The Data Has Gone From Bad To Worse

The data flow reflecting the economic impact of measures to curb the spread of COVID-19 has gone from bad to worse. We outlined some of the extraordinary data movements in our previous global macro report on March 30 (1). Since then, our downside fears on the data have played out. On the broad themes:

• Services will be hit harder than manufacturing;
• Discretionary consumer spending will be hit harder than spending on necessities; and
• Smaller business will be hit harder than larger ones.

Moreover, lockdowns and social distancing constraints now look to be in place longer than expected, which will cause a much sharper decline in activity than previously thought. In light of these developments, we have further lowered our 2020 growth outlook and raised our 2021 forecasts to partially compensate. More on the recovery path appears below.

More recent and broader-based high-frequency indicators now confirm an imminent very sharp decline in activity. Purchasing manager indices (PMIs) for March show post-global financial crisis (GFC) lows (see chart 1). Social mobility restrictions have had an outsize effect on airlines, restaurants, hotels, and cinemas. The resulting decline in the global services PMI was the largest ever recorded. Surveys from individual countries and regions tell similar stories.

Chart 1

Manufacturing PMIs have also declined sharply, albeit less than in the GFC. This reflects slumping demand, personnel shortages, supply constraints, and forced closures of nonessential businesses--all limiting production capacity. The global decline in March was less severe than in February, reflecting signs of stabilizing output in China after a record slump. Elsewhere, manufacturing PMIs are following services PMIs lower. Separate indices showing falling orders suggest that more weakness lies ahead.

Within the services sector, the response to measures to contain the spread of COVID-19 has been highly uneven. The U.S. labor market provides a clear example (see chart 2). The drop in nonfarm payrolls in March (which captured only the partial, first-month effects of social distancing) fell disproportionately on the leisure and hospitality sectors. Although this sector accounts for 11% of total nonfarm employment, it saw 65% of the drop in payrolls in March, suggesting a "beta" of around six (65%/11%). No other major sector of the economy saw a beta higher than unity.

Chart 2

Policymakers have launched a fresh round of measures in the past few weeks as the data confirms a sharper than previously expected downturn:

• In the U.S., the Federal Reserve announced it will help bolster credit creation to small and medium businesses as well as state and municipal governments, through coordination with the Treasury (2). This will be achieved by purchasing loans and establishing liquidity facilities.
• In Europe, governments agreed to a three-prong approach, including European Commission support for short-term national working schemes, European Investment Bank lending guarantees for small and medium enterprises (SMEs), and European Stability Mechanism credit lines with reduced conditionality (structural reforms).
• In China, a range of targeting measures have been rolled out, but the broad stimulus can be quantified by looking at financial conditions, which have swung one standard deviation looser than year-end 2019. Also, the flow of credit to the nonfinancial sector has risen 4 percentage points of GDP in the past three months.

The results of these policy interventions have been broadly positive. Volatility is down, risk appetite has stabilized, equity indices have begun to recover, and market conditions remain relatively orderly compared with just a few weeks ago.

A key policy area, in our view, is the nexus between the SME sector and the labor market, which will be important for the recovery. If SMEs remain in business and employees retain their jobs, this will support a faster rebound once health conditions allow. Here, Europe and Asia appear to be more successful in keeping firms in business and workers on payrolls, albeit with reduced hours. In the U.S., jobless claims have skyrocketed, reaching more than 16 million in the past three weeks alone, and we see the U.S. unemployment rate heading well into the teens in the coming months.

Our 2020 Global GDP Forecast Goes Negative

As a result of the sharp deterioration in macroeconomic prospects, we have significantly lowered our GDP forecasts for the year (see table 1). We now see a global contraction of 2.4% before a rebound to growth of 5.9% next year.

The bulk of the lower forecast stems from changes to the advanced economies, and their spillovers--to each other and the rest of the world. The forecasts for the U.S. and eurozone were lowered by about 4 and 5 percentage points, respectively. This sharp downward revision stems from the following factors.

First, most countries have extended their social distancing restrictions (or their more extreme version: lockdowns). As we noted in our previous global update, a downside risk was that these restrictions would be extended, and that the effects on activity would be nonlinear. This risk has now materialized.

Second, exit strategies appear to be more gradual than expected. We see evidence of this in China, as well as some re-intensification of restrictions in countries where the infected rate has begun to rise again after appearing to be under control. The gradual lifting of restrictions implies a more gradual recovery.

Third, sharper slowdowns in each region or bloc interact with each other through the traditional trade and financial spillover channels, as well as health channels. These second-round effects pull economic activity down further.

U.S.  The longest U.S. economic expansion on record has ended with another record, the sharpest contraction in economic activity since World War II. COVID-19 has quarantined more than 90% of the U.S. population. Beyond the demand shock from social distancing, other developments are also dragging growth, including Boeing's suspension of 737 MAX production over safety concerns and the oil-price plunge tied to oversupply and the standoff between Saudi Arabia and Russia.

