The Great Recession was preceded by the build-up of a large stock of household debt in the US. Before its onset, there was an extended period of rising house prices. When the recession took hold, it was accompanied by a sharp decline of house prices. Figure 1 illustrates these developments for the US. Similar swings in house prices were also observed in other economies, such as Ireland, Spain, and the UK (Jordà et al. 2015).
Figure 1. House prices and mortgage debt in the US
Source: Federal Reserve Economic Database.
Top panel: Case-Shiller 20-city and 10-city home price index relative to the consumer price index for all urban consumers and all items less shelter. Bottom panel: mortgage debt of households and non-profit organisations per GDP. Note: Shaded areas are recessions dated by the NBER.
In the aftermath of the Great Recession, these developments have been mirrored in at least two focal points of policy debates. First, the deleveraging of bloated household balance sheets and its repercussions on consumer demand have become a major concern in macroeconomic policymaking. Second, financial policymakers have been reminded of the challenges presented by booms and busts of house prices – and in particular, of financial stability issues of those institutions involved in the intermediation of collateralised household debt.
Whodunit: What caused the large swings in house prices?
The real side of the economy raises hardly any suspicion. It would be too far-fetched to argue that the swings in house prices were caused by changes in real fundamentals, such as changes in productivity or demography. Recognising that changes in fundamentals would need to match the order of magnitude and the pace of the observed housing price swings, any search for real suspects seems truly out of place in the case of a sudden housing bust.
The financial side of the economy holds more promise for detection of the culprit. Which piece of evidence could explain the contrasting developments of house prices during the 2001 recession and the 2007–2009 recession? The bottom panel of Figure 1 reveals the traces of a candidate explanation: the large increase in mortgage debt.
The link between household debt and macroeconomic fluctuations
Empirical evidence suggests that macroeconomic fluctuations depend on the level of private debt. Credit booms in the private sector of developed economies between 1870 and 2008 are associated with a higher probability of financial crises, deeper recessions, and slower recoveries (Schularick and Taylor 2012, Jordà et al. 2013), and the Great Recession in the US was more severe in regions with larger household debt (Mian and Sufi 2010, 2014). In recent work, we develop a formal analysis to scrutinise the patterns of interplay between household debt and economic fluctuations (Hintermaier and Koeniger 2015). Our analysis is set within the framework of an economic model. This allows us to inspect the mechanism responsible for the build-up of balance-sheet conditions which hold the potential for unloading in a fully-fledged housing crisis. Besides identifying crucial risk factors, our analytic approach informs the design of macro-prudential strategies regarding household debt.
Household debt: Not just an amplifier, but an underlying cause of confidence-driven crises
Research at the junction of macroeconomics and finance has long – at least since the seminal work of Kiyotaki and Moore (1997) – recognised that collateralised debt in the private sector may amplify economic fluctuations. The existing paradigm for the analysis of macro-financial linkages is based on the notion of fundamental shocks that hit some of the fundamental features of the economy, such as fluctuations of productivity. This paradigm assigns the economy’s financial side the role of a supporting actor, acting with some intensity as a shock-amplifier.
Our work goes beyond this paradigm in a specific way – we develop a framework that accommodates the potential role of fluctuations in confidence as an additional source of uncertainty in the economy. Our analysis keeps track of the evolution of balance sheets. These balance sheet conditions of an economy determine the circumstances under which severe economic fluctuations may actually be confidence-driven. Our approach allows the financial side of the economy to play a much more prominent role – we show that the level of debt may turn out to feature the leading part, independently representing the underlying cause of confidence-driven crises.
Consumer confidence shapes house prices in a vulnerable economy
Figure 2 shows the relationship between the financial asset position – negative for debt – and the house price for the economy with collateralised household debt. The three scenarios considered in Figure 2 differ with respect to the loan-to-value ratio applicable to mortgage debt, the interest rate at which debt financing is available, and with respect to real income growth in the economy.1 These are the crucial features necessary to determine the circumstances and the extent to which house-price fluctuations may be purely confidence-driven.
Figure 2. Relationship between financial positions and house prices
Notes: Shapes of the relationship between the financial asset position and the house price for three scenarios: low loan-to-value ratio, benchmark, low-interest-rate environment with growth. Whenever there is a proper range of house prices for some specific financial position, this represents the extent of purely confidence-driven variation. These shapes are characterised according to the method in Hintermaier and Koeniger (2015).
The top panel of Figure 2 considers an economy where the limit on the loan-to-value ratio is low. A curve shows the dependence of the house price on the financial position of households. In some parts, house prices react quite strongly to variations in the financial position. This mirrors the particular bite of the constraint on collateralised borrowing, which provides for the crucial macro-financial linkage of the model.
