It is an empirical fact that housing wealth constitutes a major share of households' wealth portfolio. Although recent research has shown that as households get older the relative share of housing wealth in their portfolio falls as they invest relatively more in financial assets, the economic significance of housing wealth to households cannot be discounted. Because the dynamics of housing wealth is critically dependent upon the dynamics of house prices, it is clearly important not only to understand how changes in house prices affect the economy, but also which underlying factors are critical for the dynamic behaviour of house prices.
Much of this need to understand the determination of housing market equilibria and the implications of changes therein stems, on one hand, from the more general concern of how asset prices affect the economy and, on the other hand, from the more specific concern of the potential wealth effects of housing on consumption and, hence, on aggregate demand. These wealth effects should not, however, be taken for granted. Aside from being quantitatively important to households, housing is also of special interest for another reason. Most consumers live in the houses they own and value directly the services provided by their home. So the benefit of an increase in house prices is directly offset by an increase in the opportunity cost of housing services. An increase in house prices does not, therefore, generally shift the aggregate budget constraint outwards. It is not obvious, then, that there is a traditional wealth effect from housing in the way that we think of a wealth effect arising from a change in the value of households' financial assets.
There are many reasons why house prices and consumption may move together, one of which is that they may both be driven by the same underlying shock, eg a change in the expected future level of income. Alternatively, house prices may have a direct impact on consumption via credit market effects. An increase in house prices makes more collateral available to homeowners, which in turn may encourage them to borrow more to finance desired levels of consumption and housing investment. Although various observations can motivate an examination of such a credit channel, we can bypass the listing of these observations simply by arguing that informational problems in the house financing system or, more generally, in the financial markets are the key factors that underlie the existence of the channel. This then implies, in particular, that households potentially face binding borrowing constraints when making consumption and housing investment decisions. These constraints may take different forms, from outright limits to the amount of borrowing, via collateral constraints or loan to value constraints, to specific down payment requirements. In addition to the question this raises of how borrowing constraints affect households' optimal consumption and investment decisions, there also emerges the interesting, albeit much less studied question of how these constraints influence the dynamics of house prices.
In their forthcoming Bank of Finland discussion paper, 'On the importance of borrowing constraints on house price dynamics', Essi Eerola and Niku Määttänen focus on this issue. More specifically, they construct and calibrate an OLG model to analyse how borrowing constraints in the form of down payment requirements affect the dynamics of house prices. In their specific OLG model with owner housing, young households need to borrow in order to finance their consumption, and differences in household size create large differences in household leverage, including across households of the same age. The model is then solved for the house price dynamics under two alternative scenarios given that the economy is hit by shocks to aggregate income and interest rates. The two scenarios make different assumptions about the existence of binding borrowing constraints that households face. Whereas households are unconstrained in the first scenario, in the second one they can borrow only up to a certain fraction of the value of their house. The authors are particularly interested in situations where a substantial fall in house prices dramatically reduces the net worth of leveraged households.
To sharpen the focus on the effects of borrowing constraints, Eerola and Määttänen make a couple of simplifying modelling assumptions that not only allow them to solve for the fully non-linear dynamics very precisely, but also allow them to describe analytically how borrowing constraints affect aggregate housing demand. More specifically, the authors assume perfect foresight and the absence of transaction costs and other non-convexities in the household problem.
Eerola and Määttänen calibrate the model to the Finnish data and compare the model dynamics with recent experience in the Finnish housing markets. Finland is a particularly interesting test case, since the country has recently been hit by two major shocks, a credit market liberalization in the late 1980s and a severe recession in the early 1990s. Both episodes were associated with large changes in house prices. Consequently, computing the house price dynamics following similar shocks in the model and comparing the outcome against actual data on house prices provides a convenient test of the quantitative relevance of the model.
Eerola and Määttänen derive a set of very interesting results from the model as well as from the calibration exercise. First of all, they find that the model captures a large part of the increase in house prices that coincided with the credit market liberalization. Consequently, the observed increase in house prices can be seen as an equilibrium response to the empirically plausible relaxation of borrowing constraints. This result also suggests that the model captures much of the actual relevance of borrowing constraints on aggregate housing demand.
The authors also find that in some situations remaining borrowing constraints can substantially shape house price dynamics. More precisely, the importance of borrowing constraints depends heavily on the underlying source of house price dynamics, in particular whether changes in house prices are primarily driven by shocks to aggregate income or interest rates. Moreover, because of borrowing constraints, house price dynamics display important asymmetries between large positive and negative income shocks. These results are related to fact that the share of borrowing-constrained households changes with the underlying shock. |