After slashing its economic outlook in April and early May, expecting a GDP crash of as much as 40% in Q2, Goldman turned cautiously optimistic around the time of peak shutdowns in the US, and not only saw a light at the end of the tunnel in mid May, but also launched its own "reopening scale" index on May 9, a number which started off at 1 and is currently at 2 and rising with the bank continuing to see modest improvements at stay-at-home businesses, offset by a slight deceleration in back-to-business segments.
However, in recent weeks, Goldman's optimism has become increasingly more muted, and on Saturday the bank downgraded its Q3 GDP forecast due to a "dramatic resurgence of Covid over the last two weeks, with confirmed daily new cases surpassing 50,000." AS a result, Goldman's chief economist Jan Hatzius has "revised" his GDP forecast to reflect these new developments: "A pause in the consumer services sector in July and August should limit the 2020Q3 rebound to 25% quarter-on-quarter annualized, vs. 33% in our previous forecast."
However, in keeping with the traditional bad news-good news strategy, the bank also upgraded its 2021 numbers by 1-1½pp, both "because some reopening activity will be delayed and because prospects for an earlier vaccine have improved recently", in other words because of "hope." Still, on a full-year basis, Goldman's forecast now implies a -4.6% GDP drop in 2020, worse than the -4.2% before, and an unchanged +5.8% rebound for full year 2021, the same at the bank's prior forecast (expect many more downward revisions to this number).
Below are some additional details from Goldman which all other sellside banks will shortly imitate as they, too, begin downgrading their forecasts due to a "resurgence" in covid cases:
The US has experienced a dramatic resurgence of Covid over the last two weeks as a spike in the Sun Belt has pushed the daily number of new cases above 50,000. In response, officials have paused or reversed reopening in states containing more than half of the US population, and reopening will likely be delayed in much of the rest of the country as well. High-frequency alternative data show a dip in consumer activity in the worst-affected states in the final week of June, and this downward trend will likely persist in early July. In this week’s Analyst, we recalibrate our economic forecast to reflect the impact of the deteriorating virus situation on reopening policy, social distancing, and consumer activity.
The Covid Resurgence
Over the last few weeks, the Covid situation in the US has worsened significantly to the point where the US is now a notable outlier among advanced economies. The US is now performing poorly on most key measures: new cases are growing rapidly; prevalence of COVID symptoms is rising; the estimated effective reproductive number Rt stands at 1.10 nationally, meaning that case growth is accelerating; the positive test rate is well above 10% in some states; and available hospital capacity is diminishing.
In recent weeks we have tracked the virus situation at the state level to monitor the risk of a pause or reversal of reopening. We focus on four recommended gating criteria for reopening from the Centers for Disease Control and Prevention: 1) declining prevalence of COVID-like illness symptoms; 2) declining new confirmed cases; 3) a positive test rate below 10%; and 4) available hospital capacity above 30%.
Exhibit 1 shows that state performance in meeting these gating criteria has worsened significantly since late May. Although most states have a positive test rate below 10% and maintain adequate available hospital capacity, the number of new cases and the prevalence of COVID-like illness symptoms are rising in states containing most of the US population. The worrisome trend in these early indicators signals that lagging indicators like hospital capacity might deteriorate further in coming weeks. At this point, five states (representing 15% of the population) meet none of the recommended gating criteria and another 11 states (24% of the population) meet only one. Only three states meet all four criteria.
Goldman then points out that while "state governments have emphasized that a return to an economically painful full lockdown would only be a last resort", Goldman agrees that there is indeed a high bar for states to shut down large portions of their economies again, "diminishing available hospital capacity leaves state governments with little choice but to reassess reopening and potentially pursue incremental restrictions."
As the bank shows in the next chart, "over the past two weeks, states representing about 60% of the US population have responded to the worsening virus situation by pausing or reversing their reopening plans."
