Is COVID-19 the death-knell for physical retail? - Capital Economics
US Commercial Property

Is COVID-19 the death-knell for physical retail?

US Commercial Property Focus
Written by Andrew Burrell
Cancel X

Retail has been hit hard by the COVID-19 crisis and lasting changes to online spending will bring further pain. While our estimates suggest that the impact is likely to be less severe than structural change in offices, the rental outlook is expected to remain weak and a faster online transition will reinforce this. In the longer term, we expect the sector will continue to underperform, though how much will vary between markets and will be driven by both online migration and other market conditions. Overall, the UK is most exposed, along with a few European markets, while the US may also be vulnerable later in the decade.

  • Retail has been hit hard by the COVID-19 crisis and lasting changes to online spending will bring further pain. While our estimates suggest that the impact is likely to be less severe than structural change in offices, the rental outlook is expected to remain weak and a faster online transition will reinforce this. In the longer term, we expect the sector will continue to underperform, though how much will vary between markets and will be driven by both online migration and other market conditions. Overall, the UK is most exposed, along with a few European markets, while the US may also be vulnerable later in the decade.
  • Retail property was in trouble before the virus, with many locations seeing values under pressure well in advance. Lockdowns provided a further severe jolt, with widespread closures and consumers adapting by shifting to online channels. While we don’t think that online shares will remain at their current elevated levels, there is a strong argument for the virus giving a nudge towards greater use, particularly in areas where adoption has lagged such as groceries. And new living and working arrangements after COVID-19 may also favour shopping patterns which could further boost the growth of online sales.
  • But, despite the gloomy headlines, the lasting impact of COVID-19 on retail may be less severe than structural changes in other property sectors. This assessment reflects a number of factors. First, the move online is long established and so retailers have already had many years to adapt to the challenge. Second, unless there is a major technological shift to open up new online possibilities, the transition is likely to remain fairly gradual. Unlike our office analysis where the shift to more remote working is expected to have a significant hit, our US and UK retail scenarios indicate that faster online migration will suppress offline demand, but the impact is fairly modest.
  • That said, in most markets, retail fortunes are expected to remain poor and a faster online transition will only reinforce this. The sector has been hard hit by the initial virus shock. In the next year or so, rents are likely to be hit as much by virus-related measures, notably local restrictions, social distancing and travel bans, as online trends. In the medium term, we expect the sector will still underperform, though how much will vary between markets and will depend both on online migration and on other underlying conditions.
  • We remain most concerned about the UK, where retail has already been hard hit and conditions suggest further pain ahead. This reflects weak underlying sales growth combined with our expectation of accelerated online leakage. But a historic legacy of over-capacity is also important, as is a CVA process that will work to hasten the correction. From our analysis, it is difficult to see any return to rental growth in the UK over the next five years, or to be hopeful of much improvement over the rest of the decade.
  • In the US, by contrast, after the current downturn, we expect that rents will recover, albeit sluggishly. Here, the macro background looks more helpful than in the UK, with a more upbeat outlook for consumer demand. Add to this a lower starting point for internet use and it is only after the middle of the current decade that we expect deeper problems. By then, we think slower in-store sales growth reinforced by over-capacity could weigh more heavily, though we are not projecting a sustained UK-style slump.
  • Our European analysis is more high level, identifying weakness rather than making firm predictions. Here, we would point to markets that looked to be struggling before COVID-19, as well as those that, like the UK, start from a higher base for online spending, which include Benelux, German and Nordic markets. A general problem is the weaker macro-economic environment in the euro-zone, which could also tip the balance in some cities. We will cover these specific market issues in future work.

Is COVID-19 the death-knell for physical retail?

Retail property was in trouble before the virus, with many locations recording weak rents and rising yields. As a result, retail capital values have trailed other sectors globally since well before this year’s lockdowns. (See Chart 1.)

Chart 1: Capital Value Growth (% y/y)

Sources: MSCI, Capital Economics

So, it was unfortunate that the sector was also most exposed to the immediate impact of virus restrictions, at least supermarkets and ecommerce aside. Many occupiers were forced to shut for a period and rental collection rates slumped. Despite a recovery in sales, the sector is still working through the damage and few retailers are likely to emerge unscathed. What’s more, the crisis has led to some dire warnings about prospects after the virus passes.

