The credit and housing boom in Sweden, 1995-2015: Forewarned is forearmed

Fredrik Andersson, Lars Jonung

30 May 2016

a

A

The volume of credit to households has grown rapidly since the mid-1990s in Sweden—twice the rate of nominal economic growth. The long-lasting credit boom has been accompanied by rising house prices and mounting household debt ratios. The central bank, the Riksbank, presently pursues a very expansionary policy with negative interest rates and massive purchases of government bonds despite high growth—amongst the highest in Europe. The credit boom raises a number of questions. Can a hard landing be avoided? Will this time be different for Sweden?1

Financial deregulation and boom-bust 1985-1995

The Swedish banking system was heavily regulated by a comprehensive set of controls of domestic and cross-border flows of credit until 1985. The Riksbank held the total volume of credit roughly at a constant level in relationship to GDP (Figure 1). This was a period of considerable financial stability with fairly small movements in real stock prices and house prices.

Figure 1. Total private non-financial sector debt and household debt, 1961Q1–2015Q3

Source: BIS and the Swedish Riksbank

Financial deregulation in the second half of the 1980s changed this picture. The once stable credit ratio increased rapidly. The outcome was a boom-bust cycle from 1986–94 with devastating macroeconomic consequences. The deep depression forced the Riksbank to abandon the pegged exchange rate for the Swedish krona. The krona has been floating since then. In January 1993, the Riksbank adopted an inflation target of 2%.2

A new credit cum asset boom 1995-2015

Following the financial crisis of 1992-93, the Swedish economy started on a long recovery. Since the mid-1990s, total credit to GDP has been on an upward trend, almost doubling from 1995 to 2008 (see Figure 1). Household debt to household income has steadily risen for more than 20 years, from 90% in 1995 to 179% in 2015. The Riksbank (2016) estimates this ratio will reach 193% by 2018. The Swedish level has far surpassed the peak of 2008 in the US and has now reached the same level of the UK in 2008 (see Figure 2).

Figure 2. Household debt in relation to disposable income in Sweden, the UK and the US, 1995–2014, with the Riksbank’s forecast for Sweden until 2019

Source: Andersson and Jonung (2015).

The rapid growth in credit is accompanied by a sharp rise in asset prices. Real house prices have grown annually by 4.9% and real stock prices by 6.7%, while economic growth has been 2.5% between 1995 and 2015. The growth in real house prices in Sweden for 1995–2014 surpasses that of the US and the UK by a wide margin (see Figure 3). Contrary to Sweden, these two countries experienced deep financial crises in 2008 which arrested the upturn in house prices. Although Sweden avoided a crisis in 2008, house prices declined slightly in 2008–09, before reverting into a new rise that has tended to accelerate in recent years.

Figure 3. Real house prices in Sweden, the UK and the US 1995Q1–2015Q3

Source: Andersson and Jonung (2015)

The role of monetary policy

The close association between rising house prices and a rising volume of credit raises the question of causality. Our econometric tests suggest that the expansion of credit has contributed to higher real house prices.3 The effect is most clearly discernible in the Stockholm region. There is also a feedback from rising house prices, increasing demand for credit. Other factors have impacted on house prices as well, such as restrictions on the supply of housing, rent controls, mortgage interest deductibility, population growth, and a strong rise in household income. But all in all, credit expansion is an important driver behind house price inflation.

The unprecedented growth of credit during the past 20 years is closely associated with an expansionary monetary policy. The Riksbank has reduced its repo-rate from 9% in 1995 to -0.5% in 2015 (see Figure 4). The expansionary stance is fuelled by Riksbank purchases of government bonds. By now, approximately 30% of all SEK denominated government bonds are held by the Riksbank. This level exceeds the purchasing program pursued by the ECB (Riksbank, 2016).

The effect of the falling policy rate on real house prices is illustrated by Figure 4. It displays real house prices in Sweden as a whole (black dotted line) and in Stockholm (black solid line) – where prices have increased the most – as well as the policy rate of the Riksbank. The blank areas in Figure 4 correspond to periods when the Riksbank lowered the policy rate compared to the previous year, and the shaded areas correspond to periods when the Riksbank increased the policy rate compared to the previous year. A clear pattern emerges—the growth in house prices stabilises or falls approximately six months after the Riksbank starts to increase the policy rate. House prices begin to rise six months after the Riksbank starts to lower the policy rate.