To wit, nearly all states had eased restrictions by June, but over the last week states including Florida, Texas, and California moved toward tighter restrictions with targeted measures including closing bars and limiting restaurant occupancy. Texas has also limited elective medical procedures to free up hospital capacity.
As Kostin further adds, whereas the US took a more bottom-up approach to reopening than most countries, with policy set mostly at the state and city level and as a result, there were bound to be setbacks in at least a few parts of the country as the economy reopened, "the recent virus news and the extent of reopening reversals have already been much worse than we anticipated, and further restrictions will likely be required in some states to bring the virus under control."
This, Goldman concedes, "calls for a reappraisal of the economic outlook."
What does this reappraisal mean in practical terms? Nothing less than a significant trim in expectations for a consumer comeback. Here's Goldman:
A combination of tighter state restrictions and voluntary social distancing in response to the virus situation is already having a noticeable impact on economic activity. Mobility data from Google show that states with a more severe deterioration in their public health situation—defined here as those states meeting none or only one of the gating criteria for reopening as of June 28—saw a modest decline in retail and recreation activity and workplace activity toward the end of June that began even before authorities tightened policy (see Exhibit 3). Activity was flat on average in other states.
Other timely alternative data sources also indicate that the consumer comeback has stalled as the virus situation has worsened. Mobility data from Apple show a pause in activity in the worst-affected states alongside a continued upward trend in the others, and Homebase data show declines in the number of businesses open and employees working, especially in states experiencing virus outbreaks.
To be sure, the recent declines are minor compared to the collapse in activity in March and April, but as Goldman warns, "they indicate a break from the steady upward trend since mid-April. With the virus situation still worsening in many states and additional restrictions likely to be required, the recent trend downward in the worst-affected states will probably continue in early July." Additionally, "bringing the virus under control is likely to take some time, and national consumer activity is unlikely to pick back up until that happens."
To incorporate these developments in our economic forecast, Goldman now assumes that the consumer comeback will pause in July and August—specifically, that the level of consumer services spending in those months will equal the June average.
We reach this forecast by first assuming that late June trends continue into early July and then taking a GDP-weighted average of activity levels across states.
Which leads us to the next question: can the US get back on track, and more importantly, when? According to Goldman, the current public health situation raises legitimate doubts about the near-term outlook, "but we think it would be a step too far to conclude that the US has reached or surpassed the peak level of reopening consistent with keeping the virus contained, and we do not think it is appropriate at this time to make major changes to the outlook beyond the next two months." Kostin takes this view for two main reasons:
- First, other advanced economies show that it is clearly feasible to resume economic life to or beyond current US levels without triggering a spike in virus cases. Exhibit 4 compares the US with a group of countries that also experienced a large first wave. The chart on the left shows that the current level of economic activity in the US puts it roughly in the middle of the pack, while the chart on the right shows that the number of daily new Covid cases per million people in the US makes it a major outlier. Similar economies have clearly found a more efficient way to balance reopening the economy and keeping the virus under control, and we think the US is likely to eventually find its way to a better approach too.
- Second, behavioral and policy changes offer fairly low-cost opportunities for the US to eventually achieve both a reduction in virus spread and further reopening of the economy. In particular, Goldman recently showed that mask wearing has a major impact on virus spread at negligible economic cost. It is admittedly hard to know how well the US will adapt in coming weeks. But last week Texas joined the roughly half of US states that have implemented a mask mandate, indicating that state authorities are willing to making politically controversial policy changes to address the current health crisis.
In summary, Goldman expects the US economy to get back on track in September although should the "resurgence" in cases accelerate, we expect this timeline to be shifted substantially, and the inflection point to take place some time after the November election.
* * *
As a result of these considerations, and until there is more clarity on the progress of covid in coming months, Goldman is revising its GDP forecast "to reflect the worsening of the virus situation and the likely resulting pause in the consumer recovery discussed above." The bank has made three updates to the sector-level GDP model that it uses to track the impact of Covid on the economy.
