Financial Stability 1/2003
3 Macroeconomic developments and credit
risk

3.1 The macroeconomic environment
3.2 Credit risk associated with loans to the household sector
3.3 Credit risk associated with loans to the corporate sector
Ramme:
- Indicators of the price level in the housing market
3.1 The macroeconomic environment
The Norwegian economy has recorded strong domestic cost inflation for some
time. Profitability in internationally exposed industries was reduced further
as a result of the appreciation of the krone through 2002. Moreover, markets
for Norwegian export goods have been characterised by low demand and falling
prices. Many Norwegian enterprises have laid off employees, closed down operations
or signalled plans to relocate production abroad. Seasonally adjusted figures
show a reduction in manufacturing production of about 1% in the first quarter
of 2003 compared with the fourth quarter of 2002. At the same time, the decline
in production in service industries has lasted longer than assumed earlier.
LFS unemployment has risen from 3.6% in 2001 to a seasonally adjusted 4.1%
in February 2003. The business sector's expectations concerning future economic
developments have been lowered (see Chart 3.1).
Chart 3.1
Sharp wage growth in 2002 resulted in higher household income. This provided
scope for both strong consumption growth and higher saving according to Statistics
Norway’s latest national accounts figures. However, these figures are
highly uncertain. High electricity prices last winter are resulting in lower
real income growth this year. Moreover, as a result of the rise in unemployment
and the slowdown in the level of economic activity, consumer confidence concerning
developments in the country’s economy has been reduced (see Chart 3.2).
Chart 3.2
Gross fixed investment fell in 2002, particularly in service industries. Manufacturing
investment, on the other hand, increased, primarily as a result of the upgrading
of production facilities in the metal industry.
Growth in mainland GDP slowed from 1.7% in 2001 to 1.3% in 2002. In the March
2003 Inflation Report, growth was projected to be the same in 2003 before
rising gradually in the period to 2005. The growth forecast for the Norwegian
economy has been revised downwards compared with the previous Inflation Report
(see Table 3.1). This can be ascribed to both weaker global growth and reduced
confidence in the future among households and enterprises.
Table 3.1
A sharp fall in asset prices may constitute a threat to financial stability.
The oil price is an important asset price for Norway. Petroleum wealth, measured
as the present value of the future return in excess of a normal capital return
on oil production and pipeline transport, is about 20% higher than the total
value of the housing stock. A strong oil price shock could lead to considerable
adjustments among enterprises, households and foreign operators. The oil price
has declined following the war in Iraq and the normalisation of the situation
in Venezuela. However, the oil price is still high in a historical context.
The decline is therefore not likely to result in substantial adjustments.
Growth in total credit to the general public (municipalities, non-financial
enterprises and households) has fallen slightly in the last half-year. In
February 2003, year-on-year growth in total credit to the public stood at
7%. Domestic credit, which accounts for about 80% of total credit, rose by
8.7% from March 2002 to March 2003. Growth in credit from foreign sources
came to a halt towards the end of 2002. In February 2003, the year-on-year
growth was 0.6%. Credit growth is far higher for households than for enterprises
(see section 3.2 and 3.3 below). In 2002, total credit rose marginally as
a share of GDP (see Chart 1.5). Domestic credit as a share of mainland GDP
has also risen.
Chart 1.5
3.2 Credit risk associated with loans to the household sector
Continued high growth in household debt…
Weaker economic growth and higher unemployment have not curbed growth in household
debt to any extent. In the year to end-March 2003, household gross debt grew
by 10.8%, compared with 11.2% six months earlier (see Chart 3.3).
Chart 3.3
…has contributed to lower net financial wealth
High debt growth since 1999, combined with the decline in equity prices since
spring 2002, has resulted in a reduction in household net financial wealth
(see Table 3.2). At the end of 2002, net financial wealth less insurance claims,
which are illiquid, amounted to a negative NOK 168bn, NOK 142bn lower than
at the end of the first quarter of 2000.
Table 3.2
Household debt increased by nearly NOK 100bn in 2002. A difference of NOK 30bn
between estimates in the national accounts and financial market statistics
for household net investments in financial assets makes it difficult to pinpoint
this debt growth. According to the national accounts, net investments in financial
assets rose sharply last year, with growth in financial assets exceeding debt
growth by NOK 30bn. Lower figures in financial market statistics for net investments
in financial assets may indicate that income growth has been lower, or growth
in consumption and fixed investment higher, than the national accounts data
imply.
Indicators of the price level in the housing market
A sharp rise in private sector debt and asset prices is an early warning
of a potential financial crisis.1 The
IMF has shown that bubbles that burst in the housing market result in a
financial crisis more often than stock market bubbles and also have greater
negative real economic consequences.2
It is therefore important to follow house price developments.
