House Prices and Consumer Spending ... A Danish Perspective
Warning: long post!
In one of my recent posts I took up the question about the effect on consumer spending from a rise in housing prices. My source was an article from The Economist, which discussed this topic from a US perspective and in relation with the slowdown of the US economy. In short; given the recent slump in US house prices the stronger correlation between private consumption and fluctuations in house prices the higher the risk of a real US recession. In a US context the consensus seems to be that the effect from rising house prices on private consumption is relative large and that it has been rising. But what about in other countries and regions?
Brad Setser has been doing some interesting empirical work on this lately about talking the differences and similarities between the housing bubble in the US and Europe. In a recent article over at Brad's blog Charles Gottlieb of the European Capital Market Institute elaborates on the European housing market concluding that Europe is ready to take over from the US in driving global demand forward. I have argued elsewhere why this is unlikely and I won't go into that discourse here. However, a specific interesting thing noted by Gottlieb's piece is that the US consumers seem to be better/increasingly ready to convert wealth accumulation accrued from rising real estate into private consumption. Brad Setser also touches upon this in his recent article about the subject in which he compares France and the US concluding that the two countries share a lot of similarities concerning the developments in the housing market. However, Brad returns to the initial point ...
Fortunately, as emphasized in my previous contribution, French domestic demand seems to be less sensitive to home price movements than US domestic demand (in part because French credit market as not as well developed). Thus French households are not as strongly exposed to potential drop in house prices as US households are, even if a mix of low interest rates, tight supply (at least in some areas) and speculation has propelled housing market in both countries to the sky.
All this interests me from a theoretical perspective but also particularly from an empirial evidence because if there are theoretical squabbles about the transmission mechanism between house prices and private consumption then it should also be clear that there exists notable differences between countries and regions as initially demonstrated by Setser and Gottlieb cited above. In short; this has prompted me to dig and what better place to begin than my home country Denmark.
In that regard we might want to take a short glance at the Danish mortgage market which is very similar to the US ditto.
With a standard Danish mortgage contract, it is possible to borrow long-term (up to 30 years) at fixed rates with an option to make penalty-free prepayments. This option is also embedded in the US mortgage contract. US and Danish mortgage markets are globally exceptional in this regard. The main consequence of this option element in the US-Danish style
contract is that investors are exposed to prepayment and thus reinvestment risk. On the one hand, over longer periods the risk characteristics of the typical callable mortgage bonds are found to be similar in the two markets. On the
other hand, the Danish market’s performance has not been much affected in periods with significant refinancing, which is a well documented characteristic of the US mortgage market.
Basically the Danish housing finance market (Danish link!) is tied up to and financed through the bond market. As such there are two main issuers of bonds in Denmark. The Danish Central Bank and credit institutes of which the latter is of special importance here as they are the primary vessels of financing real estate. The financing is mainly done through the issuing of convertible bonds often with a maturement time of 30 years. Apart from loans based on convertible bonds the Danish credit market has developed considerably in the recent 10 years. In 1996 the credit institutes was enabled to issue non-convertible bonds with shorter maturement and on which the interest rate is subject to change during the payment period. In the recent years this last type of loans have been an increasingly large share of the credit insitutes outstanding loans. Finally in 2003 yet a new loan type was introduced; in free translation you could call these 'non-downpayment loans.' Basically these loans are offered with either a fixed or variable interest rate and furthermore with the specific feature that the downpayment of the loan can be switched off (i.e. you only pay off the interest) either in the beginning of the loan period or by random annual basis decided by loantaker through a 'ticket system.'
At a first glance this does not seem to have a specific effect on the transmission mechanism from rising house prices to private consumption albeit with the exception of the last type of loan introduced in 2003 which intuitively does seem to incite loantaker to increase consumption in some years (or at least gives her the possibility). However, if we turn the tables we might see an example of how the continuing development in credit markets can contribute to a general rise in house prices. As such, the debate in Denmark is split between those who argue that the new loan types are a source of flexibility whereas others see the new loans particularly the 'non-downpayment' loans as a proxy for a spend now, pay later mentality which bids up property prices (see the notion of herd behavior below). Furthermore, many has suggested that given the current low-interest rate environment households with variable interest rate loans risk seeing their financial expenses rise considerably if the interest rates go up.
So do we have a housing bubble in Denmark then ... ? Well it could seem so as Denmark comfortably takes possession of the 1st place in the lates house price index from The Economist. Also a recent survey the OECD survey (PDF!) sums up the mainstream view ...
'Since 1995, Danish house prices have increased substantially with an average yearly growth rate of 8% (6% real). The ratio of house prices to disposable income has risen over the same period by 70%. This has supported consumption growth, not least because the increased flexibility of the mortgage market made it easier to borrow against higher house price values. Over the last year, the speed of price increases has been particularly strong.'
Also an ECB paper from 2003 suggests that Danish households, at least, as a function of the credit markets have the possibility to tap into higher house prices.
In all five countries [Denmark, Sweden the UK, the Netherlands, and Portugal], it is estimated that, in certain periods, there has been significant house equity withdrawal. In Denmark, the Netherlands, Sweden and the United Kingdom, house equity withdrawal has generally coincided with significant upturns in housing prices, as would be expected if the “credit” channel was in operation. All four countries have high accumulated debt levels and in all four the mortgage system offers instruments that permit households to tap their housing wealth directly.