The main purpose of this Focus is to bring a longer-term perspective to the post-COVID-19 retail outlook. As any changes brought about by the virus will be long lasting, we will look at this not just over our normal five-year window, but into the next decade.

We concentrate on three aspects. First, and relatively briefly, we examine what we have learnt so far in the early stages of the crisis and how this shapes the immediate outlook. Second, and the heart of this paper, will be an analysis of how the crisis may have triggered longer-term changes in spending behaviour and how this could affect the retail outlook after COVID-19. This section will concentrate on the influence of UK online sales and the impact on in-store spending. In the third and final section, we will bring together the previous elements to establish the implications for our retail rental forecasts across UK, US and European markets.

The near-term challenges for retail

Our main concern is the long-term view, but understanding the starting point is crucial for this. Retail property came into the crisis weaker than other sectors. And most of what has happened since March has been more negative for the outlook. Admittedly, huge uncertainties still surround the recovery, but we expect retail to emerge by far the weakest of any commercial property sector for several reasons. (See again Chart 1.)

This summer’s strong rebound in retail sales across many developed economies was a surprise. The release of virus-related restrictions brought a v-shaped recovery, with volumes rising close to pre-pandemic levels. This was in contrast to other economic trends, notably the sluggish revival in industrial production. (See Chart 2 and our Global Economics Update.) As a result, our retail sales expectations for the next two years or so are slightly more upbeat than early on in the crisis.

Chart 2: Advanced Economy Retail Sales and Industrial Production (Feb. 2020 = 100)

Source: Refinitiv

But we remain cautious about the demand outlook for several reasons. For one, the latest data may give a distorted picture. It appears that there has been a marked substitution away from non-retail spending categories during COVID-19. Unable to buy leisure and travel services, consumers have spent more on groceries and other goods instead. It is likely that as normal patterns of demand are re-established, sales growth will slow sharply or even stall – in fact, there is already early evidence of this in some economies.

In addition, retailers care most about physical sales, while strong online demand has in part driven the current bounce. There was no choice in lockdown as many stores were temporarily shut. Since the summer re-opening, the online contribution has declined from its spring highs, but growth remains higher than it was pre-COVID-19.

Further ahead, much will depend on the path of the virus. Recent increases in cases and any subsequent tightening of restrictions are likely to be negative for retail and may reverse the shift back to physical stores. Further out, the removal of government support in many economies over the next year is also likely to weigh on spending. Overall, while the sales recovery has exceeded expectations, there is an underlying fragility. And the impact on physical stores will be less strong than the headline figures imply.

Living with the virus brings other lasting effects for retailers. Most important are likely to be distancing restrictions. These will reduce retail capacity over a one or two-year horizon and will potentially limit sales too. Of course, it is hard to isolate this impact from the more general economic damage. And the summer rebound in sales suggests that landlords and shops are already adapting to the challenges.

Distancing measures will also not hit all retailers equally. Open high streets or local centres appear best placed along with retail parks, particularly those anchored by supermarket or other essential stores. Restrictions will remain a concern for enclosed shopping centres, which have suffered most during the lockdown thanks to their heavier reliance on luxury and leisure. This suggests shopping centre performance will continue to lag in the recovery.

Distancing will also have a severe effect on bars and restaurants in the short to medium term. Where this is important in driving footfall to high streets and shopping centres this will worsen the other factors outlined above.

Another lasting legacy from COVID-19 is travel restrictions. These were eased in some countries during the summer, though a pattern of re-imposition of quarantines after the recent virus surge suggest that progress will be erratic. The impact of tourism on retail is complex. At the macro level, some countries and markets are more open to arrivals from overseas. (See Chart 3.)

Chart 3: Direct Contribution to GDP of Tourism and Hospitality (%)

Sources: WTTC, Capital Economics

There are also more specific concerns. For instance, prime high streets, which increasingly specialise in luxury stores, are more reliant on overseas demand. They have been hard-hit as tourism and business travel has dried up. Shopping centres and leisure linked to transport hubs have also been affected significantly by this and by the drop in day-to-day office travel.