1 Borio, C. and Lowe, P. (2002): “Asset
prices, financial and monetary stability: exploring the nexus”,
BIS Working Papers No. 114.
2 World Economic Outlook, April 2003.
House prices in Norway rose by an annual average of 9.1% between 1995 and
2002. In other European countries, like Ireland, the Netherlands and the
UK, annual growth has been more than 10%.
An initial approach for evaluating the level of house prices is to look at
them in relation to develop-ments in building costs and wage growth (see
Chart 1). Deflated by the building cost index, prices for existing dwellings
in Norway are now about 20% higher than the peak level in 1987. If, on the
other hand, a wage index is used as the deflator, real prices are still
somewhat lower than the peak level.
Chart 1
Another approach is to consider the price of a dwelling as the sum of discounted
expected profit flows associated with the dwelling, i.e. the value of housing
consumption or rental income. In the same way as for equities, it is possible
to calculate a P/E ratio for dwellings by dividing average house prices
by average rental prices.3
1 Leamer, E.E. (2002): “Bubble
trouble? Your home has a P/E ratio too”, UCLA Anderson Forecast,
June and Krainer, J. (2003): “House price bubbles”, FRBSF
Economic letter, No. 2003-06, March 7.
Due to the lack of good, comparable time series data for the level of house
and rental prices, we have constructed two time series for house prices
and rental prices. House prices are based on data for Statistics Norway’s
average square metre price for existing detached houses and adjusted using
the rise in Norges Bank’s house price index for existing dwellings.
Rental prices were obtained from Statistics Norway’s Living Conditions
Survey for 2001. Rental prices are adjusted using the rise in paid rent
from the CPI. Nominal rent increased steadily from the trough in 1994 up
to 2002, but appears to have levelled off somewhat in 2003. It is nevertheless
far below the previous peak level in 1988 when rent showed an annual rise
of close to 8%. At the end of March, the year-on-year rise in paid rent
was marginally higher than the year-on-year rise in the CPI.
The P/E ratio for the Norwegian housing market has been rising over the past
ten years (see Chart 2). At the end of 2002, the P/E ratio was 6.5% higher
than at the peak in 1988. If the P/E ratio is again to reach its long-term
average, house prices would have to fall considerably and/or rental prices
would have to increase. Households are willing to pay more for dwellings
when it is cheaper to service debt. In isolation, it is therefore reasonable
to assume that the P/E ratio rises when interest rates fall.
Chart 2
The overall impression from several indicators is that the price level in
the housing market is historically very high.
Higher housing wealth results in higher total wealth
Despite the decline in net financial wealth, the rise in housing wealth has
resulted in markedly higher total household wealth than at the end of the
first quarter of 2000. Pressures in the housing market were reduced over the
past year. House prices fell by 0.9% in the year to April 2003 (see Chart
3.3).
Chart 3.3
Why is household borrowing so high?
Dwellings are the most important component of household wealth. It is natural
to look upon high growth in credit to house-holds in connection with historically
high house prices. Since 1993, the value of the housing stock has risen substantially
and is now almost as high in relation to income as at the end of the 1980s
(see Chart 3.4). Up to three years ago, the value of housing also increased
at a faster pace than debt. Household debt as a share of housing wealth (debt
ratio) is therefore far lower than was the case ten years ago (see Chart 3.5).
Chart 3.4 Chart
3.5
Homeowners who have seen a rise in the value of their dwellings have the opportunity
to raise additional loans with the dwelling as collateral for consumption
purposes, or perhaps to invest in financial or fixed assets. This has probably
occurred to some extent. However, it is likely that households have not continuously
adjusted consumption and saving, and the composition of financial balances,
on the basis of movements in house prices. Adjustments in financial balances,
and particularly debt, have been made to a greater extent through purchases
and sales of dwellings.
When house prices have previously shown a strong and prolonged rise, both debt
and financial assets can increase when dwellings are sold. This can be illustrated
by looking at a transaction between two households that have owned dwellings
(at different prices) for a long time. Both will probably have a low debt
ratio initially. The buyer of the more expensive dwelling will probably have
to debt-finance a large part of the price difference. The seller of the more
expensive dwelling does not have as much debt to repay and can use the funds
made available for consumption, or perhaps investments in financial or fixed
assets. The debt ratio of the two households as a whole will probably increase.
As dwellings are sold in the market, the debt ratio will gradually be adapted
to the previous rise in house values. If turnover in the housing market is
maintained, we may thus experience a period of higher credit to households
even if house prices remain stable.
