ticket sales.
- Liquidity events: None so far. A third area we are monitoring closely are liquidity events that grow out of either supply or demand disruption. In 2008-09, this actually was the core of the crisis, as the temporary downturn in real estate markets, of which the banking system held levered positions, produced a temporary impairment of bank capital ratios. Because banks largely are funded overnight through demand deposits and/or the commercial paper market, this exposed them to a systematic withdrawal of cash at the same time, spawning a massive chain reaction of called loans and defaults that shut down borrowers, cut off credit and tanked the economy. This time around, the banking system is healthy. But in some ways, we do have a temporary systematic force at work: the potential for retail businesses suffering from a sudden drop in sales and cash flows. Another weak spot could be oil producers suffering from temporarily depressed spot prices and volumes. Many of the former are small businesses that both power the economy and, along with some of the latter, are highly levered. So we are watching for signs of stress that could suggest they are in danger of going under through, again, anecdotal bottom-up evidence from our stock and high-yield teams, along with such indicators as the National Federation of Independent Business small business confidence index and Challenger Gray’s measure of layoffs due to bankruptcies. Our base case is that given the systemic nature of the problem, corporate landlords and bank lenders are likely to extend terms until the storm passes rather than throw all their loan and rental customers under the bus at the same time. We’ll see. If bankruptcies begin to rise, it would be a sign that a liquidity crisis is emerging. And that in turn could lead to the next area we are watching.
- Negative feedback loops: None so far Given the size of the publicly traded corporate credit markets and their increased importance in funding companies around the world, a freeze-up here could create a liquidity crisis that leads eventually to bankruptcies, rising unemployment, lower demand, lower earnings and still higher credit spreads, creating a reflexive feedback loop in the capital markets that takes on a life of its own. This is why we expect the Fed to remain focused on liquidity infusions (more below). In a related vein, we are watching the stock market itself, both for signs of compelling valuation and for potential negative wealth effects if the decline becomes long lived. And we are watching the potential for political feedback loops. That is, if the virus is quickly contained and the economy resumes its upswing, President Trump’s approval ratings should rise and the prospect of four more years of his pro-growth policies likely would buoy markets. On the other extreme, if things get so bad that voters become disillusioned with capitalism as an economic model, Bernie Sanders could rise in the polls and begin to take the delegate count lead again in the Democratic primaries. Joe Biden, now the Democratic front-runner, is probably viewed by the markets as market neutral.
- Employment declines: Low so far. The fourth area we are watching is employment, the heartbeat of the economy. If businesses disrupted by the forces noted above begin to take more drastic countermeasures, such as laying off workers, our base case of a likely short economic disruption could extend to something worse, potentially a recession. The first sign of this happening will be employment. This will show up in key indicator series, especially weekly jobless claims data and, monthly, ADP’s tally of payroll growth among small companies (under 50 employees), the government’s count of nonfarm jobs in the retail and leisure & hospitality industries, and Challenger Gray’s survey of layoff announcements. So far, so good.
The chess game of policy reaction
A third broad area of the present crisis is public policy reaction to mollify the impact of the pandemic on the global economy. Here, most of the talking heads have been, frankly, less than helpful. They conflate random various public policy tools, codifying them all as simply “Stimulus measures.” Specifically, we are watching for following potentially helpful policy reactions:
- Liquidity provision measures. Already underway, these key measures are vital. If we are right about the likely path here, the major near-term risk to the economy is that a short-term liquidity crisis, particularly among small businesses, morphs into store closures, bankruptcies and rising unemployment. So the monetary authorities in particular need to flood the system with liquidity to help these little businesses survive. This is why we liked the Fed’s sudden 50 basis-point cut, which some characterized as using its “stimulus bullets” too early. This cut should not be viewed as a stimulus bullet, but a liquidity bullet designed to avoid the need for a stimulus bullet. We think the Fed should further bolster liquidity through QE injections, in particular through direct purchases in the credit markets. This is where the real crunch seems to be occurring and could be the Achilles Heel in the capital markets. Other measures we are watching for include similar central bank action around the world (Hong Kong, Australia, Malaysia and the Fed so far this week), and regulatory relief by the appropriate overseers to encourage larger lenders and landlords to extend terms for borrowers being squeezed by the exogenous and systemic force of nature represented by the virus. (Very different from cutting off a sole borrower whose product/service is simply failing in an otherwise healthy economic environment.)
- Short-term fiscal measures that can help with liquidity problems. On the fiscal side, governments have the ability to provide short-term cash to smaller businesses in the form of tax holidays, such as Social Security tax payments. We expect the Trump administration to utilize these in the weeks ahead and if we see them, we will mark this in the progress column.
- Larger-scale fiscal measures to jump-start the economy if it heads into recession. Frankly, none are needed so far, as we are not in a recession. But should one come, the government certainly has the ability to legislate for large-scale tax or spending measures to spur the economy. We hope we don’t get to this stage but are comforted knowing that over the long term, such measures are available and likely would be employed.
Investment template: No time for panic
Against all this uncertainty, we are watching for opportunities under the assumption our base case (a short but sharp economic pullback followed by a dramatic back-half recovery) is correct. But we are pacing our investments to allow for a change in course should the areas we are monitoring turn more negative. We already have averaged in 2 points further to equities in our PRISM® stock-bond portfolio model on the first big downdrafts of last week (putting us at 70% of our maximum overweight in stocks), and have plans to add a third point if markets decline a full 20% from their highs. (In the past, 20% downdrafts have proven excellent long-term entry points, even in the 2008-09 crisis, by the way.)
At the portfolio level, we are similarly using big market down days to upgrade our portfolios, adding ever-higher quality long-term growth stories at significantly reduced prices. Again, we are advising moving deliberately and confidently into equity positions rather than trying to call a precise bottom amid the current volatility. We are entering the “Be Not Afraid” moment. Follow with us the progress of the crisis for sure, and adjust your strategy accordingly. But above all, remain confident and careful. This is no time for panic.
A commonly understood behavioral science flaw is called “reporting bias:” humans typically overestimate probabilities of an outcome when only one side of a binary outcome is reported regularly. The classic example is the lottery. Assume for the moment a lottery is in place where 415,869 tickets are sold. On the night of the draw, news stations around the country report the winner. Although the odds of winning the lottery are 415,868 to one, the average TV listener, watching the exuberant winner, overestimates his or her odds of success, and is likely to bet again next week when the next 415,000 tickets are sold. If on the other hand, the news stations were required by law to announce each lottery loser every minute of every day before they announced the winner, it would take 289 days before the winner was announced: “And now, Aunt Sally’s bet: lost. And now, Joe Smith’s bet: lost. And now, Margaret Jones’ bet: lost. And now …. etc., etc., etc.” Everyone would far better understand their odds, and no one would bet on winning. If we think about the odds of dying from Covid-19 based on the number of cases currently reported, we effectively would be playing a similar low-odds lottery. It would take nearly 9.5 months of minute-by-minute announcements every hour of every day of friends and neighbors who did not die from the virus before finally naming someone who did. Contrast this to today’s second-by-second news flow of the latest case contracted in … “New York!”… “Seattle!”… “New Jersey!” as the 24/7 social media frenzy fuels coverage that’s far more concerned with clicks, eyeballs and “likes” than with accuracy, facts and fairness. We think it is likely, with time, that the public in general will notice that despite the dire and breathless news flow, most people in their circle of life remain uninfected and healthy. With this, and maybe warmer weather, the panic phase we currently are in is likely to diminish.
March 4 2020
March 4 2020