Figure 4. The policy rate (%) of the Riksbank (right-hand axis) and real house prices (index=100 for 1995), (left-hand axis), 1995–2015

Source: Statistics Sweden.

The expansionary Riksbank policy is explained by its adherence to the inflation target of 2%. Low inflation has caused the Riksbank to lower its policy rate despite the stimulating effects on debt and house prices. Since the early 2000s, members of the Riksbank board have been concerned by a possible conflict between financial stability and price stability. In the past five years, tension has been running high within the board.4 One group has promoted leaning against the wind while another group has argued for the sole focus on the 2% target. Recently, the majority has explicitly moved to the position that the inflation target should be the one and only goal of the Riksbank.

Where is Sweden heading?

Presently the Swedish economy is booming while fiscal and monetary policy remains highly expansionary and financial imbalances are increasing. How will this end? Of course, we do not know. Financial history tells us, however, that periods of a prolonged rise in credit and asset prices usually lay the ground for a correction, often in the form of a financial crisis.5

We can calculate the size of a likely correction by estimating a long run sustainable debt level. Households’ interest ratio has remained stable at 4% since the early 1970s in Sweden, with an exception of the financial crisis during the early 1990s. By assuming this share is the preferred long run interest ratio in the future, and by assuming that the post-tax interest rate returns to the pre-crisis average of 3%, the sustainable long-run debt ratio for households would be around 130% of disposable income. This number should be compared to the present level that is close to 180%.6

This simple calculation suggests a reduction by 50 percentage points, corresponding to six months’ disposable income for the average household, in order to get back to a sustainable path. For households with large loans taken in recent years, the adjustment would be much bigger. A decline by 50 percentage points exceeds the reduction in the interest ratio during the Swedish financial crisis in the early 1990s as well as the correction in the US and the UK after the 2008 financial crisis.

We do not see that such a correction can take place through higher economic growth. Growth has been high in Sweden in recent years while the household debt ratio has been rising—not falling. Rather, the most likely adjustment will take place through higher real interest rates, most probably through rising international rates being transmitted into Sweden. Rising interest rates will induce households to amortise more on their loans to reduce the burden of debt. The risk with such a process is that it can turn into a bust with falling consumption, rising savings and rising unemployment. Falling house prices will foster a downward cumulative spiral, well-known from the history of financial crises.

Will this time be different?

There is a growing awareness about the risks stemming from two decades of consistently growing debt ratios and house prices. But this awareness has not yet been transformed into decisive policy action. New measures concerning amortisation rules, soon to be enacted to reduce household indebtedness, may actually contribute to the crash they are intended to prevent. Of course, we cannot rule out the possibility that the Swedish authorities may find a way to arrest the house price boom without bringing about a bust and a financial crisis. If they do, this time will be different.

References

Andersson, F NG and L Jonung (2015) “Krasch, boom, krasch? Den svenska kreditcykeln” (“Crash, boom, crash? The Swedish credit cycle”), Ekonomisk Debatt, 43(8): 17-31.

Goodfriend, M and M King (2016) Review of the Riksbank’s monetary policy 2010-2015, 2015/16:RFR7.

Jonung, L, J Kiander and P Vartia (eds) (2009), The Great Financial Crisis in Finland and Sweden: The Nordic experience of financial liberalization, Edward Elgar, Cheltenham.

Reinhart, C M and K S Rogoff (2009) This time is different: Eight centuries of financial folly, Princeton University Press.

Endnotes

[1] We allude here to the title of the monumental study of financial crises by Reinhart and Rogoff (2009).

[2] Finland, Norway, and Sweden went through roughly the same boom-bust cycle in the early 1990s. See Jonung et al (2009).

[3] See Andersson and Jonung (2015).

[4] Goodfriend and King (2016) have recently documented the lively clashes within the board.

[5] For the historical pattern, see for example Reinhart and Rogoff (2009).

[6] See Andersson and Jonung (2015) for the details of these calculations.   

a

A

Topics:  Europe's nations and regions Financial regulation and banking Macroeconomic policy

Tags:  Sweden, Riksbank, financial crisis, credit boom, asset boom, household debt, house prices, growth

Associate Professor, Department of Economics, Lund School of Economics and Management

Professor, Knut Wicksell Centre for Financial Studies, Lund University

Events