Our article entitled, “Earnings Models: has the Rally Outpaced reality?,” intimated that the market’s swift advance since the late-March low was predicated on a quick earnings recovery. Investors are already paying high premiums for a share of these earnings. More specifically, FactSet’s most recent measurements set the forward 12-month price/earnings ratio for the S&P 500 at 21.0 – far above the 10-year average of 15.1. In other words, the market is trading as if it expects a “V”-shaped economic recovery – a fast recovery that corresponds to the fast decline - as the base case. Certainly, some earnings recovery was inevitable because economic activity was artificially and systematically suppressed during the lockdown. Energy prices recovered throughout May, as Americans were allowed back into their cars. Mortgage applications and restaurant bookings have seen a nice pop from pent-up demand. American Airlines recently announced that it was scheduled to fly 55% of its capacity by July, up from 20% in May. We expect certain industries (restaurants, retail, cinemas, hotels, airlines) might take months to fully recover, despite the near-term recoveries. Therefore, the market expectation of a full and rapid recovery seems challenging. Wall Street and government analysts now have enough information to construct more accurate models that proportionally weigh the continuing impacts of coronavirus on the real-world economy, and that is what we’re going to review today.
Know your enemy: the coronavirus
While the lockdowns were painful, it did buy time for medical professionals to ramp up testing and avoid a catastrophic influx of patients beyond capacity of our health-care system. Moreover, we are also several weeks into promising treatment and vaccine trials. Investor sentiment has clearly been trading sharply up on hopeful news for COVID-19 treatments, even if the general public won’t directly see the benefits of these treatments for weeks or months to come. Despite the time needed to develop and test medical treatments, the scientific and medical community has made meaningful strides even without a silver bullet to cure the disease.
We collectively have learned a lot about how to reduce infection rates. Early in the outbreak, the primary disease vector had been thought to be contaminated surfaces, and now we know the virus spreads more commonly through the shared air or person-to-person contact.
As a result, we’ve largely reversed the recommendations on masks. Originally, masks were regarded as unnecessary for the general public due to lack of efficacy (and the legitimate need to reserve effective PPE for healthcare professionals). Since then, additional evidence and case studies over the past few months have demonstrated that masks, however imperfect, are effective in lowering infection rates. Medical professionals are honing their techniques on treatments as well. Early in the outbreak, mechanical ventilators were considered the key resource for treatment. New evidence and studies have shown that COVID-19 is as much of a cardiovascular disease as a respiratory disease; ventilators were potentially overused early in the outbreak, and now blood thinners and anticoagulants are being investigated as a potential therapy for severe cases.
The contours of the public-health crisis are important for data models because changes in the public-health forecast will largely determine our economic fortunes. For instance, if the US gets an uncontrolled second wave of the virus in the fall, that could lead to a resumption of lockdown measures in some states. Conversely, a successful prevention, early-testing, and tracing regimen will minimize the impact to consumer confidence and spending, encouraging a robust recovery.
Confident economic modeling
There are still plenty of unknowns about the novel coronavirus. We are still fine-tuning our understanding of infection-fatality rates and case-fatality rates. There are truly unknowable questions, such as the duration of immunity granted by a (hitherto nonexistent) vaccine or antibodies from a prior infection. Still, the evidence presented thus far regarding the COVID-19 coronavirus puts it squarely in the realm of probable misfortunes that inevitably occur every few generations. In other words, while the public-health crisis is frightening and tragic, the new coronavirus itself is not unusually lethal (like its cousins, SARS or MERS).
While the US is technically still in the first wave, the encouraging data on weekly hospitalization and deaths suggests that the virus’ spread can be controlled with disciplined application of the measures we’ve taken thus far. This fact is encouraging as a caring human being, hoping to minimize pain and hardship. It is also encouraging from a data modeler’s point of view because it limits the potential scope of potential economic actions and reactions to what has occurred thus far. The apex of economic activity was probably hit in January, while the low hit the US in April. Prior to a lockdown, nobody really knew how the economy would manage. At least now we have a sense of what an economy running at 70% strength looks like. We know what the economic impact is for the various prophylactic measures a state or nation can enact. Moreover, the patchwork response of each of the 50 states provided a diversity of inputs (limits of individual mobility and curtailed business activity) and outputs (health consequences) that modelers can use to fine tune their estimates. Other countries (Italy, China), which were ahead of us in the outbreak, are also ahead of us in the recovery, which creates guideposts for our own eventual recovery. Variance in national response (e.g., Sweden’s economic crisis despite its “soft lockdown”) provides even more data points. Given the wealth of data and precedent, we can be more confident in the estimates generated regarding our own rates of recovery.
Earnings & GDP forecasts post-COVID-19
May 2020 will probably mark the bottom of the US economic activity, but it will take time to figure out the full extent of the damage. FactSet reports that an unusually high number of S&P 500 companies – 1 in 3 – have withdrawn earnings-per-share guidance for 2020. While this will stymie economic modelers and increase uncertainty, the experience of other countries exiting lockdowns provide macro-level guidance on earnings growth. In Q1 2020, the blended earnings decline for the S&P 500 was -14.6%, but the impact of the coronavirus was only being felt in the latter half of the quarter. As an example, Citigroup’s current estimate for earnings decline bottoms out in Q2 (-42.9% year-over-year decline), with a muted recovery in 3Q (-24.8% YOY) and 4Q (-12.4% YOY). Put it all together, and several analysts are forecasting S&P 500 earnings in 2020 to be roughly -24% lower than they were in 2019.
