Abstract
In recent years, the IMF has released a growing number of reports and other documents covering economic and financial developments and trends in member countries. Each report, prepared by a staff team after discussions with government officials, is published at the option of the member country.
Housing and Business Cycles: Is the UK Different From Other Advanced Economies?1
In contrast to other OECD countries, housing cycles in the UK are marked by sharp movements in prices and an inelastic response of residential investment, owing notably to supply constraints. Housing cycles in the UK also tend to have a large impact on economic activity, with booms generally associated with a worsening of household balance sheets and a rise in relatively high-risk mortgages. Alleviating supply-side constraints, notably pertaining to planning restrictions, is imperative for a moderation of housing cycles in the UK, while risks to financial stability in the context of the current house price inflation could be addressed by pursuing targeted macroprudential measures.
A. The UK Housing Market: A Historical Perspective
1. The housing sector plays an important role in the UK economy. The value added of the real estate sector—related closely to the housing market as a sector—has been increasing steadily over time in the UK. While in 1990 it represented about 6 percent of GVA, in 2013 it reached 12 percent. Moreover, when all industries related to the housing market (finance, real estate, and construction, or FREC) are taken into consideration, the share rises to about 25 percent of GVA, substantially higher than the average for the OECD economies taken together and one of the highest among the G7 economies, underscoring the increasing importance of the housing sector in the UK economy.
Housing-related Sectors: Finance, Real-Estate, and Construction
(Percent of GVA)
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A001
Source: Haver Analytics.2. House price increases in the UK stand out among the OECD economies. Over the past 30 years, real house prices have increased the most in the UK when compared with other OECD economies. Indeed, over this period, annual house price increases have averaged 3 percent in real terms, compared with 1 percent for the OECD as a whole. This divergence in house price increases was particularly pronounced from the mid-1990s through the Great Recession. Furthermore, house prices in the UK have also been a lot more volatile when compared with other advanced economies.
Real House Prices in OECD Economies
(Index, 100 = 1980)
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A001
Sources: OECD; and ONS.3. In the context of a strong demand for housing, rapid price increases reflect the impact of serious supply constraints in the UK housing market. Residential investment in the UK as a share of GDP is among the lowest across the OECD economies. The sluggish response of residential investment to a strong demand for housing is attributed to supply-side constraints. In particular, restrictive planning regulations, in combination with inadequate incentives for local authorities to grant building permits, have resulted in a low house price elasticity of residential investment in the UK.2 Obtaining a planning permit in the UK takes about 25 weeks, longer than in the average OECD economy (NAO, 2008).3 There is some evidence that past attempts by the government to limit the time involved in processing planning permits have resulted in an increase in the rejection rates of permit applications, and an increase in the time spent in the appeal process (Corder, 2008 and NAO, 2008).4
OECD: Residential Investment and Planning Costs
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A001
B. Housing Booms and Busts in Advanced Economies
Does the UK Stand Out?
4. Housing cycles in the UK are characterized by large fluctuations in real house prices and a limited response of residential investment.
The UK stands out as having the largest fluctuations in real house prices among G7 economies.5 In fact, house price volatility is even larger than in countries that have experienced pronounced housing cycles, such as the US and Canada.
On the other hand, residential investment in the UK is less volatile than in most OECD economies. This volatility, in turn, is influenced by the elasticity of housing supply.6
The UK Housing Cycle: Volatile Prices and Subdued Residential Investment (1980–2013)
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A001
Source: Haver Analytics; and IMF Staff calculations.5. Housing cycles in the UK have become more persistent and volatile over time. The properties of the housing cycles have changed dramatically over time.
The duration of housing cycles has increased dramatically during upturns. While during the 1980–96 period, the average duration of upturns in the UK was 31 quarters, during 1997–2013 the average duration increased by 50 percent to 47 quarters.
Moreover, the amplitude of the cycle—measured by the difference in real house prices between the trough and peak—during upturns increased from 98 to 170 percent.7 This is surprising, since most macroeconomic variables across OECD economies experienced a decline in volatility over the same sample period (Blanchard and Simon, 2001).
The volatility in the housing market has been amplified by a loosening of credit conditions prior to the crisis (Igan and Loungani, 2012).
