The consensus forecast is that GDP contracted in many major economies in Q1 and will most definitely contract in Q2 as a result of the negative impact of national lockdowns on supply and demand.
The implication is that these economies will be in recession – defined at a country level as two consecutive quarters of negative quarter-on-quarter GDP growth.
A number of major economies, including Japan, France, Italy and South Africa, already recorded negative GDP growth rates in Q4 2019 and may therefore have already been in recession in Q1 2020.
Quarterly GDP data are typically released with a lag of at least a month and so we turn to other key macro variables, released weekly or monthly, to ascertain the magnitude of the current economic downturn.
In the United Kingdom, the collapse in the Composite PMI in March gives credibility to the National Institute of Economic & Social Research forecast that GDP shrank 5.0% quarter- on-quarter in Q1. To put this in context the largest quarterly contraction during the Great Financial Crisis was “only” 2.1% qoq (in Q4 2008).
In the United States, the Atlanta Federal Reserve’s GDPNow tracker, a running estimate of the seasonally-adjusted annualised rate of US GDP growth in Q1 based on available data, stood at 1.0% qoq on 9th April.
This number will likely be further revised downwards following the release of other key March data, including retail sales and industrial output (on 15th April), durable goods orders (24th April), goods trade balance (28th April) as well as housing data.
The US Conference Board’s core scenario, which is certainly not the most bearish, is that US GDP will shrink 5.8% qoq (annualised) in Q1 and by a third in Q2 – three times more than at any other time since World War Two.
Country-specific recessions: A useful but ultimately crude yardstick
At a country level a recession is defined as two consecutive quarters of negative quarter-on-quarter (real) GDP growth – an easy definition to apply even if ideally these numbers should be seasonally-adjusted.
However, the majority of countries release even preliminary quarterly GDP data with a lag of at least one month. For example the US Bureau of Economic Analysis will only publish its advance estimate of US GDP growth in Q1 on 29th April, with its third and final estimate due out on 25th June – nearly three months after the end of Q1.
So we will only know with certainty whether countries were in recession in the first half of 2020 by end-June at the earliest, with the exception of those countries which had already recorded negative GDP growth in Q4 2019 and may therefore have already been in recession in Q1 2020. These include a number of major economies, including Japan, France, Italy and South Africa (see Figure 1). Mexico was already in recession in the second half of 2019, with seasonally-adjusted GDP having contracted 0.1% qoq in both Q3 and Q4 2019, according to OECD data.
United Kingdom GDP may have contracted 5% qoq in Q1 2020
Some countries, such as the United Kingdom and Russia, publish monthly GDP data which provide a timely insight into how GDP likely fared over the course of a quarter. However, these monthly data are also published with a lag. For example the UK Office of National Statistics only released February data on 9th
April and by its own admission “The monthly growth rate for GDP is volatile. It should therefore be used with caution and alongside other measures, such as the three-month growth rate, when looking for an indicator of the longer-term trend of the economy”.
According to our estimates, UK GDP growth was near zero in January-February (see Figure2). Given the sizeable contraction in economic activity in recent weeks, even before the government formally announced on 23rd March a nationwide lockdown, GDP very likely contracted rapidly in both March and Q1 2020. The National Institute of Economic & Social Research (NIESR) estimates that UK GDP growth contracted 5% qoq in Q1 2020. The largest quarterly contraction during the Great Financial Crisis was “only” 2.1% qoq (in Q4 2008).
Based on the historical, positive relationship between the UK Purchasing Managers Composite Output Index (commonly referred to as the “UK composite PMI”), which collapsed to 36.0 in March from 53.0 in February, and quarterly GDP growth, the NIESR’s forecast for Q1 2020 is feasible, in our view (see Figure 2)1. Assuming no revisions to prior data this would imply that GDP contracted about 15% mom in March. To put this in context between May 2008 and March 2009 monthly GDP growth in the UK averaged -0.6% mom and never contracted more than 1% mom.
A word of caution. The more timely release of monthly or even weekly macro data (including trade, industrial and construction output, factory orders, investment, employment, income, retail sales) and surveys (including Purchasing Manager, consumer sentiment and business confidence indices) can provide an early insight to forthcoming GDP figures. However, some of these monthly data series are volatile, exhibit strong seasonality or are subject to material revisions. They should thus be interpreted with care when assessing the implications for underlying GDP growth.
United States GDP forecast to have contracted faster in Q2 than at any time since WWII
The Atlanta Federal Reserve’s GDPNow tracker, a running estimate of the seasonally-adjusted annualised rate of US GDP growth based on available data for the current measured quarter, was relatively stable around 3.0% qoq up until a fortnight ago (see Figure 3).
1 The UK Composite PMI may still have under-estimated the collapse in economic activity given that the survey was conducted from 12th to 27th March and thus omitted the last four days of the month (when the UK was in lockdown).
The Atlanta Fed has since revised its growth estimate to just 1.0% qoq following the recent release of very weak US macro figures for March (including employment, jobless claims and ISM PMIs for the manufacturing and service sectors). Moreover, we anticipate that this number will be further revised downwards following the release in coming weeks of other key March data, including retail sales and industrial output (on 15th April), durable goods orders (24th April), goods trade balance (28th April) as well as housing data. US investment banks are forecasting a material contraction in US GDP in Q1 while the US
Conference Board’s core scenario is that US GDP will shrink 5.8%.
However, it is conceivable (even if unlikely) that US GDP growth narrowly avoided contracting in Q1 before collapsing in Q2 as a result of the negative impact on US economic activity, and particularly household consumption, of domestic lockdown measures, the sharp downturn in US equities and sharp contraction in global trade and economic growth (for a more detailed analysis of how developments in US equities affect consumer confidence and ultimately household consumption see “The key quartet: US income, confidence, net worth and consumption”, 18th October 2019). If the United States (and other major economies) relax lockdown rules early this summer and at least some employees return to work, US GDP growth could rebound in Q3, albeit from an extremely low base.
In this scenario the US will have avoided (albeit narrowly) a recession as defined by two consecutive quarters of negative GDP growth. Of course this literal interpretation would mask the fact that US growth may contract in Q2 by far more than at any other time since World War Two. US investment banks on average forecast that the US economy will shrink by about a third in Q2 in annualised terms, broadly in line with the US Conference Board’s core scenario. The largest contraction in US GDP since 1945 was -10.4% qoq (annualised) in Q1 1958 (see Figure 4).