The long-term impacts of COVID-19

Longer term, we assume that either through vaccine or built-up immunity, the immediate concerns of the virus will pass, and life will mostly return to normal. (See our UK Economic Focus.) But, while we think that growth in overall retail demand will revert to past trends, we do expect COVID-19 will have a lasting impact on the share of online spending. And with the virus providing a nudge to online migration, this will have consequences for in-store demand.

The inexorable rise in online spending is an established trend that was expected to continue before the virus. Since the late 1990s, double-digit rates of ecommerce growth have been common across most developed economies, resulting in the online share of total retail sales steadily increasing. (See Chart 4.)

Chart 4: Online Sales (as % of retail sales)

Source: Refinitiv

The net effect of this has been to squeeze the increases in spending within physical stores (or offline demand). (See Chart 5.) With overall UK sales growth also weaker in the post-GFC period, this has been bad news for retailers.

Chart 5: UK Retail Sales (% y/y)

Sources: Refinitiv

Over the next two years, online share trends will likely be noisy as economies grind out of their COVID-19 slump. But where this trend finally settles and how online growth resumes longer term will be critical for assessing in-store retail demand over the next decade.

To explore the consequences for retail property, we have created a number of long-run online spending scenarios. Each of these is justified by different assumptions on consumer behaviour and why this has changed post-COVID. None is fully representative of our current forecast view, which will also include many other factors and judgements. Instead, they are intended to isolate and quantify the possible impacts of different spending assumptions.

In this analysis, we concentrate on the UK, which has the longest running and most detailed ecommerce data. It is also regarded as a lead nation for online adoption. And there is clear evidence that it has been hurting its retail market for some years.

Scenarios for online sales

  1. Baseline assumptions (pre-virus path)

For comparison, we need a benchmark for how online spending would have evolved before the virus struck. In our forecasts, we previously assumed that as the online potential of different types of spending became saturated, overall rates of increase in online retail would slow further. This view was based on historic trends and a range of estimates by retail or property analysts. Unfortunately though, there was no consensus on the point at which the share would stabilise. This is critical, as a more rapid transition will mean more leakage of high street demand and more pain for retailers in the short term.

To model this, we used a time trend on pre-virus online growth that implies a continuation of the steady pre-virus slowdown. Chart 6 illustrates this path. Over the next decade, the share reaches a point where online and overall sales growth equalise and the internet share tops out at roughly 30% of all sales. This is a stylised view and inevitably the actual profile will be less smooth. But it is in line with our previous thinking and gives a clear benchmark for our COVID-19 impact scenarios.

Chart 6: UK Online Sales Baseline (% of sales)

Sources: Refinitiv, Capital Economics

In all of these scenarios, we assume an outlook for total retail sales that is consistent with our macroeconomic forecasts, in this case based on the UK long-term view. (See our UK Economic Outlook.)

2. Medium Impact scenario – earlier adoption of online channels

The outlook for online demand after the virus boils down to two key assumptions:

  • How much new online spending persists after COVID-19.
  • How far online penetration will rise longer term.

We think that changes in both are likely, but each will be driven by a distinct behaviour change and so we have run two scenarios to split the effects.

Our first (Medium) scenario deals with the impact of persistence post-COVID-19. In this, we assume online demand stays elevated across a number of categories over the next 2-3 years. As a result, the online share remains at a higher level after its COVID-19 spike for most of the decade. In essence, this brings forward about two years of “normal” online growth relative to our baseline. (See Chart 7.)

Chart 7: UK Online Sales Baseline and Scenarios (as % of retail sales)

Sources: Refinitiv, Capital Economics

For this scenario, we assume that ecommerce penetration will stabilise at a similar long-term level of about 30% share after 2030. This is because we think the virus may accelerate the uptake of certain types of spending, but as the benefits of using online are not necessarily changed, we assume a similar long-term peak to the baseline.

There are good reasons to think that some of the recent substitution will be sticky and that online shares will stay higher. Within a sharp overall increase in online sales during lockdown, for instance, food, household and other goods categories (including pharmacy, newspapers and books) rose fastest. (See Chart 8.)

Chart 8: UK Online Sales by Store Type

Source: Refinitiv

empty cellempty cellempty cell

For some of these faster-growing categories, the online share had previously been lower than average. This in our view supports the case for more of this spending to stay online after the virus. Many will have purchased books, music, electricals and clothes online already, whereas far fewer have ordered groceries or health and beauty items. We think this will bring an increased likelihood that these consumers will buy online again.

