Housing Market Recovery Stronger Than Expected
Stamp Duty Holiday Introduced Too Soon?
Transaction & Housing Supply Scenarios
The housing market may have been all but extinguished from mid-March to mid-May but it’s now running hotter than a typical British summer and the government has thrown on more fuel in the form of a stamp duty holiday to keep the fire ablaze. What remains uncertain and concerning is what happens when the backlog of activity delayed by lockdown fades and the full effect of massive shocks from Covid-19 and Brexit feed through. This note looks at the limited evidence so far, questions whether the government have gone too early with their stamp duty holiday, and presents our illustrative scenarios for transactions and housing supply.
A Strong Recovery
The strong recovery in housing market activity was a surprise to many and, though the evidence is limited, it appears to be continuing. Data from MHCLG on the number of new build Energy Performance Certificates, which is usually a good leading indicator for new build activity, suggests it has returned to normal levels for this time of the year (Figure 1 below). Meanwhile, yesterday’s Rightmove release reported asking prices have risen 2.4% since March. It also showed sales agreed in the five days following the stamp duty holiday announcement were 35% above last year’s level. They had already been up 15% in June. The Rightmove house price index may just report asking prices so it can be easy to dismiss as just reflecting sellers’ optimism. However, it appears deals are being done and so the small price falls reported in the mortgage lender indices may show a bias towards lenders’ caution in this uncertain environment. Perhaps the reality on prices is somewhere in between the two.
It is uncertain just how long this recovery can last given the scale of the economic shock and likely rise in unemployment later in the year. However, for the time being, low mortgage rates will support high prices, the stamp duty holiday will support activity, forbearance will limit distressed sales and constrained mortgage lending criteria means only those with equity can buy. All that suggests housing market inequalities will be exacerbated and the medium-term outlook for transactions is lower than seen in recent years.
A Holiday Too Soon?
Given the strength of the current recovery in housing market activity, the timing of the stamp duty holiday is odd as it is broadly targeted at total activity rather than supporting specific buyer groups. As launched, it will add to the existing burst of activity over the summer and, if previous holidays are a guide, activity will then return to normal (whatever that is) during the autumn/winter until a last spike to beat the end of March deadline. The following collapse in transactions during Q2 2021 could therefore come at a time when the economy is at its weakest. Instead, it would’ve been better to introduce the holiday at the Autumn Statement, when unemployment is likely to be rising and the housing market flagging. Q2 2021 could be terrible for transactions.
We have constructed two scenarios to explore possible new housing supply trajectories over the next five years and the impact making different assumptions can have. As such, while we would be delighted if they were correct, these scenarios should not be considered forecasts of what will happen to housing transactions and new supply.
The current market is highly uncertain and there are a large number of important factors that will determine its future but our housing supply model is intentionally simple and uses total transactions as the main input. Therefore we have constructed two scenarios for transactions over the next five years across England and these are shown opposite. The OBR’s central scenario published last week is also shown (rebased by us to England) and highlights a more optimistic view of the future. Unfortunately we do not believe it is credible as it includes a spike in the second quarter of 2021 (after the end of the stamp duty holiday) and a rapid rise to a level of transactions not sustained since 2008.
Housing Supply Scenarios
The housing supply model is relatively simple in that it assumes the level of private completions excluding Help-to-Buy Equity Loan is linked to total transactions, affordable housing delivery is a combination of available government funding (assumed to continue at current levels) and S106 delivered housing is linked to private completions. The two most important variables are therefore total market transactions and Help-to-Buy completions. Check out our recent Housing Conversation for more on the past, present and future of Help-to-Buy.
Our “Recovery” scenario assumes housing market transactions continue their recovery during 2020-21 but then fall back in 2021-22 following the end of the stamp duty holiday. Transactions then return back to 2019-20 levels in 2024-25. Help-to-Buy is also expected to continue either in its current form or with a replacement that matches its recent levels of additionality. This scenario gives an average for net additional dwellings over the five year period of 208,000 p.a. though this includes a fall to 191,000 next year. This would just meet the government aspiration for one million new homes over five years but well below the previous 300,000 p.a. target.
The second “Downturn” scenario may appear much more negative than our first scenario but, aside from a sharper fall in the first year, is actually very close to what happened to net supply during the last recession. The scenario assumes a similar recovery in transactions over the summer to our “Recovery” scenario but they then fall during the autumn and into next year as the consequences of the economic shock and rising unemployment suppress housing market activity. Transactions then recover to 2019-20 levels over a six year period. The other big difference is that following the end of Help-to-Buy, we assume its replacement can only support half the number of completions. That depresses housing supply relative to the recovery in transactions. This scenario gives an average for net additional dwellings over the five year period of 150,000 p.a., well below optimum levels.