We now see the toll on GDP will be far more severe than we once thought--with the contraction showing up in the first-quarter figure and worsening substantially in the April-June period. We forecast a decline in real annualized GDP of 7.6% in the first three months of the year and 35% (annualized) in the second quarter, translating to a decline of 11.7% peak to trough.

While the recovery will kick into gear during the second half of the year, it won't be enough to offset overall economic losses caused by COVID-19. With businesses shuttered, we expect the unemployment rate to peak at 19% in May, likely the biggest jump in unemployment on record, going back to 1948. Now at the lower bound, the federal funds rate will remain near zero until sometime in 2023.

Europe.  Large European countries have extended their lockdowns--for instance, France and Italy prolonged them by three weeks to an eight-week scenario. This makes our previous assumption of six to eight weeks of lockdown look like a best case rather than a baseline. In addition, exit strategies from lockdowns appear to be much more gradual than thought initially. The EU recommends its members to have a stepwise approach at a one-month pace.

Economic activity is therefore unlikely to stabilize by the end of the second quarter, and external EU borders might not fully reopen in time for the summer tourist season. Moreover, the lockdowns don't affect the economy in a linear way: the longer they are, the more they affect investment decisions, bring the economic fabric at risk, and lower GDP prospects.

Finally, the external environment has worsened. We expect a deeper recession in the U.S., which is the main destination for European exports. Also, a lastingly lower oil price might affect demand in OPEC and Russia, which together account for more than 7% of European goods exports--half as much as the U.S. This will undermine some of the benefits that Europe derives from lower inflation associated with lower oil prices.

We expect 2020 unemployment to be higher (8.3% for the eurozone and 5.8% for the U.K.) than previously assumed. A longer contraction in activity might end in lower employment, once governments have removed the short-time working schemes. This might also come from difficulties of the short-time working schemes to cushion less well-protected workers dependent on temporary and seasonal contracts in the hardest-hit sectors like tourism, restaurants, and hotels.

Asia-Pacific.  The climb back from the COVID-19 hit has already begun in some parts of Asia-Pacific, but the progress is likely to be slow with sporadic setbacks. China has passed the first-wave COVID-19 peak and has gone some way to restarting its economy. Industrial sector operating capacity is around 90% of normal levels as workers steadily return from their home provinces. The service sector is opening at a slower pace, hampered by social distancing rules that look set to remain at least for now. Policy stimulus has been building and should provide a tailwind for ongoing recovery through the second half.

Elsewhere, no economy has been spared the large cost of at least one month of stringent social distancing, which means large hits to growth and, in some economies, deflation. Together with the spillovers from weak demand in the U.S., Europe, and China, relative to our pre-COVID-19 forecast we now expect a hit to 2020 growth of 4-6 percentage points, bringing Australia to -5.7% and Japan to -3.6%. Among emerging markets, we now expect both India and Indonesia to grow by 1.8% this year, previously unthinkable numbers for two high-speed economies. Still, the downside risks of an imminent balance-of-payments crisis have reduced as the squeeze in the U.S. dollar subsides.

Emerging markets (EMs).  Our more pessimistic view on the U.S. and the eurozone will translate into a larger shock to external demand for EMs. This is particularly true for Mexico, which is exposed to the U.S., and Poland and Turkey, which are exposed to the eurozone. For Turkey, the hit to international travel is also relevant. In addition, we now expect a much deeper downturn in EMs owing to more stringent containment measures and their effects on domestic demand as the virus has spread widely across key EMs, and most countries. Even those where the number of infections is still not very large have implemented widespread containment measures.

The weakest quarter for most EMs in EMEA (Europe, the Middle East, and Africa) and Latin America will likely be April-June. Thereafter, we expect an uneven recovery, depending crucially on the policy response. Economies that have been quick to implement social distancing polices, and have complemented those with robust economic stimulus, should see stronger recoveries (for example, Chile). Generally, countries with delayed public health responses and limited stimulus programs will see weaker recoveries (Mexico). Brazil is somewhere in the middle, while Russia has been hit by two shocks--falling oil prices and COVID-19.

The Contours Of The Recovery

While the near-term data flow will be bleak, infection curves are flattening and there is now some light at the end of the tunnel. As a result, the attention of policymakers and markets has turned to the recovery path. How quickly will growth rebound? Will we get back to the pre-COVID-19 output path, or move to a lower one? Will potential growth take a hit?

