Lisa Panigrahi and Danny Walker
The average house in the UK is worth ten times what it was in 1980. Consumer prices are only three times higher. So house prices have more than trebled in real terms in just over a generation. In the 100 years leading up to 1980 they only doubled. Recent commentary on this blog and elsewhere argues that this unprecedented rise in house prices can be explained by one factor: lower interest rates. But this simple explanation might be too simple. In this blog post – which analyses the data available before Covid-19 hit the UK – we show that the interest rates story doesn’t seem to fit all of the facts. Other factors such as credit conditions or supply constraints could be important too.
Chart 1: Real house prices in the UK
Interest rates have fallen substantially in the UK over the past few decades. The rate on 10 year government bonds – the ‘risk free’ benchmark – is under 1% today, compared to 5% in 2000 and 12% in 1980. In real terms, the 10 year rate is around 7 percentage points lower than it was in 1980. Rates have fallen in many advanced countries, not just the UK. In fact, as Chart 2 shows, the fall in UK rates since 1980 is slightly less than the average across a sample of advanced economies.
Chart 2: Nominal long-term rates in advanced economies
In a standard asset pricing framework, a fall in risk free rates mechanically translates into a rise in asset prices. Discount rates are made up of a risk free rate plus a risk premium. This framework would predict large rises in the prices of many assets in response to the fall in risk free rates in recent decades. Given the fall in UK risk free rates from 1980 to today, a fictional asset that pays out a constant 9% of its price in cash every year and with a constant risk premium of 4%, would have quadrupled in price all else equal.
An influential academic paper from the 1980s showed that housing can be modelled like any other asset. The ‘user cost’ of owning a house should be equal to the rent. If prices were too high, there would be an incentive to rent, which would lead to a fall in prices. A recent post on this blog based on a paper by Miles and Monro (2020) found that the fall in rates over the past four decades, combined with the increase in incomes, might explain all of the rise in UK house prices since then.
But there are a number of facts that this framework doesn’t seem to explain.
The timing of the fall in rates and the rise in house prices doesn’t line up. Chart 3 plots the path of both variables for the UK. There are two periods that stand out. In the decade from 1990 to 2000, UK house prices rose by just 30% while rates more than halved, falling from 11% to 5%. Then from 2000 to the start of the financial crisis in 2008, house prices doubled while rates barely budged. Since the crisis rates have fallen to just above zero and house prices have risen by a further 20%. They have remained flat in real terms on average.
Chart 3: Nominal UK average house prices and nominal long-term rates
UK house prices have risen by more than in almost any comparable country. As Chart 4 makes clear, only three advanced economies – Australia, Norway and Spain – have seen a bigger rise in prices than the UK since 1980. At the other end of the spectrum, house prices in Japan are barely above their 1980 level. All of these economies have seen a similar fall in interest rates to the UK. If UK house prices had grown at the same rate as in the average advanced economy, they would be at half the level they are at now.
Chart 4: Real average house prices in advanced economies
The framework doesn’t account for house price growth in other countries. We have calculated what a user cost model would predict for house price growth in the UK and other countries since 1980. The model we use is the same as the one used this previous post and explained in some detail in this working paper, with a different data source. We input actual data on average rents and interest rates for each economy and hold all other factors constant over time. As Chart 5 shows, this model works relatively well for the UK, predicting only slightly higher price growth than we observe in the data. However, for countries like Germany, Italy and the US, the model predicts house prices of more than double the actual level in the latest data.
Chart 5: Predicted house prices in advanced economies vs actual prices
Other UK asset prices have risen by the same amount as in other countries. Given the growth in UK house prices, you might expect other UK asset prices to have risen a lot too. The most accessible risky asset class for households, apart from housing, is equities. Chart 6 compares UK equity prices to other advanced economies. Since 1980 equity prices have risen by almost exactly the same amount in the UK as in the average advanced economy.
Chart 6: Nominal equity prices in advanced economies
There has been large variation in house price growth across regions of the UK. Chart 7 shows how house prices have evolved at the local authority level in the UK over the period for which the statistical authorities produce data. London has seen the highest growth in house prices of any region. The City of London, Newham and Waltham Forest have all seen prices quintuple since 2000. In Blackpool, Burnley and the Isles of Scilly prices have risen by less than half as much over the same period. On a regional basis, house prices in the North East have grown the slowest. A previous post noted that rates alone would not explain this regional variation.
Chart 7: Nominal house prices in UK local authorities
The framework doesn’t account for differences in house price growth across UK regions. We again use a user cost model to predict regional house price growth, taking into account how rents have varied across regions and over time. We look at price growth since 2005, which is the period for which the statistical authorities produce data. And we hold all other factors constant, apart from the fall in rates, which applies to all regions. Chart 8 shows that though this model performs fairly well for the West Midlands and the North West, it under-predicts house price growth in London, the East and South East, and over-predicts it for the other regions. This may be because for regions like London, expectations of rental growth have risen over this period, or because houses in these regions are increasingly seen as safer assets than those in other regions.
Chart 8: Predicted house price growth in UK regions vs actual growth
So what might the framework be missing?
All economic models involve abstractions and simplifications – if they didn’t they would be useless. But given the importance of housing to economic growth and people’s wellbeing, and that changes in house prices could impact financial stability, we need a model that captures the key drivers.
