Wickman-Parak: The property market and the financial crisis
The American actor Danny Kaye once said: “It takes more than money to make you happy. You also need shares, gold and property.” The experience gained in the 1990s has taught us that the chances of becoming happy, defined in this way, can fluctuate wildly.
The question that I address in my speech is whether the situation that we are witnessing on the property market today could develop into a repetition of the property crisis that we experienced in the early 1990s. My answer is that even though we can see risks on the property market, there are no real signs that the current recession will entail a recurrence of the property crisis of the 1990s. Before I explain why, however, allow me to begin by telling you why the Riksbank analyses the property market. Then I will move on to discuss how we assess the current state of the market, both in terms of the housing market and the commercial property market. In the case of the commercial property market, I will also look back and make a number of comparisons with the crisis of the 1990s.
The Riksbank, financial stability and the property market
The Riksbank has a deep-rooted institutional memory of the bank crisis of the 1990s, which in many respects was a property crisis. Rough calculations show that approximately two-thirds of the banks' loan losses in the 1990s stemmed from the property sector.(1) The origins of the Swedish bank crisis, unlike the present crisis, were domestic, not international. It was also against the background of the Swedish bank crisis that the Riksbank began to analyse and monitor the stability of the Swedish financial system in a more systematic and comprehensive way. The Financial Stability Reports that we have since begun to publish have become an important tool for us. In these reports, we present our results and pinpoint the risks that may exist to the stability of the financial system.
These days, a considerable proportion of the banks' lending is to Swedish property companies; to be more precise around nine per cent of their total lending. Although the proportion of such loans is smaller now than in the 1990s, when the proportion was approximately one third, it is still so large that it constitutes a significant potential credit risk for the major Swedish banks. Let me therefore begin by saying something about the current situation on the housing and commercial property markets and their importance to financial stability.
Low interest rates have stabilised the housing market…
Between 1996 and 2008, house prices increased by almost 160 per cent. Good income growth – household disposable income increased by almost 65 per cent during the period – was an important contributory factor behind this price increase. However, another significant factor was that there was a general fall in mortgage rates during the period so that the interest ratio, that is post tax interest expenditures in relation to disposable income, also fell. There was also an increase in competition between the mortgage market participants. As a result of this increase in competition, interest rate margins were squeezed and credit conditions became more generous; for example households were able to get interest-only loans and the lenders accepted higher loan-to-value ratios. This meant that the amount that the borrower no longer needed to amortise could instead be used for interest payments on somewhat larger loans.
The rate of increase in house prices began to slow down towards the end of 2007. The Riksbank's gradual interest rate increases from the start of 2006 and onwards probably contributed to this. Towards the end of last year, house prices began to fall, partly as a result of an unusually large supply of housing that was not met by a corresponding increase in demand. The latter was a natural consequence of the general loss of households’ confidence in the wake of the financial crisis. Since the beginning of 2009, the prices of single-family dwellings had fallen by two per cent and the prices of tenant-owned apartments by nine per cent compared to the prices at the end of 2007.
Since the beginning of 2009, however, it appears that the situation on the housing market has stabilised somewhat. Prospective house buyers have become more active and there has been a decline in the time it takes to sell houses. Signs of budding optimism are also confirmed in SBAB's (Swedish Housing Finance Corporation) latest survey of estate agents. This reveals that demand is expected to increase and that the time to sale is expected to decline even further at the same time as prices are expected to increase somewhat.
…but a weaker labour market may lead to price falls in the period ahead
The strongest contributory factor to the recovery of the housing market is probably that mortgage rates have fallen dramatically since December last year as a result of the Riksbank's repo rate cuts. One could almost say that house prices just now are artificially boosted by the low mortgage rates, but, when those on redundancy notices actually become unemployed the demand for housing will fall and this will have an impact on prices. Therefore, we should not be surprised if there is a further fall in prices in the period ahead. However, a collapse that would even approach the magnitude of the collapse experienced during the1990s, when real house prices fell by almost 30 per cent from their highest to their lowest level, is out of the question according to our assessments.
