In the near-term, we think that the ongoing consequences of CVAs and the large share of over-rented property will exacerbate the structural reduction in demand for retail property, causing rental values to fall sharply. But, with consumer demand expected to hold up and retail construction subdued, we think that there is scope for rental values to rebound as early as 2022, sooner than the consensus expects.
- In the near-term, we think that the ongoing consequences of CVAs and the large share of over-rented property will exacerbate the structural reduction in demand for retail property, causing rental values to fall sharply. But, with consumer demand expected to hold up and retail construction subdued, we think that there is scope for rental values to rebound as early as 2022, sooner than the consensus expects.
- The current retail rental downturn has been driven by a structural shift towards online spending. This has meant that, with fewer sales made in store, the current stock of retail properties is too high. It has been exacerbated by profitability concerns of retail tenants who have pushed for rent cuts and closed stores, notably using Company Voluntary Arrangements (CVAs).
- Over the next year or so, we expect these pressures will continue to weigh on rents. For one, retail stores continue to close. With almost 40% of retail property over-rented, these are likely to be re-let at lower rents. Further, even though the growth of CVAs appears to have peaked, we don’t think that this will limit retail rental value falls, at least in the near-term. Indeed, CVAs lower overall market rents, which increases the incentive for other retailers (including profitable ones) to push for rent concessions. In addition, history suggests that, even if insolvency is avoided now, a high proportion of retailers will likely still fail in the future, implying further rises in vacancy.
- Nevertheless, the profitability concerns of retailers look set to improve. Admittedly, further increases in the National Living Wage will increase cost pressures. But the retail labour market appears to have loosened, which should reduce overall pressure on wages. In addition, falls in retail rental values will reduce retailer costs, both directly and indirectly through a reduced business rate burden at the 2021 revaluation.
What’s more, unlike previous retail rental value downturns, consumer demand has held up.
- This has led to reasonable rates of retail sales growth, which has supported rental values notwithstanding the oversupply of property. Unless there is a no deal Brexit, we think that an improvement in economic growth and continued increases in real wages will support retail spending and therefore rental values.
- That said, the structural reduction in the number of retail properties is likely to be slow, which will offset any recovery in the near-term. However, our forecast for retail floorspace based on the outlook for in-store retail sales growth and assumptions about floorspace productivity suggests that floorspace could start to reduce as early as 2022. Given that retail construction has been subdued, once floorspace reduces, solid demand will reduce vacancy, putting upward pressure on rents.
- Overall, we think that retail rental values will fall by more than the consensus expects in the near-term. However, further out we think that there is an opportunity for rental values to rebound sooner than expected. On balance, we think that retail rental value growth will average minus 3.8% y/y this year and next, improving to an average of 0.8% y/y between 2021 and 2023.
How prolonged will the downturn in retail rents be?
Changes in the way people shop have shifted the focus of retailers towards online platforms. In fact, in September 2019, around 19% of retail sales were made online, compared to less than 5% a decade ago. With fewer sales made in store, the current stock of retail property is greater than what is needed now.
This structural shift has been exacerbated by cost pressures on retail tenants, including higher rents, business rates and wages, which have led to a rise in administrations and the increased use of Company Voluntary Arrangements (CVAs).
Together, these pressures have meant that landlords have faced increasing vacancy and a wave of rent reductions. In response, MSCI all-retail rental values have fallen for the past five quarters. This Focus looks to assess how much longer rents could fall.
This rental downturn differs from past cycles
One way to provide insight into this question is to look at previous retail rental downturns. Notably, in these downturns, it took around seven years for real rental values to stabilise. (See Chart 1.) This suggests that the current downturn has further to run.
Chart 1: Real Retail Rental Values (Peak = 100)
Sources: MSCI, Capital Economics
However, this downturn differs from those of the past. For one, it has been more modest. So far, rental values have fallen by less than 10% from their peak, compared to falls of 16% over the same period in the previous two cycles.
In addition, previous downturns were driven by a recession-generated collapse in consumer demand. But in this cycle, consumer demand has held up. Instead, the root cause of the current downturn is the structural reduction in demand which has meant that the supply of retail property is too high. Given that this downturn is more structural than cyclical, how much longer could it last?