The UK Housing Cycle Has Changed Over Time
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A001
Sources: Haver Analytics; and IMF Staff calculations.6. House price volatility in the UK has been influenced by restrictive planning regulations and favorable credit conditions. The properties of the UK housing cycles can be interpreted through a simple model of demand and supply for housing. Large fluctuations in prices with a limited response in quantities are a result of large shifts in the demand curve in a market where the supply is inelastic. In fact, some key features of the UK housing market indicate that this is the case.
Planning costs are relatively high compared to the average OECD economy, and tend to reduce the elasticity of residential investment to house prices.8
Moreover, mortgage loans with a high loan-to-value (LTV) ratio have generally been more pervasive in the UK, and have contributed to boosting the demand for housing (IMF, 2011).
To test the influence of these factors on house price volatility, we split the sample of OECD countries into different groups—countries with high and low planning costs, and countries with high and low LTV ratios—and compute the average annual price change in each subsample.9 This exercise shows that house price volatility tends to be high in countries with high planning costs and high LTV ratios, suggesting that these two factors could be playing a pivotal role in influencing the UK housing cycle.
Planning Restrictions and LTV Ratios Influence the Housing Cycle
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A001
Sources: World Bank; Haver Analytics; and IMF Staff calculations.7. Households’ balance sheets expand dramatically during house price booms, resulting in a higher financial vulnerability of the household sector. For instance, as a consequence of sustained increases in house prices ahead of the current crisis, housing wealth sharply increased in the decade prior to the event. The increase in housing wealth went hand in hand with higher mortgage debt. However, notwithstanding the large increase in mortgage debt, housing net worth (the difference between housing wealth and mortgage debt) reached 200 percent of GDP in 2007. But the increase in net wealth on aggregate masked an underlying vulnerability in the household sector. In particular, the mortgage market in the UK provides relatively high loan-to-value ratios for first-time buyers, making this segment of the household sector more vulnerable to swings in the housing market. This vulnerability, in turn, magnifies the impact of housing shocks on economic activity.
UK: Household Sector Balance Sheet
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A001
The Great Recession and its Aftermath
Impact of the crisis
8. Advanced countries that experienced the most pronounced housing booms before the crisis exhibited the largest decline in GDP relative to trend. In the run-up to the crisis (1997–2007) real house prices in the UK increased by 150 percent, more than in other OECD countries, except Ireland. Post-crisis, GDP in the UK was 18 percent below the pre-crisis trend in 2012. On average, OECD economies experienced a decline in GDP relative to trend of 14 percent. Some countries, such as Switzerland or Germany, experienced a milder pre-crisis housing boom and a smaller decline in de-trended GDP. This pattern illustrates how large swings in the housing market are correlated with economic activity.10
Change in Real House Prices and GDP Deviation from Trend in OECD Countries
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A001
Source: Haver Analytics.9. During the Great Recession (2008–10), the UK’s housing bust was more severe than in most G7 economies. Real house prices in the UK declined by 15 percent while residential investments as a share of GDP declined by 50 percent. Only the housing bust in the US was more severe. In contrast, Canada and Germany experienced a housing boom shortly after the Great Recession. More recently, UK’s real house prices and residential investment have been recovering, but they remain significantly below the pre-crisis peak.
G7: Housing Market During the Great Recession
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A001
Sources: Haver Analytics; ONS; and OECD.10. The housing bust had a significant macroeconomic impact in the UK, particularly on private consumption and household debt. The UK exhibited the largest decline in private consumption and one of the largest reductions in household debt ratios among G7 economies, which possibly reflects not only wealth effects from lower house prices, but also the prevalence of tight credit conditions in the aftermath of the crisis. On average, countries that experienced a large housing bust (Italy, Japan, and the UK) also experienced a deep recession. On the contrary, countries that experienced a housing boom (Canada and Germany) exhibited a shallow recession. The US experienced a different boom-bust pattern as its economy was able to grow at a relatively fast pace after a housing bust, owing to a pick-up in investment and exports and a fast rebound in productivity.