Turnover figures showed a slight rise in 2002 (from an already high level)
(see Chart 3.6). According to figures from the real estate industry and the
Oslo Housing and Savings Society (OBOS) for the first four months of 2003,
turnover was as high as in the same period last year. This has probably contributed
to continued high credit growth in the first few months of 2003.
Chart 3.6
Debt burden earlier and now
Even though households have solid collateral in the form of housing wealth,
they are limited by developments in income and interest rates. Interest and
capital payments on debt must be serviced by the income available to the household.
If debt increases in relation to disposable income, borrowers become vulnerable
to negative macroeconomic shocks.
The debt burden of households has risen sharply since 1999 (see Chart 3.7)
and is approaching the peak level recorded during the previous banking crisis.
With a view to determining the debt level that can lead to debt-servicing
problems, it is probably most relevant to compare the debt burden at a time
when considerable problems arose. Some indicators may suggest that the critical
debt burden was reached some time before the peak.
Chart 3.7
On the other hand, there are factors that imply that households today can withstand
a higher debt burden than during the banking crisis. If household debt is
seen in relation to dispos-able income less expenses for “necessities”,
we obtain a measure of how large the debt is in relation to the “remaining
income” that is available to service the debt. As a very rough measure
of necessities, we have selected expenses for food, non-alco-holic beverages,
clothing and footwear. As a share of total con-sumption, expenses for these
consumer goods have declined substantially since the 1980s. This means that
the “remaining income” has risen at a faster pace than disposable
income. The red curve in Chart 3.7 shows that the debt burden is lower historically
when expenses for necessities are deducted from disposable income. It is also
normal to assume that households want to use an increasing share of income
on dwellings when income rises. This will result in a higher debt burden.
Chart 3.7
The change in monetary policy regime from an exchange rate target to an inflation
target has probably made it less likely that households will be exposed to
a “double shock” in the form of higher unemployment and higher
interest rates, as was the case during the banking crisis. In isolation, this
may mean that households can withstand a somewhat higher debt burden than
prior to the banking crisis. The debt burden of Norwegian households is high
by international standards (see Chart 3.8). However, structural differences
across countries, for example in relation to the scale of owner-occupied dwellings,
make it difficult to make such comparisons.
Chart 3.8
Household interest expenses accounted for a little more than 7% of cash income
(disposable income plus interest expenses) at the end of 2002 (see Chart 3.9).
The reduction in interest rates since last autumn will reduce the interest
burden in 2003.
Chart 3.9
Considerable differences between different households
The financial situation of the household sector as a whole is satisfactory.
However, there are considerable differences between groups of households.
Some groups have very high debt in relation to income and will thus be in
a high-risk position in terms of debt-servicing problems. In earlier reports,
we have pointed to the vulnerability of particularly exposed households with
high interest expenses in relation to income. As shown in the November 2002
Financial Stability report, about 30% of total household debt of a little
more than NOK 1 000bn is attributable to households with an interest burden
of more than 20%. In addition to having a high interest burden, these households
have small reserves in the form of financial assets (see Chart 3.10). This
applies in particular to groups with a high interest burden and low and middle
income. For example, households in the income decile 5 with a high interest
burden have on average financial reserves that amount to only 8% of their
total debt.
Chart 3.10
Developments ahead
The projections for the next two years are based on the assumptions underlying
the March 2003 Inflation Report and an unchanged sight deposit rate of 5%.
Moreover, it is assumed that debt growth in the household sector will be gradually
reduced from the current level and will be equal to nominal income growth
at the end of 2004 and through 2005. Under these assumptions, the debt burden
will increase at a somewhat slower pace than in recent years, but will nevertheless
reach a historically high level at the end of 2005 (see Chart 3.7). Relatively
low interest rates entail somewhat lower interest expenses as a share of cash
income at the end of 2005 than at the end of 2002 (see Chart 3.9).
Chart 3.7 Chart
3.9
All in all, the financial situation of the household sector has deteriorated
over the last six months. The wealth position of the household sector as a
whole is nevertheless satisfactory. However, high house prices and house sales
continue to stimulate strong growth in debt from an already high level. A
situation where debt growth is much stronger than growth in income cannot
be sustained over time. Credit risk associated with loans to the household
sector is therefore considered to be somewhat higher than in the November
2002 Financial Stability report. It is, however, still moderate.
3.3 Credit risk associated with loans to the corporate sector
Substantial rise in the number of bankruptcies…
Sluggish trends in the Norwegian economy through 2002 contributed to an increase
of 25% in the number of bankruptcies from 2001 to 2002. The number of bankruptcies
continued to rise in the first quarter of 2003.
Many of the enterprises that declared bankruptcy are small unincorporated firms.
Developments in 2002 indicate that larger enterprises are also declaring bankruptcy.