In our view, the answers to the recovery path question begin at the end of the process. The potential growth path of the economy--to which it must ultimately return--is determined solely by the supply side (3). Deviations from that path come from the demand side of the economy or from transitory supply factors. It might also take into account sectoral changes, say more resources for health security, fewer for casinos. Once we determine the post-COVID-19 potential GDP path, we then derive the recovery path by connecting the nadir to the new potential path.

What determines the new, post-COVID potential output path relative to the pre-COVID path? We consider three possibilities (see chart 3).

Path A: Back to the original path--demand shock only.  The components of potential growth, and therefore the supply side of the economy, are unchanged by the COVID-19 episode. This means the economy will return to the pre-COVID path after the adjustment period ends (that is, when the output gap closes). The loss in output can be measured by the area between the path A scenario and the old baseline path.

Path B: Shift to a lower, parallel path--demand shock with loss in the level of output.  The growth of labor, capital, and productivity remain unchanged from the pre-COVID period, but the level of one or more of these factors is lower. Examples include lower human or physical capital due to obsolescence as a result of the changing composition of growth, or a step-down in productivity. The implication is that while the potential rate of growth post-COVID is the same as pre-COVID, the level of potential output is permanently lower. As an example, this happened after the 1990 U.S. recession.

Path C: Shift to a lower, shallower path--demand shock with loss in level and rate of output growth.  Along this path there is both a lower level of potential GDP as in path B, and a reduction in potential GDP growth. The latter would likely be the result of slower productivity growth, as happened in the U.S. following the GFC. Path C diverges from the old baseline owing to new, lower potential growth as well as partial, proportional convergence. Note that this path does not recover the year-end 2019 level of GDP in the forecast period.

Chart 3

We can put some numbers around these scenarios to give a sense of the likely volatile growth ahead. Let us consider a U.S.-style economy with a 2% annual potential growth rate, or 0.5% per quarter. Output is indexed to 100 at the end of 2019, when the economy was at full capacity so the output gap was zero. Thus, the economy in the pre-COVID quarter is at its textbook "steady state."

Assume the COVID shock happens in the second quarter only and results in a 10% drop in activity in that quarter (or 34% annualized) (4). If the recovery period is six quarters (for simplicity), then we reach post-COVID-19 at the end of 2021. Along the recovery path, growth rebounds to over 13% in the second half of 2020 in scenario A, and over 10% in scenarios B and C (5). The point here is not precision; rather, we will see wild swings in growth in both directions over the next year or more as the economies bounce back from the COVID-19 shock, even if the recoveries are, in some sense, incomplete.

Still Lots Of Risks To Consider

The balance of risks to our baseline remains on the downside. The recovery trajectory in our baseline is unidirectional, albeit at various speeds and duration, and with differing levels of completeness relative to the pre-COVID path. While there is a range of possible recovery paths, implicit in these paths is the assumption that the return to normal of the health situation, the return to normal of household and firm behavior, and the return to normal of policy stances are relatively smooth, with no backtracking (6). That need not be the case. There are both health and economic events to consider in looking at alternative, downside scenarios.

Uneven health recovery.  The health situation could improve more slowly than assumed, or move in an oscillating pattern. This would have direct effects on social distancing policies and, therefore, activity (recall that some pandemics unfold in waves). The result could be a "W" pattern for the recovery and a slower and choppier path to normalization with more lost output relative to the original baseline. More ominously, we may not get a vaccine or treatment in the forecast period, meaning a return to normal might be impossible.

Premature austerity.  Another risk is that, in light of the large rise in gross public debt (7) in response to the COVID-19 recession, fiscal austerity is implemented before private demand has sufficiently recovered. This is precisely what happened following the GFC once output started to recover. This will delay the return to potential growth and employment and result in lower-than-desired inflation and additional lost output.

Mismeasuring output gaps.  As noted above, in some scenarios the economy won't return to its pre-crisis path of potential output. This path is unobservable and difficult to gauge, at least in the short run after a major downturn. Therefore, policymakers may have the wrong estimate for the path and, therefore, the wrong policy settings.

For example, if the path of potential GDP has shifted downward as a result of the COVID-19 shock, the economy may actually be on the new path, but the authorities may think the old path is their target. In this instance, they may be running more stimulatory policies than necessary, raising inflation and issuing more debt than necessary, which would require future tightening, and the possible associated rise in volatility and a slowdown in activity.

Exit path for extraordinary measures.  As the recovery matures and potential output is regained, the monetary and fiscal authorities will need to exit the extraordinary stances put in place to combat the crisis. As post-GFC experience has shown, normalizing central balance sheets requires clear communication of objectives and timing. Similarly, the fiscal authorities will need to be clear on their exit from extraordinary labor market and SME policies in order to let the market, rather than the government, determine economic outcomes. Both of these could affect the path of the economy beyond the recovery period.

All told, while not precluding a faster-than-expected recovery, in light of the factors above, we would place the balance of risks on the downside.

Endnotes