Post-crisis Recovery
11. Price-to-Income and Price-to-Rent ratios suggest an overshooting in house prices. Standard valuation ratios indicate that house prices remain high relative to income and rents, suggesting an overshooting of house prices. The extent of overshooting can be estimated by calculating the difference between current housing valuation ratios and a benchmark long-term valuation ratio. An assessment based on two benchmarks, the average of the ratios over the past 15 and 30 years, shows that the overshooting is in the range of 10–30 percent.11
UK: Price-to-Income and Price-to-Rent Ratios
(Index, 100 = 2010)
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A001
Source: OECD.12. The current UK housing recovery occurred in the presence of weak credit growth, suggesting a greater role of cash transactions. The current housing recovery is comparable with previous historical episodes. Real house prices in this recovery are close to the average of the previous two recoveries, and residential investment has been increasing at the same pace as in previous episodes. However, it is puzzling that housing recovery is taking place in a context in which GDP growth is weaker than in the past, and the ratio of credit to the private sector to GDP (measured by M4 and M4LXex lending) is declining. The increase in house prices in a context of weak credit growth suggests that cash transactions, in particular by foreigners, are playing an increasingly important role in the housing recovery.12
The UK is Experiencing a Strong Housing Recovery Compared to Past Episodes
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A001
Sources: ONS; and IMF staff calculations.13. The UK housing recovery remains unbalanced. Demand is outpacing supply, particularly in the London market. There is a large disparity among regional house price indices. While real house prices in London have already reached their pre-crisis peak, house prices in other regions have not yet recovered. The differences in house prices might in part reflect regional disparities in the UK recovery. Furthermore, there is an increasing imbalance between demand and supply. While residential investment is recovering at a sustained pace, housing transactions have been growing in recent quarters resulting in an acceleration of house price inflation. This imbalance is likely to be more acute in the London property market, where prices are growing faster than anywhere else.
UK: Real House Prices.
(Index, 100 = 2007Q1)
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A001
Source: ONS.House Transactions and Residential Investment
(Index, 100 = 2007Q1)
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A001
Source: RICS; ONS.C. Housing and the Business Cycle
14. House price shocks in the UK have a strong impact on consumption and household debt. A Vector Autoregression (VAR) model was estimated to assess the impact of house price shocks on the UK economy.13 In response to a 10 percent increase in house prices, private consumption responds strongly, increasing up to 2 percent in 5 quarters. The household debt ratio reaches a peak response of 10 percentage points of disposable income after 12 quarters. The increase in private consumption can be explained by wealth and collateral effects. In particular, the existence of mortgage equity withdrawal in the UK allows households with positive equity to extract part of their housing wealth to finance consumption expenditures (Benito et al., 2006). Furthermore, the high elasticity of household debt to house prices captures the fact that mortgage credit expands rapidly during housing booms, resulting in a higher household debt leverage ratio.
UK: Consumption Response to a House Price Shock
(Percent)
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A001
Source: IMF staff calculations.UK: Household Debt Response to a House Price Shock
(Percent of disposable income)
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A001
Source:IMF staff calculations.15. In the UK, residential investment is less responsive to house prices and household debt is more elastic to house prices than in most OECD countries. A cross-country comparison of the transmission of house price shocks shows that the UK is different from other OECD economies. The response of residential investment in the UK is half of the median response for OECD economies. On the other hand, the response of household debt is more than twice as high as the response in the median OECD economy. This cross-country comparison highlights two key facts about the UK economy. First, housing supply constraints seem to be more binding in the UK resulting in a low elasticity of residential investment to house prices. Second, household debt is highly elastic, reflecting the fact that mortgage credit expands rapidly during housing booms.
UK: Real GDP During the Great Recession
(Index, 100 = 2008)
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A001
Source: IMF staff calculations.16. The housing bust in the UK was responsible for a third of the decline in GDP during the Great Recession. Using a historical decomposition analysis of the VAR model, we simulate the impact of housing shocks on GDP.14 The simulation shows that housing shocks reduced GDP by 3 percent, about a third of the fall in output experienced during the Great Recession. More importantly, the simulation suggests that the effects of housing busts are persistent and account for the weak recovery in the aftermath of the crisis.
17. In the early 1990s, the UK economy experienced a fast recovery from a housing bust due to a rapid increase in exports and investment. The historical decomposition analysis shows that the impact of the early-1990s housing bust on GDP was 2 percent. However, the recession was short-lived and the GDP recovery was robust. The negative impact of the housing bust on GDP was compensated by a rapid expansion in exports and investment during the first four years of the recovery. This episode illustrates how the rebalancing of the UK economy towards investment and exports can play an important role in ensuring a fast recovery in the aftermath of a housing bust.