Although experience shows that house prices often covary between countries, there are several factors that indicate that house prices will fall somewhat less in Sweden than in other countries. Among other things, Sweden's investments in housing in relation to GDP have been at a very low level compared to those in countries such as Denmark, Spain and the USA. There is also a shortage of housing in approximately half of the municipalities of Sweden. Another factor that indicates this is that Swedish households buy their houses to live in, not to rent out. In the United Kingdom, where house prices are now falling substantially, there has been a large market for buy-to-let. This means that a house or an apartment can be purchased to be rented out on a commercial basis. This element does largely not exist on the Swedish housing market.
An important question with regard to financial stability is to what extent falling housing prices will affect the loan losses of the banks. We do not believe that this will lead to any significant increase in the credit risk on the bank's balance sheets as our assessment is that the households will still be able to service their debts. In recent years, the Riksbank has devoted a lot of work to analysing how the Swedish household sector's debt servicing ability is affected by changes in interest rates and unemployment. Our analysis is based on so-called micro data; that is data relating to individual households. The analysis shows that the ability of Swedish households to service their mortgages is not particularly affected by higher unemployment. This is due to a combination of an extensive social safety net and the fact that the largest part of the household sector’s debt is held by households with two wage-earners. There are of course exceptions, for example borrowers that for one reason or another are not entitled to unemployment benefit or households where both wage-earners have the same employer, which increases the risk that both will become unemployed at the same time. However, despite these exceptions, we find it difficult to see that the ongoing recession could lead to a situation in which many households would no longer be able to service their loans. This conclusion is also supported by Sweden's experience during the 1990s when only a very limited proportion of the banks' loan losses could be related to the household sector despite the substantial fall in house prices and the rapid increase in unemployment.
At present our policy rate, the repo rate, is extremely low and in this context I would like to emphasise for presumptive house buyers that at the moment the interest rate is much lower than the average one would normally expect over an business cycle. It is important to remember that mortgage rates will go up when the economy recovers. People should bear this in mind when calculating their potential housing costs.
The prices of commercial properties are falling
It we now shift our focus to the market for commercial properties, the first thing we can see is that prices have fallen. More precisely, the prices of commercial properties fell by eight per cent in 2008 compared to 2007.(2) It is not only in Sweden that prices have fallen and, historically speaking, there is a high degree of covariance in prices on the property markets between countries. Examples of commercial property markets where prices have fallen more than in Sweden are those in Ireland, the UK and Norway. On these markets, prices have fallen by between 10 and 30 per cent. The prices of commercial properties have also fallen dramatically in the Baltic countries.
The category of commercial properties that was hit hardest in the crisis of the 1990s was that for office premises. It is therefore interesting to compare how this market has developed during the current recession compared to the crisis of the 1990s. The real price for office premises in the centre of Stockholm fell by almost 70 per cent from its highest level in 1989 to its lowest level in 1993. If we compare this with recent developments we can see that the corresponding prices fell by almost 15 per cent between the fourth quarter of 2007, when the highest price was recorded, and the first quarter of 2009.(3) So, although we can see that there are major price falls at present they are, so far, much more modest than the falls recorded during the crisis of the 1990s. And even if it is reasonable to believe that prices may continue to fall in the period ahead, we do not believe that we are looking at a collapse in prices that can be compared to this crisis.
Major differences compared to the crisis of the 1990s
Nowadays, it is common to attempt to compare the present situation with the crisis of the 1990s. Let us therefore take a closer look at the similarities and differences between developments on the property market now and during the property cycle in the 1980s until the point at which the crisis broke out. One of the important factors that made it possible for prices to increase so much during the second half of the 1980s was the deregulation of the credit market. As there was a high, pent-up demand for loans, this deregulation led to a rapid growth of credit. A large proportion of the new loans were property related. In 1992, lending to property represented almost one third of the major banks' total outstanding loans. In the late 1980s, there was also a rapid increase in the construction of new properties and when the crisis set in there was a dramatic increase in vacancy rates.
