Press play to listen
The volume of retail sales – the value of the sale of goods adjusted for domestic inflation – rose a faster-than-expected 14% mom in June. In level terms UK retail sales were only 1% below the average recorded in the 12 months to February 2020.
This points to a very rapid recovery in domestic economic activity, albeit from a low base, and tallies with the surge in the UK composite PMI from 30.0 in May to 47.7 in June. There were greater opportunities to spend on goods as the lockdown was eased and the government continued to pump billions into the economy to support employed and self-employed workers and those on benefits and ultimately disposable incomes.
The corollary is that global GDP growth likely accelerated from 1.8% mom in May and we stick to our forecast of record-high growth of 10% mom in June. This would still imply that GDP in June was at the level which prevailed in late-2009 and that UK GDP contracted a record 20.6% qoq in Q2 according to our estimates.
In “normal” times UK retail sales figures provide a timely but only partial picture of economic activity. The historical correlations between retail sales and the broader measure of household consumption and between retail sales and GPD have been quite weak.
This should come as no surprise as retail sales exclude the sale of services (three times as large as goods sales) and motorvehicles and spending abroad and thus only capture a fraction of household consumption, which in itself only accounts for “only” 60% of GDP.
But these are not “normal” times. In the run-up to and after the introduction of the lockdown on 23rd March, households’ opportunities to spend on services and cars and to travel abroad were very limited, while fixed investment dwindled. As a result retail sales (sale of goods in UK) and GDP contracted and recovered in lock-step in February-May.
The correlation between retail sales and GDP growth likely remained strong in June but may weaken slightly in July given the partial re-opening of the hospitality, leisure and entertainment service industries and pick-up in outbound travel from the UK. Over time this relationship will likely “normalise” but for now retail sales remain a timely and powerful indicator of headline GDP growth in the UK and a focus for market participants.
UK retail sales and overall economic activity are recovering quickly – from very low base
The seasonally-adjusted volume of UK retail sales rose 14% month-on-month in June, a record monthly increase which was far greater than the consensus forecast of 8% mom (see Figure 1). The volume of retail sales – the value of the sale of goods adjusted for domestic inflation – was up 28% in June from April, the first full month of the UK national lockdown when retail sales slumped to their lowest level since December 2005. As a result the volume of retail sales in June 2020 was only 1% below the average recorded in the 12 months to February 2020, according to our estimates.
This rebound in retail sales points to a very rapid recovery in domestic economic activity, albeit from a low base, and tallies with the surge in the UK composite PMI from 30.0 in May to 47.7 in June which was in line with our expectations (47.0) but far greater than expected (41.0) – see Figure 2 and United Kingdom: Anatomy of economy on lockdown life-support (18th June 2020). There were greater opportunities to spend as the lockdown was eased and the government continued to pump billions into the economy to support employed and self-employed workers and those on benefits and ultimately disposable incomes.
The corollary is that global GDP growth likely accelerated from 1.8% mom in May and we are sticking to our forecast of record-high growth of 10% mom in June (see Figure 2). If correct this would still imply that GDP in June was at the level which prevailed in late-2009 and that UK GDP contracted a record 20.6% qoq in Q2 according to our estimates.
Retail sales are timely data release but historically have not correlated closely with GDP
In “normal” times UK retail sales figures in isolation provide a timely but only very partial picture of UK economic activity. However these are not “normal” times and retail sales have been and are likely to remain central to the current recovery in economic growth.
Analysts and financial markets pay close attention to the monthly release of retail sales data in the UK (and for that matter in most economies) as they are a timely proxy of household consumption and by extension GDP. The Office of National Statistics typically releases retail sales figures between the 17th and 24th of the following month (July figures are due to be released on 21st August). In comparison UK household consumption data are released only quarterly as part of the first estimate of GDP – for example Q1 2020 household consumption data were only released on 13st May, by which point retail sales figures for January-March were already available. Moreover, the ONS typically releases monthly GDP data, which do not include an expenditure-side breakdown, with a six-week lag – i.e. about three weeks after the corresponding monthly retail sales figures.