Admittedly, these categories include essential items such as food and medicine, where purchases can be less easily delayed and where online was often the only recourse in the spring. But this is balanced by the fact that the stores selling these items were also less likely to be forced to shut.

We think that the most interesting candidate for behaviour change is groceries. Historically, online food sales lagged other categories and the share of total spending has been low at around 5%. Of course, lockdown also removed some issues with delivering perishables with more workers at home. But given that online sales remain strong and that food accounts for over 40% of total retail sales, the shift here seems significant.

Low take-up of online grocery is not just a demand issue. Many supermarkets found it difficult to profit from deliveries, which have relatively high costs, and have not always pushed this option. So, during the lockdown when demand surged, retailers struggled to increase limited capacity. But supermarkets seem now to recognise online demand as a key source of future growth and are likely to invest more to prevent a repeat of these shortages.

High Impact scenario – structural changes bring a higher online share

In this second scenario, we assume that COVID-19 not only pushes forward some online demand as previously outlined, but also that online penetration reaches a higher level. This requires broader behaviour change to justify. Here, we think that major shifts in how we may live, work and play are most relevant.

Our recent Focuses highlighted the potential for dramatic demand shifts in the office sector over the next decade as companies and employees adapt to remote working. (See here and here.) At the very least, this will affect where people spend. As workers travel less to city centres, they will become more reliant on neighbourhood stores, and also critically on online purchases. And as workers will be at home more often, delivery becomes more convenient, which will remove obstacles to higher online penetration.

The COVID-19 crisis is also expected to lead to a rethink on housing. We have argued that city living will become less popular and there will be a migration to rural and suburban locations, where more space makes remote working more palatable. (See our Focus.) In practice, we think these shifts may be limited by relative price movements and affordability constraints. But even a small migration would bring more online spending, as new locations will likely be less accessible, and this drift will reinforce the remote working impacts.

More home working and less urban living appears to be bad news for most physical retail. Top destination centres and historic shopping streets may be sheltered. But stores that depend heavily on daytime worker populations, as well as anything linked to transport hubs is expected to struggle.

A changing balance between home and work makes a higher level of online demand more viable longer term. Given the number of factors involved, it is difficult to be precise. And the evolution of ecommerce may in turn be shaped by new technologies and new behaviours. But in our view, something like the highs seen during lockdown seem a reasonable assumption for our High scenario, with online spending topping out at about 35% of all spending in the UK. (See again Chart 7.)

It is also worth noting that this scenario by encompassing both the impacts of the initial disruption and expectations of structural change is probably closest to our current thinking.

Table 1 summarises the scenario results for total sales and offline spending. The baseline shows in-store demand much harder hit in the early COVID-19 period than overall retail spending. Thereafter, physical sales growth is positive, but subdued, for most of the rest of the decade compared with total sales. It is only after 2030 that demand fully recovers.

Table 1: UK Retail Sales Scenarios (Values, %p.a.)

Total Sales

Offline sales

Baseline

Medium

High

2009-13

3.1

1.9

2014-17

3.5

1.6

2018-19

2.8

1.4

2020-21

0.3

-1.1

-1.6

-1.7

2022-24

2.4

0.5

0.4

0.1

2025-30

2.5

1.3

1.5

0.7

2031-35

2.5

2.4

2.4

2.3

Sources: Refinitiv, Capital Economics

Our Medium scenario moves fairly closely with the baseline, albeit slightly weaker, but the higher impact case has a much slower recovery. In the long-term, all converge on a steady state as online demand moves into line with total spending growth. On the face of it, the differences in year-by-year growth rates look relatively modest. But the cumulative impact on physical retail sales is significant, with sales roughly 10% points lower on the High scenario by the end of the decade.

The impact of online growth on retail rents

The main reason for looking at retail sales in such detail is to try to understand how changes in online spending will influence rents. But even with a clearer idea of in-store spending, this analysis is by no means easy. The relationship between retail sales and rents has always been weaker than the key macroeconomic drivers for office markets, even after allowing for online migration. And since 2018, the link appears to have broken down completely. (See Chart 9.)