If we compare this with the situation today, we can see that although lending to the property companies increased in the early 2000s, it did not increase to at all the same extent as prior to the property crisis. Nor has the construction of new properties increased to the same extent as it did at the end of the 1980s. Vacancy rates did increase in the years following the turn of the century and the dotcom crisis, but they were not at all at the same levels as during the crisis of the 1990s. Furthermore, as a percentage of the major Swedish banks' total lending the Swedish property companies now represent a smaller proportion than they did then, even though they make up the largest business sector that the Swedish banks lend to.
However, the most important difference compared to the property crisis of the 1990s relates to the interest rate levels, which are now much lower as a result of the fact that today's monetary policy focuses on defending an inflation target rather than a fixed exchange rate. During the property crisis we saw high nominal interest rates which in combination with rapidly falling inflation led to high real interest rates. Many property companies were squeezed between falling incomes and the rapidly-rising interest rates. The costs of the property companies' loans increased significantly and this ultimately meant that several, primarily highly-leveraged, property companies found it difficult to service their debts. This led in turn to increased loan losses for the banks. Today, the interest rates that the property companies have to pay on their loans are much lower than they were in the 1990s. At the same time, the property companies in general are less leveraged than they were in the 1990s and their financial position is generally stronger now than it was then.
In the early 2000s, the banks also took a different line with regard to credit assessments than they did before the property crisis of the 1990s. Then, the banks largely focused on the values of the collateral, values that were quickly undermined as prices fell. Now, the banks focus more on the customers' cash flows and thus on their ability to service their debts.
In this context it is also worth noting that there appear to have been more foreign investors and financiers in the 2000s than there were in the 1990s. When we now talk about the fact that it is more difficult to find funding as a result of the financial crisis, it also appears to be primarily the foreign investors and financiers that have withdrawn from the market. It is also the foreign investors that have had the highest proportion of borrowed capital in their purchases, which also means that they are the ones who have been hit hardest, relatively speaking, by the financial crisis and the tight funding situation.
…but also certain similarities
Although there are major differences between the situation then and now, there are also a number of similarities. In simple terms, we could say that the price of a property is determined by two factors: the expected net operating income for the property and the return requirement of the investor. The fact that some investors are prepared to pay more for a property than others may therefore either be because they expect the rental income to increase or because they do not require the capital they invest to generate as much in return. From the mid-1980s to 1990, prices increased much more than rents. The same pattern was repeated during the period of price increases from 2005 to 2007. We can thus say that the development of prices during these two periods was not driven by what we usually refer to as fundamental factors. That is, prices did not increase as a result of an increase in the demand for office premises, lower vacancy rates and higher rents. The price increases were instead due to the fact that investors no longer demanded the same high level of return on their investments, and at the end of 2007 risk premiums were almost zero. Developments between 2005 and 2007 were, however, not at all as dramatic as they were in the 1980s, when the property bubble grew. The yield levels in the 1980s were sometimes even lower than the yield on a government bond, which meant that those who invested in commercial properties did not require any compensation for the risk they took for investing in property.
Another similarity between then and now is how events on the financial markets can affect the commercial property market. One could perhaps say that the crisis on the property market began in the autumn of 1990 when the finance company Nyckeln was unable to refinance its loans. When Nyckeln and the other finance companies could no longer continue lending to the property companies, this marked the beginning of the end for credit-driven price increases. As in the 1990s, the current financial crisis has made it more difficult for the property companies to find funding.