There could be a role for changes in credit conditions. The framework assumes that people are not credit constrained, meaning they can exploit arbitrage opportunities by buying up rental properties. If there are frictions in practice, this could mean that credit conditions matter for house prices. Mortgage debt expanded rapidly as house prices rose in the UK before the crisis, so this could be an important channel for the UK.
The framework does not include an explicit role for supply elasticity. In practice, evidence suggests that housing supply responds very differently to prices in different countries and regions. The Barker review of UK housing supply pointed to low elasticity of supply in the UK. Where supply elasticity is low, as is the case in the UK, the same change in rates will have a larger impact on prices and rents. More work is likely required to document UK supply elasticities in detail and to explore what different elasticities have meant for growth in prices and rents over time and across places.
It’s hard to measure how expectations and risk premia vary over time and place but this could be important. Within the framework, an increase in rental growth expectations or a fall in risk premium would imply an increase in price. There is some evidence to suggest that returns on housing have been similar to other risky assets and relatively constant over the long term.
Other relevant factors might include maintenance costs and taxes. Maintenance costs have risen in line with inflation in the UK. UK property taxes are small and have been relatively flat – but there is evidence that higher taxes would likely reduce house prices. All of these are subject to considerable uncertainty.
Regardless of the consequences of Covid-19 on house prices, the long-term drivers discussed here will remain crucial parts of policy questions in coming years. Future research might be needed to better quantify some of these factors and to build them into a richer modelling framework that considers interactions between interest rates, credit conditions, rents and house prices.
Lisa Panigrahi and Danny Walker work in the Bank’s Macro-Financial Risks Division.
If you want to get in touch, please email us at bankunderground@bankofengland.co.uk or leave a comment below.
Comments will only appear once approved by a moderator, and are only published where a full name is supplied. Bank Underground is a blog for Bank of England staff to share views that challenge – or support – prevailing policy orthodoxies. The views expressed here are those of the authors, and are not necessarily those of the Bank of England, or its policy committees.
Everything stems from the quantity of credit banks and regulators permit to be allocated to housing. Debt serviceability comes from % disposable income able to cover annual debt service. Annual debt service can be altered by changing the debt term length (or even interest only mortgages!), interest rates and debt quantum (debt to income or LTV). A study needs to be done on the history of bank lending standards on these 3 key variables, as well as the change in the amount homeowners are spending on servicing mortgage debt as % disposable income. There will be a threshold which cannot be crossed where lending standards are so loose and the % income spent on housing is so high, it cannot go any further and it detracts from spending and consumption that drives GDP transactions, having a negative total effect on the economy. This would in in turn impact incomes and the ability to service the mortgage debt taken on. Do we reach a condition where we have low/flat GDP growth as unproductive spending on housing takes up such a large portion of incomes?
This all then feeds back into the value of the land the house sits on as the £ per square foot to construct the bricks and mortar has risen with inflation, the difference in house price growth is attributable to the change in the land price the house sits on. Landowners have been the major beneficiaries over the years.
This is a big problem as the more consumers spend on servicing mortgages the less they can consume or invest in productive assets. Spending so much of one’s income on a house that has already been built detracts from spending on GDP contributing transactions (investing in companies to create new products, consuming more goods).
It also over time increases inequality as lower incomes spend more as a % disposable income on rent (as can never save enough for the deposit required for high house prices) and so can never accumulate an asset on their personal balance sheet (swap debt for equity as pay down mortgage vs just spending rent). Non-asset owners remain non-asset owners.
As so much of UK banks assets are mortgages and the UK’s wealth sits in housing I am surprised further studies have not been taken by the BOE. Whilst factors such as supply, population growth and urbanisation do play a part, we need to uncover the weighting each variable plays. Bank credit standards and availability will be the resounding factor.
Spot on. As someone who has been financing mortgage loans for over 20 years, in and outside the UK, it is very clear to me that a structural change is needed to the UK mortgage market (and I am trying!). The key is affordability of debt, and responsible lending, to ensure a borrower can always repay his debt (aside from personal life events). The regulators have introduce the responsible lending rules and interest rate stress requirements to make this happen and this would put a natural cap on how much can be advanced by the banks, and hence how much properties can be bid up. This avoids a worsening housing bubble, because the house price itself is not a factor in the decision making. What these rules have done is dislocate the market, and make properties relatively unaffordable, mainly due to the interest rate stress requirements. There is an easy solution: rates are so low, offer people Fixed for Life mortgage loans, and people can afford 6/7 times their income, and hence can afford to buy a property at current price levels. The current lenders don’t like these products though, they are good for the consumers, but not good for the lenders.
One of the questions posed (chart 3) is why house prices fell slightly from 1990 to 1995 and then were stable from 1996 to 2000, despite favourable rates of interest.
I think that the answer is that there was a huge increase in nominal prices from December 1987 to December 1989 of 39% (Nationwide stats), followed by a fall of 18% over the next three years in the recession. Buyers of houses experienced ‘negative equity’ and there were quite large numbers of repossessions. Mortgage lenders then tightened their standards in the early 1990s.
The experience of house prices falling in 1987/89 affected the behaviour of buyers during the whole of the following decade of the 1990s.