- First, the bank takes on board the latest official data for May released since its last forecast revision, especially the personal spending report, the construction spending report, and the durable goods report.
- Second, Kostin uses an array of alternative data sources already available for June to update his assumptions about the pace of recovery last month. Building on earlier work, the bank combines consumer transaction data from Second Measure and Opportunity Insights and consumer mobility data from FourSquare and SafeGraph into a composite measure that it uses to forecast June spending levels in the corresponding consumer services categories, as shown in Exhibit 5. The chart indicates that the pace of consumer recovery in the first two months of reopening, May and June, was quite strong. The banks also uses real-time data on activity at construction sites from OxBlue and the latest auto and aircraft manufacturing schedules to fine-tune construction and manufacturing assumptions for June.
- Third, the bank now pencils in a flat path of consumer services spending in July and August due to the deteriorating virus situation and reopening rollbacks in some states. While the bank's economists had already anticipated a sharp deceleration in the pace of recovery after June once the easy gains from reopening were exhausted, Kostin admits that "this is nonetheless a significant downgrade" from his previous forecast.
The chart below shows Goldman's revised estimate of the virus impact on the level of GDP by component. The solid red line shows its current estimate of the total impact of the virus hit, while the dotted red line shows the bank's previous estimate for comparison. The orange line shows the GDP forecast in levels by adding potential growth to Kostin's estimate of the virus impact. The forecast implies that real output will return to the pre-virus level by 2021 Q3, but a 1-2pp output gap relative to potential will remain at end-2021.
As noted above, the biggest near-term change relative to Goldman's previous estimate is the pause in the consumer services rebound in July and August, shown by the flat path of the dark blue bars in those months. However, the bank - leery of delivering too much bad news at one time - expects this to be offset by faster growth in 2021, primarily because some reopening activity will simply be delayed. Meanwhile, "recent improvements in the prospects for earlier delivery of a Covid vaccine also increase our confidence that 2021 will see strong growth. These offsetting revisions result in little net change to the level of GDP by end-2021."
The next chart shows that Goldman's forecast translates to quarterly annualized GDP growth of -5%/-33%/+25%/+8% for 2020Q1-Q4 (vs. -6.5%/-33%/+33%/+8% previously) and +8%/+6.5%/+5%/+4% for 2021Q1-Q4 (vs. +6.5%/+5%/+4%/+3% previously). Goldman's Q3 2020 forecast of +25% is still strikingly strong because of the statistical overhang created by the sharp upward trajectory from the April bottom through June—even if GDP were flat in Q3 at the June level, Kostin sees QoQ annualized growth rate still at +17%.
Additionally, Goldman's revised forecast implies 2020 growth of -4.6% on a full-year basis (vs. -4.2% previously) and -3.7% on a Q4/Q4 basis (vs. -2.6% previously), and 2021 growth of +5.8% on a full-year basis (vs. +5.8% previously) and +5.9% on a Q4/Q4 basis (vs. +4.6% previously).
In short, the bank sees the deterioration of the virus situation as primarily shifting growth from Q3 to 2021, although it cautions that "downside risks to our baseline scenario have clearly increased."
* * *
Of course, it's not just GDP that is impacted by covid, although when it comes to jobs, the picture is decidedly more optimistic. As Kostin concludes, he also adjusted the bank's forecast for the unemployment rate to reflect the large positive surprise in the June employment report. Following large declines in May and June, the official unemployment rate now stands at 11.1% and the “true” unemployment rate including misclassified workers stands at 12.3%, the latter down sharply from 19.5% in April and 16.4% in May. This strong rehiring momentum, coupled with the fact that over 80% of the newly unemployed since February still report themselves as on temporary rather than permanent layoff, points to substantial further progress in the labor market in the second half of the year. Indeed, despite the downgrade to Goldman's Q3 growth forecast, the bank now expects the official unemployment rate to reach 9% by the end of 2020 (vs. 9.5% previously), and to drop a further 2% to 7% by the end of 2021.