UK: Real GDP During the 1990 Recession
(Index, 100 = 1990)
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A001
Source: IMF staff calculations.Average Growth of GDP Components, 1992–95
(Percent)
Citation: IMF Staff Country Reports 2014, 234; 10.5089/9781498304337.002.A001
Source: ONS.D. Conclusions and Policy Implications
18. The UK housing cycle is highly volatile as a result of tight housing supply constraints and fluctuations in credit conditions. The UK is characterized by having one of the largest fluctuations in house prices and one of the smallest volatilities of residential investment among OECD economies. These properties of the UK housing cycle are a result of both housing supply constraints and fluctuations in credit conditions. This volatility in the housing market, in turn, contributes to a more pronounced business cycle in the UK, as shown during the recent financial crisis. A more stable housing market requires policies that address both supply constraints and excessive fluctuations in mortgage credit.
19. Housing supply-side constraints can be alleviated through changes to the planning system and tax reforms. The new National Planning Policy Framework introduced by the government is creating the incentives for local councils to increase available land for construction. There are early signs that this change in the planning system is contributing to the recovery in housing construction. However, key inefficiencies remain in the way property and land are taxed. The current tax regime discourages institutional investment in rental accommodation, and undeveloped land is exempted from business taxes, which incentivize land hoarding. A reform of property and land taxation could improve the efficiency in the use of land and encourage an expansion in housing supply.
20. Targeted macroprudential policies could address financial stability risks stemming from the housing market. Although mortgage credit as a share of GDP has been declining in the current housing recovery, there are signs that there is a buildup of financial risks: loan-to-income ratios are increasing in London and among first time buyers.15 Macroprudential policies are the first line of defense against systemic financial risks. Given the uncertainty of the transmission of mechanism of macroprudential policies, there are merits in adopting macroprudential policy measures early and gradually.16 These policies would not only increase the resilience of the banks’ balance sheets, but would also reduce the volatility of the housing cycle.
Appendix 1. The VAR Model
1. A seven-variable macroeconomic VAR model is estimated to evaluate the impact of housing shocks on economic activity. The VAR model for the UK is estimated with quarterly data for the 1987–2013 sample period. The VAR includes the following variables: CPI inflation rate, residential investment, private consumption, GDP, interest rate, the ratio household debt to income, and house price. Data for the UK and other OECD countries were obtained from the Office of National Statistics (ONS) and Haver Analytics. All variables, except for the inflation rate and the interest rate, are expressed in logarithms.1 The system is identified following the standard recursive ordering (Cholesky decomposition), with the order following the listing of the variables mentioned above. We follow the Akaike criterion for lag selection, and estimate the model with 2 lags.2
2. The VAR model is related to an extensive literature on housing and business cycles. Several studies have estimated the impact of house price shocks on economic activity, and the role of monetary policy in stabilizing the housing cycle. Igan and Loungani (2012) estimate a VAR model for a sample of OECD economies, and find that house price shocks tend to have a larger impact on output in countries where mortgage credit is more accessible. Musso et al. (2011) compare the impact of housing shocks in the Euro Area and the US, and find that in the latter housing shocks have larger effects on consumption. In addition, house prices in the US are more sensitive to changes in the monetary policy rate. Giuliodori (2005) estimates the effects of the monetary policy shocks on house prices across European countries and finds that the housing market channel of monetary policy tends to be stronger in countries with more developed mortgage markets. Sa et al. (2011) find that in a sample of OECD countries, the effects of capital inflows on house prices is amplified in countries with more developed mortgage markets.
References
André, Cristophe, 2011. “Improving the Functioning of the Housing Market in the United Kingdom,” OECD Economics Department Working Papers, No. 867.
Barker, Kate, 2004. “Review of Housing Supply: Final Report,” HMT.
Benito, Andrew, Jamie Thompson, Matt Waldron and Rob Wood, 2006. “House prices and Consumer Spending,” Bank of England Quarterly Bulletin, pp. 142–154.
Blanchard, Olivier and John Simon, 2001. “The Long and Large Decline in U.S. Output Volatility,” Brookings Papers on Economic Activity, No. 1, pp. 135–174.
Caldera Sanchez, A., and A. G. Johansson, A. G., 2011. “The Price Responsiveness of Housing Supply on OECD Countries,” OECD Economics Department Working Paper, No. 837.
Cheshire, Paul, 2014. “Turning Houses into Gold: the Failure of British Planning,” Centrepiece, Vol. 19, No. 1, pp. 14–18.