Retail rental values are not yet at their trough
In the near-term, retail rental value declines look set to worsen. For one, it seems likely that retail stores will continue to be closed. Indeed, assuming the same pace of store closures seen so far this year continues, there could be around 1,600 store closures by year-end. (See Chart 2.)
Chart 2: Retail Store Closures (000s)
Sources: Centre for Retail Research, Capital Economics
Admittedly, this number of store closures is lower than in 2018 and is not that high compared with history. But store closures haven’t been the only determinant of rental values this cycle. Another important factor has been the increased use of CVAs. These have involved negotiations between landlords and tenants which have led to rental concessions and sped up rental value declines.
That said, the use of CVAs by retailers this cycle appears to have peaked as retail sales have picked up. (See Chart 3.) However, as discussed in our Update, we don’t think that this will limit rental falls, at least in the near-term.
Chart 3: Number of Retail CVAs and Retail Sales
Sources: ONS, Refinitiv
For one, PwC research into past retailer CVAs has shown that more than half have resulted in another insolvency period. Therefore, even if insolvency is avoided now, a high proportion of retailers will likely still fail in the future. This means that vacancy is likely to increase further.
In addition, as has already been seen from Next, Primark and more recently John Lewis, creditworthy retailers also have an incentive to push for rent or service charge cuts because CVAs lower the overall market rent.
Even without CVAs, with almost 40% of retail property over-rented, as stores close or as leases expire, they are likely to be re-let at lower rents, if they can be re-let at all. As such, it seems likely that the share of vacant retail units will continue to climb above its Q2 rate of 14.2%, putting further downward pressure on rental values.
This is consistent with indicators of the health of the market. For example, the net balance of RICS surveyors reporting increasing retail inducements, such as rent-free periods, has been a reasonable predictor of turning points in retail rental values in the past. The latest survey suggests that retail rental value growth won’t improve in the next few quarters. (See Chart 4.)
Chart 4: Retail Inducements and Rental Values
Sources: RICS, MSCI
Demand is less of a drag this cycle
Further ahead, we are relatively positive about the outlook for retail rental values. The key reason for this is that, as mentioned earlier, unlike previous rental downturns, this cycle has not been driven by weakness in consumer demand.
In the past, outright declines in retail sales corresponded with a fall in retail rental values. (See Chart 5.) This reflects that the performance of retail sales indicates how able landlords are to increase rents. Indeed, if retail sales growth reduces so does retailer profitability. In some cases, this will lead to retailer insolvency and increases in vacancy, putting downward pressure on rental values. More recently, with the introduction of turnover leases, this connection between retail sales and rents has been further strengthened.
Chart 5: Real Retail Rental Values and Retail Sales Growth (% y/y)
Sources: Refinitiv, MSCI
But, since late-2017, retail sales growth has held up. And we expect it to remain solid over the next few years. Indeed, unless there is a no deal Brexit, an improvement in economic growth and continued increases in real wages are expected to support incomes and therefore retail spending. (See Chart 6.)
Chart 6: Retail Sales and Real Wage Growth (% y/y)
Sources: Refinitiv, MSCI
Admittedly, with the unemployment rate at near-historic lows, we don’t think that there is much scope for further increases in the pace of wage growth. But even as the unemployment rate increases, it will remain low by historic standards, so we also don’t think that wage growth will slow too much. This bodes well for retail rental values, notwithstanding the oversupply of property.
Cost pressures are set to wane
In addition, the cost pressures which have caused profitability concerns for retailers and exacerbated retail rental value declines look set to ease. In particular, the fall in retail job vacancies from their peak suggests that the retail labour market has loosened a little. (See Chart 7.) This change is not drastic but points to reduced pressure on wages.
Chart 7: Retail and Wholesale Vacancies (000s)
Admittedly, the increase in the National Living Wage (NLW) from £8.21 to £10.50 by 2025 will put further upward pressure on wages. In real terms, this equates to average growth of 2.3% p.a. over the next six years, compared to the real increase of 3% y/y in 2019. The number of NLW jobs is also set to rise, with the age eligibility increasing by 2025 to include those aged 21-25.