Chart 9: UK Retail Rents and Retail Sales (% y/y)

Sources: Refinitiv, MSCI

Further, unlike office markets, consistent data on physical measures for retail (vacancy, completions and stock) are harder to come by in European markets including the UK. This means that the detailed supply and demand analysis of our previous Focuses is hard to replicate. And there are other problems. High streets, shopping centres and retail parks are very different assets and hard to aggregate meaningfully. All of this makes high-level analysis more problematic than for offices.

Nonetheless, using our historical series for offline spending, we ran a regression analysis and derived some estimates for each scenario. We would caution against putting too much weight on the figures. The relationships are statistically weak, and these rounded numbers are intended only to give a broad direction of travel. (See Table 2.) Our current rental forecast, which doesn’t map strictly with any of these scenarios, is included for comparison.

These estimates suggest a more upbeat short-term rental outlook than our current forecast. But given the economic upheavals created by COVID-19, we would not read too much into these differences. And, more generally, over a five-year horizon, the results suggest that we may have been too optimistic about the recovery and hint that UK rents could decline for much longer.

The rental outlook also appears fairly bleak after 2025. In the High scenario, where offline demand remains more sluggish, the implication is that rents may continue to fall. Beyond 2030, online leakage should be less significant, as high street spending grows at a similar rate to overall retail demand. This suggests a return to rental growth rates of about 1-1.5% a year, still relatively unimpressive.

Table 2: UK Retail Rental Value Growth (% p.a.)

Current forecast

Scenarios

Baseline

Medium

High

2009-13

-1.7

2014-17

1.1

2018-19

-2.7

2020-21

-4.9

-3

-4

-4

2022-24

0.9

-1

-1

-2

2025-30

0

0

-1

2031-35

1

1

1

Sources: Refinitiv, Capital Economics

The results raise several points for our forecast. First, we have probably over-estimated the resilience of UK retail over the next five years and should re-examine our numbers. Second, the downside to UK retail rents is likely to outlast any immediate economic recovery. In the High case, which most closely reflects our current thinking on online spending, it looks like rents will continue to fall in the latter half of the decade as well.

Other factors are also undermining rents

Our estimates provide useful guidance for the post-virus future. They are, however, only calibrated on the basis of demand changes and the relationships with rents are fairly weak. It seems likely that online trends alone provide only a partial explanation of the movements in UK retail rents. In the immediate pre-COVID-19 period, for example, UK demand held up, but rents fell at a faster rate than at any time since the GFC. (See again Chart 9.)

It has also likely that supply factors will be important. For instance, it has been argued that years of strong growth before the late-2000s may have left an over-capacity in UK retail. This glut was exposed by the deep GFC recession and has been steadily undermined by the shift online. It is certainly true that the UK has a large retail sector. As a proportion of the commercial investment universe, it accounts for a third of total capital values. In most other developed economies, for instance, the share is a fifth or less. (See Chart 10.)

Chart 10: Retail Market Shares (%)

Source: MSCI

But it is not easy to quantify this overhang of supply or to isolate its influence on rents. Various retail vacancy estimates show that occupancy has fallen in retail over recent years, while generally rising elsewhere. (See Chart 11.) This has reversed a historical pattern of below-average vacancy in retail. But the change has been fairly gradual. And while vacancy never reverted to its pre-GFC norms, it did drift lower until fairly recently and has not yet reached its previous peaks.

Chart 11: UK Vacancy Rates (% of Floorspace)

Source: MSCI

But if there is a legacy of over-supply in UK retail, this will weigh on rents. It may also take a long time to correct. Even if development is weak, the unwinding of the excess will likely take years through change of use and stock withdrawal. This will worsen the rental response to stagnating physical demand. Pre-virus, we had been optimistic that the worst of the rental correction would be over by the early 2020s, but after the upheavals of this year, we are less convinced. (See our UK Commercial Property Focus.)

Other conditions may also weaken the UK rental outlook. For example, MSCI calculate that about 40% of retail properties are still over-rented, compared with about 30% of the all-property index. Given recent trends, this is perhaps unsurprising as market rents have fallen. But retail over-renting has been apparent for much longer, despite over a decade of poor performance. With a high proportion of rents above market rates and retailer margins threatened by online, further downward pressure looks inevitable.