Regarding gross domestic product, the Congressional Budget Office projected a 38% drop in GDP. Of course, that drop won’t be sustained for the entire year; several businesses are already tentatively reopening after forced closures in April. GDP is expected to start climbing in the latter half of the year. For instance, Goldman Sachs predicts a bounce-back of 29% in 3Q 2020. Put it all together, and you end up with a median forecast close to the American Bankers Association’s median projection, which is a 5.4% annualized decline of GDP for 2020. This estimate is reflected by Moody’s chief economist Mark Zandi, who calls for a 5.1% GDP retraction in 2020.
Unemployment forecasts post-COVID-19
Barring truly extraordinary circumstances, late April probably marked the low point in US unemployment. In the past 10 weeks, about 42 million jobless claims were filed. The total number of people receiving aid is currently 21 million, down a bit from the 25 million peak. ADP’s recent unemployment report suggested about 2.7 million private-sector jobs were lost in May, far better than the 8 million expected by some economists. The Department of Labor’s jobs report on June 5 surprised analysts to the upside, showing that the unemployment rate was only 13.3%, whereas many analysts expected a much worse number. The economy created 2.5 million jobs last month. The implication of these reports is that the labor force is being called back to the work faster than originally projected.
We’ve most likely seen the worst of the unemployment numbers for a while. We all hope that the damage to US employment is temporary and that unemployment will continue to decline in the coming months. The more important question is how rapidly (and to what level) the employment picture will improve. These estimates were made before the DOL report on May unemployment was released, so the estimates will assuredly need to be revised. However, to give you a sense of the range, a recent Bankrate survey saw end-of-year unemployment forecasts from a low of 6% to 20%, but most estimates were far more contained than these outliers. For instance, Oxford Economics predicts the unemployment rate will remain a dismal 10% by December. JPMorgan expects the 10% unemployment range will last at least until early 2021. There are more optimistic estimates: Federal Reserve regional president Charles Evans hopes for a 9% unemployment by the end of the year. There are more pessimistic estimates: the aforementioned American Bankers Association median projection of 10.9% by December 2020.
While these dismal estimates might be seen as overly pessimistic given the May unemployment number of 13.3%, it’s important to know that the Department of Labor suspects that the actual unemployment number is closer to 16.3% due to discrepancies in how employers are filling out their reports regarding furloughed workers. Employment is also hard to factor correctly during times of stress. Employees may be working reduced hours (to limit employees in a workspace) or for reduced wages (in response to weaker demand) as a result of the lockdown, but that employee is still listed on a payroll. Poorer, half-employed workers still imply a limited potential. The U6 number, which shows unemployed and underemployed, is still very high, at 21.2%. Thus, the consensus year-end target of 10% unemployment is still very possible. We’ll be looking for undated estimates as data continues to trickle in.
Many unknowns remain
For all the good news about our ongoing recovery, there are still many unknowns to consider. The biggest unknown, of course, regards the coronavirus. If we get a second wave, all expectations for recovery will be reviewed and deferred. States and local governments should be doing all they can to balance the needs of the economy while encouraging behavior to continually suppress the virus rate of transmission.
Employers are eager to get back to work, but there will be a period of demand monitoring, which will slow the recovery after the initial bounce-back. As a result of the pandemic, consumer-spending patterns have changed. About half of us are still wary about eating out, shopping, and going to social events. As a reminder, about 70% of economic growth is driven by consumer spending. If Americans expect lean times, they reduce spending. If Americans reduce spending, it lowers economic activity, thus creating a self-fulfilling prophecy. In this instance, personal-savings rates have skyrocketed to 33% in April - much higher than normal, and more than enough to limit growth.
There will be lingering effects on businesses that will take months, or even years, to determine. For example, we’ve been investigating the effect of this pandemic upon real estate. If workers aren’t required to congregate at an office, will employers stop paying for expensive midtown office space? We’ve learned that most leases for high-end office space have long contracts, say 20 years. Therefore, it could take decades to determine if the collective demand for premium office space peters out.
On a shorter time horizon, let’s consider the relationship between individual households and their creditors – banks, mortgage holders, car loans, credit-card companies, and the like. For the past few months, unemployed workers were given a temporary patch to mitigate the damage from losing their job through the CARES Act. These unemployed workers were able to maintain their creditor payments (rent, mortgage payments, credit-card bills) because they were getting state benefits (say, $300 a week) augmented by federal payments (an additional $600 per week). The additional federal benefits stop in July. Once those payments stop, how many creditors will be left without payments? Some economists are worried about a late-summer wave of defaults that have been hidden by the CARES Act, but not put off forever. Banks aren’t overleveraged like they were prior to the financial crisis of 2008, but a wave of defaults will still hurt the system. Smaller businesses without deep relationships with banks and other lenders, may be forced to close without extended credit. It is for this reason that Moody’s Analytics suspects one out of every eight businesses (primarily small businesses) to fail as a result of the slowdown.
Other factors include a global reorganization of supply chains. In short, most economists expect shorter global supply chains with fewer international dependencies. This may shore up the security requirements, but it will create higher prices for raw materials for goods all along a supply chain, ultimately increasing end prices for many items.
These factors, and more, will be evaluated in the months to come.