Corder, Matthew and Nyssa Roberts, 2008. “Understanding Dwellings Investment,” Bank of England Quarterly Bulletin.
Giuliodori, Massimo, 2005. “The Role of House Prices in the Monetary Transmission Mechanism Across European Countries,” Scottish Journal of Political Economy, Vol. 52, No. (4), pp. 519–543.
Harding, Dan and Adrian Pagan, 2002., “Dissecting the Cycle: A Methodological Investigation,” Journal of Monetary Economics, Vol. 49, No. (4), pp. 365–381.
Hilber, Christian, and Wouter Vermeulen, 2012. “The Impact of Supply Constraints on House Prices in England,” London School of Economics, SERC Discussion Paper 119.
Igan, Deniz, and Prakash Loungani, Prakash, 2012. “Global Housing Cycles,” IMF Working Paper No. 12–217.
Musso, Alberto, Stefano Neri, and Livio Stracca, 2011. “Housing, consumption and monetary policy: How different are the US and the euro area?” Journal of Banking & Finance, Vol. 35, pp. 3019–3041
OECD, 2011. “Housing and the Economy: Policies for Renovation,” Economic Policy Reforms: Going for Growth.
Sá, Filipa, Pascal Towbin and Tomasz Wieladek, 2011. “Low Interest Rates and Housing Booms: The Role of Capital Inflows, Monetary Policy and Financial Innovation.” Working Paper No. 411. Bank of England.
Prepared by Ruy Lama and Stephanie Denis (EUR).
OECD (2011) and André (2011).
While the National Audit Office (NAO) estimates an average time to process planning permits of 25 weeks, the Doing Business Indicators (DBI) consider an average time of 13 weeks. The estimate of the NAO is higher than the DBI since it takes into account the time involved in the appeal process once a planning permit is rejected. In addition, the DBI database shows that the monetary cost of obtaining building permits in the UK is 66 percent of per capita income, higher than the cost in the average OECD economy (56 percent of per capita income).
Corder (2008) showed that the rejection rate of major housing projects increased from 15 percent in the mid 1990’s to 35 percent in 2008. One reason for the increase in the rejection rate is that the government introduced a target of 13 weeks to process planning permits. Local councils reached the target by increasing the rejection rate of permit applications.
We compared the average annual change in house prices and residential investment during upturns and downturns. The turning points for the upturns and downturns were defined using the BBQ algorithm from Harding and Pagan (2002). Our sample included 18 OECD countries.
The elasticity of housing supply depends on policy factors (planning costs and land-use regulations) and non-policy factors (land availability and population density). See Caldera Sanchez and Johansson (2011) for an estimation of housing supply elasticities from OECD countries.
The amplitude is measured as the difference between the peak and trough of real house prices.
In addition, the sluggish supply of housing in the UK is explained by the uncertainty over planning outcomes. While in the UK a development requires permission from the local planning authorities and is subject to delays, other OECD countries rely on rule-based zoning systems (Barker, 2004 and Cheshire, 2014).
The threshold for high planning costs and LTV ratios is defined by the sample median.
Although it is challenging to identify the direction of causality between house prices and economic activity in the data, in section III we estimate a VAR model to quantify the impact of house price shocks on GDP.
The high price-to-rent and price-to-income ratios in the UK reflect in part the historically low long-term rates.
M4LXex is a measure of credit to the private sector excluding the effects of securitizations and loan transfers.
See appendix for details on the estimation of the VAR model.
The simulation includes shocks to both house prices and residential investment to fully capture the extent of the downturn in the housing market.
See the Selected Issues Paper “Macroprudential Policy: Lessons from Advanced Economies.”
The Financial Policy Committee (FPC) has recently recommended a macroprudential policy measure: a cap on mortgages with high loan-to-income ratios. The new macroprudential policy measure states that no more than 15 percent of new mortgages could have a loan-to-income ratio of 4.5 or higher. This measure will be effective from October 1, 2014.
Sims, Stock and Watson (1990) show that if variables are cointegrated, the VAR model can be consistently estimated in levels.
We conducted a sensitivity analysis and estimated two alternative version of the VAR model: (i) a model where variables are de-trended and shocks are identified using a Cholesky decomposition; and (ii) a model in levels where the shocks are identified using generalized impulse response analysis. Both versions of the model provide similar results to the baseline model used in the paper.