Other factors have also exacerbated the UK rental adjustment, notably Company Voluntary Agreements (CVAs). Use of this insolvency procedure spiked in 2018, as retailers sought rent concessions from creditors. Levels dipped last year, but COVID-19 is likely to start a new, more severe wave if past downturns are anything to judge. That will not only impact on these struggling retailers, but also cause market ripples as others use CVA evidence to put pressure on their landlords. (See Chart 12.)

Chart 12: Retail CVAs

Sources: Insolvency Service

Bringing this together, it is clear that by accelerating the online transition, COVID-19 will contribute to another dismal decade for UK retail. While there is a high degree of uncertainty, our estimates suggest that our current view of a fairly quick recovery may be over-optimistic. In fact, it may be well into the next decade before rental growth returns to “normal”.

But we also think online sales are not the only problem, as UK markets face a toxic combination of weak demand, excess-capacity and over-renting. And the prospect of a cyclical surge in CVAs will exacerbate the rental correction over the next few years.

Of course, there may be mitigating factors. Some physical retailers have strong online offers and will benefit from the shift (or at least be hurt less). But if market-leading department store John Lewis is a guide, these companies are unlikely to expand their expensive physical networks. Instead, they will rationalise further and use more of what is left to deliver online purchases.

Implications for other markets

Although we have used the UK for our most detailed analysis, broad points can also be drawn about other global markets. So far, very few have seen as big a rental correction as the UK, but most have underperformed. In the US, for instance, there has been persistent gloom about prospects for some years. But the rental fallout had been modest before this year in either local (neighborhood) centers and larger malls (regional centers). (See Chart 13.)

Chart 13: US Retail Rents (% y/y)

Sources: REIS, Capital Economics

That said, the weak economic outlook this year and next is forecast to bring a rental downturn of a similar depth to that seen after the GFC. But after this cyclical upheaval, rental growth is expected to resume, albeit at a slow pace of around 1% a year. (See our US Commercial Property Outlook.)

In a recent Update, we analysed the US ecommerce data using a similar approach to our UK scenarios. The results are updated in Table 3. The baseline assumes a continuation of pre-COVID-19 trends, taking online penetration towards 25% by the 2030s. This is in line with pre-virus industry estimates.

There is also a High case, which embodies our post-pandemic assumptions about accelerated online take-up, home working and new living preferences. Here penetration rises to 30% over the next decade. (Note that US retail sales include auto spending and so online shares will be lower, so these figures are not directly comparable with the UK.)

Table 3: US Retail Sales (Values, %p.a.)

Total

Offline sales

sales

Baseline

High

2010-19

3.6

2.9

2.9

2020-21

2.2

0.9

0.7

2022-24

4.3

3.0

2.7

2025-30

3.0

1.8

1.3

2031-35

2.9

2.2

1.5

Sources: Refinitiv, Capital Economics

On these estimates, the outlook for in-store demand in the US is somewhat stronger than for the UK, which in part reflects more robust total sales growth. Our US consumer outlook is relatively upbeat into the medium term and this helps support underlying store activity. On balance, given the past relationships between rents and retail sales, this evidence alone doesn’t suggest that we need to downgrade our current US forecasts. (See Chart 14.)

Chart 14: US Offline Sales (High) and Retail Rents (% y/y)

Sources: REIS, Refinitiv, Capital Economics

After 2024, however, US consumer demand is expected to be weaker and the economic background is less supportive. This looks likely to be the period that US retail is most exposed and may begin to see some of the problems that currently bedevil the UK. In our High case, for instance, offline sales growth sinks below 1.5% p.a., much lower than seen in the recent past.

While we are not convinced that a concerted UK-style rental contraction is likely, the US is another market where excess supply may be an issue. Vacancy rates show a similar pattern to the UK, for instance, with retail rates more elevated in the last cycle and drifting upward in recent years, despite weak completions and a strong economy. (See Chart 15.) While we think our short to medium-term view is about right, there may be a need for more caution when pushing predictions later into the 2020s.

Chart 15: US Vacancy Rates (% of Floorspace)

Source: REIS, Capital Economics

A detailed multinational European analysis is outside the scope of this paper. Not only is each country at a different stage of online progress, but our prime data is city based and fitting this to national retail sales trends will be less easy. In addition, Europe’s prime high streets have become increasingly specialised in luxury goods, which may further weaken any correlation between rents and national retail spending.