Ok, square and done then? Well, not quite. A study from 2005 by Jens Lunde at the Copenhagen Business School which takes on the Danish housing market in larger perspective seems to suggest a somewhat different analysis.
(quote from the abstract)
From 1998 on, the former positive correlation between house prices and the private consumption has been changed, and herd behaviour as well as the new mortgage products are among the possible causes.
Reading on we are given the first indication of a concrete transimission mechanism between house prices and private consumption (10:1) in Denmark as Jens Lunde quotes the monetary review of the 4th quarter published by the Danish National Bank (see list of references in Lunde's paper).
In accordance with the consumption relation in MONA: ”the consumption in fixed prices will increase with around 10 DKK for each 100 DKK increase in the real wealth.” (Danmarks Nationalbank, 1997b, p. 9).
Now as you might have guessed Lunde is rather sceptical about this correlation and in terms of the debate in Denmark back in the late 1990s the consensus emerged something like this:
The debate in Denmark ended with an article in the central bank’s quarterly review, where an economist in the central bank argued that on the long-term the increasing house prices would not influence private consumption, while on the short-term the house price rise could influence the propensity to consume out of income (Pedersen, 1998). This view was in accordance with the earlier critic (Lunde, 1998). But absolutely, ”the man in the street” is convinced that the rising house prices boost the owners’ economy and that the tenants are impoverished.
To be fair to Lunde he is not blind to the fact that general studies on this show a significant correlation between rising house prices and private consumption. However, if we stay in Denmark Lunde presents evidence that the correlation between rising house prices since 1998 has been negative in Denmark.
Since 1998 has been seen a deciding break as the earlier close positive correlation between the increase in the real house prices and the consumption/income ratio has been interrupted. Instead after 1998 the correlation has been negative as seen in Figure 2 below.
In conclusion (pp. 24-25) Lunde offers possible explanations as to why the link between consumption and rising house prices might be broken in a Danish context. The following are particularly important I think ...
Widespread herd behaviour or speculation. Even though the documentation is dubious the population seem to have positive expectations on, that the prices on houses and flats will continue to increase more than the consumer prices and that the interest rate level will remain low (see Section 3 above). These expectations could drive the prices somehow
without basis in fundamentals or in the private consumption.
The rather quick extension of the adjustable-rate mortgage loans [see my description above] and the easy access to raise mortgage and to finance a house buying can have released a special and remarkable increase in the prises for houses and flats and in raising new loans with security in the owneroccupied dwellings.
So have I left you more confused than when we set off? I hope not. The transmission mechanism between rising house prices and private consumption clearly merits some general theoretical work but it is furthermore clear that country-specific conditions and dynamics are very important to differentiate our analyses.
List of References
Economist, the 2006 - Home truths (Economics Focus).
Economist, the 2006 - Checking the thermostat.
European Central Bank, The 2003 - Structural Factors in EU Housing Markets.
Finansraadet.dk - Det Danske Obligationsmarked
Nykredit.com (Allen Frankel, Jacob Gyntelberg, Kristian Kjeldsen, and Mattias Persson) 2004 - The Danish Mortgage Market.
OECD Economics Department Working Papers No. 513 (Espen Erlandsen, Jens Lundsgaard and Felix Huefner) 2006 - The Danish Housing Market: Less Subsidy and More flexibility.
Setser, Brad 2006 - Is France the US of the Eurozone -- at least when it comes to housing?
Setser, Brad (Charles Gottlieb) 2006 - Is it Europe's turn to rise a housing bubble?
Danish mortgages—like those in America, but unlike mortgages in other rich countries—make it easy for borrowers to repay their loans early if they wish. The combination of fixed interest rates and an option to prepay helps to shield borrowers from interest-rate risks. If rates rise after they buy a home, they are protected by the fixed interest rate; if rates fall, they can take out a new mortgage at a lower rate and prepay the old one (see chart).
The prepayment option, like bond issuers' options to call some bonds before they mature, saddles the lender with more risk. Forecasting prepayment risk in America, for example, is complicated and requires lots of information about demographics and credit factors, along with plenty of market data. The Danish system is in many ways better than America's, especially as a template for emerging economies. Its mortgages must be financed through the issuing of bonds that adhere to the “balance principle”: the maturity and cashflows of the bonds must match those of the underlying loans almost perfectly. The bonds are callable, to reflect the prepayment risk of the underlying mortgages.
Denmark's approach allows mortgage-backed securities to be standardised into large pools, which makes them more liquid and attracts investors. Besides increasing liquidity, standardisation also makes it easier for investors who buy the bonds to forecast and deal with prepayment risk. And the system efficiently separates the business of mortgage lending from the ultimate investors—such as pension funds and insurers—who supply the funds by purchasing the mortgage-backed bonds. Denmark's mortgage lenders compete on their ability to distribute and service the loans, and this specialised competition has helped to drive down costs.
Danish mortgage bonds certainly seem appealing to investors. With DKr 1.6 trillion ($260 billion) outstanding at the end of 2005, the country has the second-biggest market in Europe for mortgage-backed securities, after Germany's Pfandbrief market. Measured as a share of GDP, the market is far bigger than Germany's or America's. The Danish system is also more stable in many ways than America's. Although it has held roughly the same form since the late 18th century—it was set up as a way to rebuild rapidly after the Copenhagen fire of 1795—there has never been a default on a Danish mortgage bond.