Confidence enters the picture in the second panel of Figure 2 – compared to the previous case, this benchmark scenario considers a setting with a higher limit on the loan-to-value ratio for mortgage debt. Now the relationship between the financial position and the house price for this scenario is not just a curve anymore. The model predicts a shape, which has a specific pattern of thickness for some debt levels. Put differently, for specific levels of debt, the analysis determines a proper range of house prices – not just a certain price. For a given financial position, fluctuations of house prices within such a range are purely confidence-driven.
The current debt level thus determines the exposure to a crisis of confidence. Our framework therefore rationalises a clear role for (debt) “quantities as vulnerability indicators in tranquil times,” pursuant to Brunnermeier and Sannikov (2014). This aspect of our results may also be viewed as formally identifying the prerequisites of a so-called ‘Minsky moment’.
A strong liquidity feedback effect is pivotal to an independent role of confidence
Households in our economy borrow against the value of their housing collateral. This exposes households to a liquidity effect, through which the house price feeds back on itself – a high house price supports high credit limits for collateral-constrained households; high credit limits support high household demand; high demand supports a high house price. This closes a cycle which may be self-reinforcing, since replacing all instances of the word “high” by the word “low” yields an equally plausible chain of statements. Credit and asset price cycles with some similar elements feature in a recent book by Turner (2015).
The quantitative analysis of our model allows us to identify the circumstances – in particular, the debt levels – for which this feedback mechanism is strong enough to actually support a multiplicity of market-clearing house prices. Forward-looking households entertain beliefs about which of these multiple house prices is going to prevail in the future. Swings in confidence correspond to swings in these beliefs, thereby driving changes in the current house price and in consumer demand.
The perils of a low-interest-rate environment — a case for macro-prudential policy
The bottom panel of Figure 2 shows that the picture changes drastically in a scenario that combines a low-interest-rate environment with expectations of real income growth. Such an economic environment provides an ideal breeding ground for confidence-driven housing debt crises. The range of debt levels at which the economy is prone to purely confidence-driven crises becomes much larger, as does the extent of downside systemic risk for house prices and the potential of looming repercussions on consumer demand.
The challenges for economic policy in such an economic environment have garnered the attention of, for instance, the Bank for International Settlements (2015) and the Sveriges Riksbank (2015). In the light of our analysis, the exclusion of loans to households for house purchase mentioned in the modalities for ‘Targeted Longer-Term Refinancing Operations’ by the European Central Bank (2014) appears as a provision to mitigate the risk emanating from vulnerable household balance sheets.
The macroeconomic liquidity feedback mechanism makes confidence-driven housing crises an instance of systemic risk. Therefore, the prevention of crises of confidence seems to be a clear case for macroprudential policy. The recommendations of the European Systemic Risk Board (2013) contain elements for dealing with the crucial risk factors that we have identified in our work.
Bank for International Settlements (2015) 85th Annual Report, Basel.
Brunnermeier, M K and Y Sannikov (2014) “Monetary analysis: Price and financial stability”, ECB Forum on Central Banking, Sintra, Portugal, Print.
European Central Bank (2014) “Targeted longer-term refinancing operations: Updated modalities”, Annex to Press Release of 29 July.
European Systemic Risk Board (2013) “Recommendation of the European Systemic Risk Board of 4 April 2013 on intermediate objectives and instruments of macro-prudential policy”, Official Journal of the European Union, 15 June, C 170/1-19.
Hintermaier, T and W Koeniger (2015) “Household debt and crises of confidence”, CEPR, Discussion Paper No. 10865.
Kiyotaki, N and J Moore (1997) “Credit Cycles”, Journal of Political Economy, 105: 211–248.
Jordà, Ò, M Schularick and A M Taylor (2013) “When credit bites back”, Journal of Money, Credit and Banking, 45: 3–28.
Jordà, Ò, M Schularick and A M Taylor (2015) “Leveraged bubbles”, CEPR, Discussion Paper No. 10781.
Mian, A and A Sufi (2010) “Household leverage and the recession of 2007–09”, IMF Economic Review, 58: 74–117.
Mian, A and A Sufi (2014) House of debt, The University of Chicago Press, Chicago.
Schularick, M and A M Taylor (2012) “Credit booms gone bust: Monetary policy, leverage cycles, and financial crises, 1870–2008”, American Economic Review, 102: 1029–1061.
Sveriges Riksbank (2015) “The housing market and household indebtedness from a central bank perspective”, speech by Governor Stefan Ingves, 19 November, locality: SNS, World Trade Center, Stockholm.
Turner, A (2015) Between debt and the devil: Money, credit, and fixing global finance, Princeton University Press, Princeton and Oxford.
1 For the purpose of illustrating the scope of confidence-driven fluctuations, we have shut down any fundamental shocks to the economy in all scenarios.