But British Property Federation research has shown that retailers are more sensitive to property costs, at least compared to the office sector. Indeed, rents and business rates account for a 52% of retail staff costs, compared to just 13% for offices. MSCI data show that retail rental costs have been falling for the past five quarters. This will reduce retailer costs, both directly and indirectly through a reduced business rate burden at the 2021 revaluation.
The retail stock will be slow to reduce
However, even as cost pressures on retailers wane, the structural reduction in demand for retail space will continue to weigh on rents. This reflects that supply side issues are typically slower to resolve as it takes time to convert, demolish or remove stock. Indeed, in the year to March 2019, retail floorspace continued to increase. (See Chart 8.) That said, as a share of total commercial property floorspace, there are signs that floorspace has stabilised at around 18%.
Chart 8: Retail Floorspace (March Years)
Determining how floorspace could change from here is no easy task. But it is encouraging that retail sales volumes have continued to grow as a share of total retail floorspace, even once excluding non-store retail sales from the calculation. (See Chart 9.)
Chart 9: Retail Sales Volumes Per Sq. Ft. (£)
Sources: Refinitiv, VOA, Capital Economics
We can estimate how retail floorspace might change using our forecasts for in-store retail sales and assumptions about retail floorspace productivity (or sales density growth). Specifically, we assume floorspace productivity continues at its 2012-2018 average pace of 1.7%, as the point where the shift towards online spending started having an impact on the market. This calculation suggests that floorspace could start to reduce in 2022. (See Chart 10.)
Chart 10: Retail Floorspace (Bn Sq. Ft.)
Sources: Refinitiv, VOA, Capital Economics
Chart 10 also shows that this calculation will differ materially depending on the assumption about floorspace productivity. For example, if productivity grew at its 2002-2018 average pace, retail floorspace wouldn’t reduce over the forecast horizon.
However, in the current environment where retailers are under pressure to make the best use of the space they have, we think that it is unlikely that productivity will reduce back to its historical pace. As such, we think it is reasonable to assume that retail floorspace will reduce within the next five years.
This calculation also doesn’t take into account the future retail pipeline. But the evidence suggests that the development pipeline is likely to be limited. Indeed, the RICS survey shows that the net balance of surveyors has been reporting falling development starts since 2016. Further, since mid-2009, real new shop construction orders have been well below their long-run trend. (See Chart 11.) This suggests that a significant increase in retail completions is unlikely to be a risk to rents.
Chart 11: Real New Shop Construction Orders
(% Difference from Trend)
When will rents reach a turning point?
As a result of our estimates of the outlook for retail floorspace, we think that retail rental values could reach a turning point as early as 2022.
This is further supported by the fact that real retail rental values are already 13% below a simple estimate of their trend, more than any time over the history of the MSCI data. (See Chart 12.) This compares to real rents 30% above their long-run trend before previous downturns. And even over the next five years we don’t think that real rents will return to their previous levels.
Chart 12: Real Rental Values (% Above/Below Trend)
Sources: MSCI, Refinitiv, Capital Economics
Although this exceptionally low starting point may not prevent further falls, all else equal, it should limit the rate of decline. Indeed, with retail sales growth expected to hold up, low real rents will provide an incentive for tenants to expand, boosting occupier demand and, in time, rental value growth.
On balance, in the near-term, we think that retail rents will fall by more than the consensus expects. This reflects our view that, even though the use of CVAs appears to have peaked, we don’t think that it will limit rental value falls in coming quarters. We have pencilled in a cumulative decline in 2019 and 2020 of 8%, compared to the consensus of 6%.
Further out, much will depend on how quickly the retail stock is reduced. Our estimates suggest that, assuming retail floorspace productivity continues at its average pace since 2012, this could happen as early as 2022. And if we are right that consumer demand holds up, given that the construction of new retail space has been at near record lows for the past five years, it seems likely that retail rental values will also rise in 2022, and at a reasonable pace. This is sooner than the consensus expects. (See Chart 13.)
Chart 13: Retail Rental Value Forecasts (% y/y)
Sources: IPF, Capital Economics
Amy Wood, Property Economist, +44 20 7808 4994, email@example.com