Chart 16: Retail Rents (%, Cumulative 2018-19

Sources: MSCI, Capital Economics

Nonetheless, while we have not done a full re-forecast, there are some general points on the impact of online developments in European markets. (See also our European Property Focus.) For one, some locations experienced falls in rents similar to the UK before this year, notably in the Nordics, Netherlands, Belgium and to an extent Germany. (See Chart 16.)

This is not proof of vulnerability, but it shows that even when economic growth was positive, retailers were struggling in these cities. Of course, most markets are expected to see lower rents this year, though this is driven by COVID-19 and lockdowns. So, the pre-virus casualties seem to be a better starting point for further investigation.

There are consistent internet penetration measures in Europe using ecommerce turnover, though they are not strictly comparable to our previous US and UK analysis. (See Chart 17.) These suggest that in Germany and France online migration has not been that different from the UK over recent years, while it looks like Sweden is ahead on these metrics.

Chart 17: E-commerce (% of Turnover)

Source: Eurostat

On face value, these lead markets could be more open to a post-virus acceleration in online growth. But as we have highlighted, current levels are not the whole story in assessing the outlook for physical sales. What matters is the expected rate of increase of ecommerce and critically where online sales eventually top out. And as these reflect many factors, not just economic, but also cultural and social, this analysis requires a country-by-country focus. Nonetheless, as there is some overlap with the markets that had already been having problems before the virus, such as Netherlands and Germany, it suggests that we should look at these first.

One final aspect to note in Europe is macroeconomic performance. With economic growth rates in the euro-zone trailing the UK and US long term, the base for retail growth is also likely to be lower. This has the potential to amplify the impact of the online shift on physical demand, most obviously in laggards like Italy and Spain, but also in other core economies like Germany and Benelux. Of course, low growth is nothing new. It was also apparent for much of the 2010s and did not prevent euro-zone retail seeing respectable rental growth. But this factor may tip the balance over the next decade in some markets.

Conclusions

Despite the gloomy headlines, the lasting impact of COVID-19 on retail may be less severe than structural changes in other property sectors. This assessment reflects a number of factors. First, the move online is long established and so retailers have already had many years to adapt to the challenge. Second, unless there is a major technological shift to open up new online possibilities, the transition is likely to remain fairly gradual. Unlike our office analysis, our US and UK scenarios indicate that faster online migration will suppress growth, but in most cases the impact is fairly modest, even under the more extreme assumptions on behaviour change.

That said, in most markets, retail fortunes are expected to remain poor and a faster online transition will only reinforce this. The sector has been hard hit by the initial virus shock. In the next year or so, rents are likely to be hit more by special virus-related measures, notably local restrictions, social distancing and travel bans, than online trends. In the medium term, we expect the sector will still underperform, though how much will vary between markets and will depend both on online migration and on other market conditions.

Of all the locations surveyed, we remain most concerned about the UK, where retail has already been hard hit and where conditions suggest further pain ahead. This reflects weak underlying sales growth combined with accelerated online leakage in our scenarios. But a historic legacy of over-capacity is also important, as is a CVA process that will work to hasten the correction. On the basis of our analysis, it is difficult to see any retail rental growth in the UK over the next five years, or to be hopeful of much improvement over the rest of the decade.

In the US, by contrast, after the current downturn, we expect that rents will recover, albeit sluggishly. Here, the macro background looks more helpful than in the UK, with a more upbeat outlook for consumer demand. Add to this a slightly lower starting point for internet use and it is only after the middle of the current decade that we expect deeper problems. By then, we think slower in-store sales growth reinforced by over-capacity could weigh more heavily, albeit that we are not yet projecting a sustained UK-style contraction.

Our European analysis has necessarily been more high level, identifying weakness rather than making explicit forecasts. Here, we would point to markets that looked to be struggling before COVID-19, as well as those that, like the UK, start from a higher base for online spending. A more general problem is the weaker macro-economic environment in the euro-zone, which could also tip the balance in some cities. But given the number of specific factors, more detailed forecasts require a country-specific analysis, which we will address in future work.


Andrew Burrell, Chief Property Economist, +44 7985 902 151, andrew.burrell@capitaleconomics.com