It should, finally, be mentioned that although the banks at the overall level, that is not only taking property-related loans into account, are facing substantial loan losses. The Riksbank's assessment in the latest Financial Stability Report is that the loan losses in the main scenario will amount to just over two per cent of the banks' lending in 2009 and 2010. In a stress scenario that portrays a much weaker development of the real economy, it is calculated that the losses would amount to 4.3 per cent of lending in the same period. These are major losses, but still much more limited than the losses made in the crisis of the 1990s. The loan losses in the period 1991 to 1994 amounted to 13 per cent of average lending in the same period and, as I mentioned earlier, two-thirds of the losses were linked to property-related loans. This time, the single largest source of losses consists of loans to the Baltic countries and the rest of Eastern Europe. In both the main scenario and the stress scenario, it is calculated that the losses associated with these loans will account for about 40 per cent of the total losses.
No new 1990s crisis, but still an uncertain situation
There are thus several reasons for the assessment that we will avoid a repetition of the type of property crisis that we saw in the 1990s; interest rates are more favourable today, the current situation has not been preceded by the kind of building boom that preceded the crisis in the 1990s and the banks' lending now focuses more on the debt servicing ability of the property companies than on the value of their assets and leverage levels are now generally lower than they were in the 1990s. Nevertheless, there is still a great deal of uncertainty on the market and it is worth highlighting the risks that we see in the period ahead.
Firstly, we are experiencing a serious recession and it is difficult to predict exactly how this will affect the demand for premises. An important factor affecting the demand for office premises is the level of employment in the service sector. As in the case of employment in general, we have seen a decline in service-sector employment. On the other hand, the decline in the service sector has come later than it has in the industrial sectors. Our assessment is, however, that employment in the service sector will decline further. Given this background, it is reasonable to expect that we will see little or no increase in rental incomes and an increase in vacancy rates. The earnings of the property companies and their ability to service their loans will thus deteriorate and there is a risk that there will be more defaults also on the part of property properties.
Secondly, falling property prices may become a problem for property companies when they have to renew their loans at the banks if they are unable to meet the banks' demands concerning the maximum permitted loan-to-value ratio.
Thirdly, the increase in the presence of foreign banks on the Swedish property market in recent years also gives some cause for concern. This is because these banks are now withdrawing from the Swedish market as a result of the financial crisis and are not renewing their loans here. The Swedish listed property companies appear admittedly on the whole to be less dependent on foreign bank funding. Nevertheless, the other players that are dependent on this type of funding may find it difficult to refinance their loans or get funding from Swedish banks. The fact that foreign financiers are pulling out could also be a further cause for concern on the Swedish market for commercial properties in general and make this market appear to be less attractive.
Fourthly, the interest expenditures of the Swedish property companies may increase more than their earnings when economic activity picks up and interest rates thus rise. This could make it more difficult for them to service their loans.
Concluding remarks
We published a new Financial Stability Report a couple of weeks ago. Compared to our assessment in the November report, we no longer see the liquidity risk as the main risk for the banks. Important reasons behind the decline in the liquidity risk include the measures taken by the Riksbank to promote liquidity and the government's guarantee programme. Following the rapid downturn in economic activity and the accompanying decline in the prospects of the banks' borrowers, the credit risk has instead taken over as the most prominent risk.
Our assessment is that the banks are facing substantial loan losses but that they will meet their capital requirements even when they are tested against a very severe stress scenario.
I refer those who are interested in our overall analysis, the stress tests and how we assess the risks for different categories of borrower to the report. I would just like to repeat what I said earlier: there is little to indicate that the Swedish household sector as a whole will give rise to any significant loan losses. On the other hand, the high level of uncertainty regarding economic prospects means that the credit risk in the business sector at large has increased. This also applies to your particular sector.
We are now experiencing a crisis that is exceptional in terms of both its depth and geographical distribution. This raises a number of questions. When will the financial markets return to normal? When will our economies recover, and will the recovery last? If an economic upturn does come, how strong will it be? These questions will be answered eventually and then we will see, to return to Danny Kaye, whether happiness will return for the property owners.
(1) See the Riksbank (2009), "Global recession and financial stability”, Financial Stability Report 2009:1.
(2) Data from Svenskt Fastighetsindex/Investment Property Databank.
(3) Data from Newsec. The real price is defined as the nominal price divided by the consumer price index.