However, the correlation between quarterly retail sales and quarterly household consumption has historically been quite weak, with an R-squared of just 0.2 for data since Q2 1997 (see Figure 3). In Q1 2020, the latest data point available, household consumption growth was far weaker than implied by retail sales growth.
The historical correlation in the past decade has only been slightly stronger, with a still modest R-squared below 0.4 (see Figure 4).
The historical correlation in the past ten years between quarterly retail sales growth and GDP growth has been even weaker, with an R-squared below 0.2 (see Figure 5), while the relationship between monthly retail sales and GDP in the ten years to mid-2019 was virtually non-existent (see Figure 6).
This should come as no surprise given that retail sales only capture a fraction of UK household consumption, which in itself only accounts for “only” 60% of GDP (see Figure 7).
For starters, UK retail sales – which amounted to £440bn (20% of GDP) in 2019 – do not include the sale of motor-vehicles (£53bn in 2019).
More importantly, retail sales do not include the sale of services which last year amounted to about £1.3trn (60% of GDP) or three times as much as the sale of goods according to our estimates. This is not unique to the UK – in most developed economies the service sector is far larger than the goods sector. However, the ONS typically releases service sector data (including the Index of Services figures and Monthly Business Survey turnover of services industries) the same day as monthly GDP data (i.e. three weeks after retail sales figures). As a result analysts and financial market participants tend to ignore their release.
Finally, UK retail sales measure the sale of goods to UK and non-UK residents (e.g. tourists) whereas household consumption excludes the sale of goods to non-UK residents but includes UK-residents’ spending abroad (see Figure7). This definitional difference can prove significant given the size of British inbound and outbound tourism. In June-September 2019 UK residents made over 32 million visits abroad, more than twice the number of overseas residents’ visits to the UK (14.9 million), according to the ONS.
Retail sales critical to UK growth story in recent months
However, Figure 8 shows a strong historical, positive correlation between monthly retail sales growth and GDP growth in the past 12 months, with an R-squared of almost 0.8. In February-May 2020 the correlation had an R-squared of about 0.93, according to our estimates. From this historical relationship, the correlation between the UK Composite PMI and monthly GDP growth (see Figure 2) and other macro indicators we derive our GPD forecast for June 2020 of 10% mom.
The volume of retail sales and GDP fell and rose in almost perfect unison during these four months mainly because in the months prior to and in particular following the British government announcing a national lockdown on 23rd March the opportunities for households to spend on services and motor vehicles at home and to spend abroad were extremely limited, in our view. Moreover, the opportunities for companies to invest in the UK were also negligible.
The key hospitality, leisure and entertainment industries remained shut until recently and most parents were forced to stay at home to look after children with schools closed throughout May, which contributed to the 23.3% year-on-year collapse in the turnover of the services sector in March-May (see Figure 9). In comparison the value of retail sales fell “only” 13.6% yoy.
Moreover, the sale of motor vehicles shrunk 65% yoy in March-May to only 279,252 based on SMMT data (see Figure 10). Finally, international travel and therefore UK residents’ spending abroad (and non-UK residents’ spending in the UK) slowed to almost zero during that period. International air passenger traffic to and from UK (international) airports was down 84% yoy in March-May according to the UK Civil Aviation Authority (see Figure 10).
As a result the relative size of the consumption of services, purchase of motor vehicles and spending abroad – critical component of household consumption – fell relative to that of retail sales, as did other key components of GDP growth, including fixed investment spending. The corollary was that, in relative terms, retail sales along with government spending became far bigger drivers of UK economic growth and that the until-then-modest correlation between retail sales and GPD was materially boosted.
We suspect that the correlation between retail sales and GDP growth remained strong in June but that it may weaken slightly in July given the partial re-opening of the hospitality, leisure and entertainment industries as well as a pick-up in outbound travel from the UK and thus spending abroad. Over time this relationship will likely weaken further and “normalise”, in our view. But for now retail sales remain a timely and powerful indicator of headline GDP growth in the UK and a focus for analysts, market participants and policy-makers alike.