Measured by the number of employees in bankrupt enterprises and these enterprises’
market value, the increase is greater than implied by the number of bankruptcies
(see Chart 1.1). The number of employees in enterprises that went bankrupt
was a little less than 14 000, an increase of 48% compared with the previous
year. The market value for enterprises that went bankrupt in 2002 showed a
rise of 74%. Measured by total sales, the increase is highest in manufacturing
(see Chart 3.11). The property management, commercial services and rental
activities sector and the sectors construction and hotels and restaurants
also showed a considerable increase.
Chart 1.1 Chart
3.11
Developments in the number of bankruptcies show that the internationally exposed
sector is experiencing greater problems than sheltered sectors. The rise in
the number of bankruptcies was particularly pronounced in manufacturing. However,
bankruptcies also showed an increase in more sheltered industries such as
property management, construction and hotels and restaurants.
...is related to previously high level of investment activity and debt growth
Debt-servicing problems and the number of bankruptcies often increase following
a period of high investment growth. This also appears to be the case now.
In many industries, investment as a share of real capital rose markedly after
the cyclical upturn began in 1993. Investment peaked in 1997-1998 (see Chart
3.12). In a number of industries, the investment rate was higher during this
upturn than in the period leading up to the banking crisis in the early 1990s.
Strong investment activity often takes place as a result of favourable macroeconomic
conditions and considerable optimism. Investment decisions may be made which
result in an excessive stock of real capital when the downturn occurs. Problems
are greatest for those enterprises that have shown the greatest willingness
to take risks and in industries where investment has resulted in the most
excess capacity.
Chart 3.12
The growth in corporate borrowing from banks is still low and must be seen
in connection with investment activity (see Chart 3.13). Investment declined
in most industries in 2002. Growth in loans to service sectors and property
management has been higher than growth in loans to other sectors, but growth
has slowed over the last five months. Growth in loans to retail trade and
hotels and restaurants has been negative since October 2002. Growth in loans
to manufacturing is also low.
Chart 3.13
Continued high risk of defaults
Moody’s KMV model shows that the probability that large unlisted enterprises
will default on their obligations is at the same level as in the November
2002 Financial Stability report (see Chart 3.14). Default probabilities in
Norway have moved on a rising trend since 2000, as has been the case in the
other Nordic countries. In March 2003, Norway had the second highest default
risk after Sweden, measured by the default probability for the median enterprise
(see Chart 3.15). The change in default risk1
sprimarily reflects changes in enterprises’ total value as a result
of fluctuations in stock markets.
Chart 3.14 Chart
3.15
1 The effect of changes in market values
and volatility on default risk
was discussed in the November 2002 report.
Fish farming is an industry that was marked by debt-servicing problems and
high default risk in 2002. A number of enterprises in the fish farming industry
recorded low earnings, partly as a result of low prices. Combined with a high
level of borrowing in recent years, this resulted in debt-servicing problems
for a number of enterprises. In Pan Fish, which is one of Norway’s largest
fish farming companies, the largest banks converted debt into equity and supplied
new equity capital in January 2003. This contributed to more stable operating
conditions for the company. In general, mergers and acquisitions of small
fish farming enterprises have contributed to increasing the level of debt
in the sector. Many of the enterprises with an expansionary strategy have
experienced problems. Debt restructuring and the supply of new equity capital
have improved the debt-servicing capacity of the industry. Developments ahead
will largely depend on developments in salmon prices.
Uncertainty associated with commercial property companies
The commercial property industry is the single largest recipient of bank loans.
At the end of 2002, these loans accounted for 11.5% of total bank lending,
which is more than a doubling over the last ten years (see Chart 3.16).
Chart 3.16
Developments in profitability in the various segments of the commercial property
market have varied. Rental prices for the segments shopping centres, shops
and restaurants showed positive developments in 2002. The hotel market, on
the other hand, has been negatively influenced by a decline in the number
of guest nights. The number of guest nights in February 2003 was 7% lower
than one year earlier. The market for office rentals is marked by an increase
in available commercial space and falling prices. The difference between the
market segments is also evident in investment. Figures from Statistics Norway
show a decline in the total floor space of commercial property starts of altogether
10.1% in the last two years. Property starts have increased in the production-oriented
segments manufacturing and warehouses and fisheries and agriculture. Other
types of buildings have shown a decline, particularly in the segments office
and commercial buildings, transport and communications buildings, hotel and
restaurant buildings and school and cultural buildings.
Developments in the market for office premises have been weak. Demand for office
premises is influenced by developments in the number employed in office-intensive
occupations. In the Oslo area, demand for office premises has fallen, partly
as a result of a reduction in the number employed in the IT and consultancy
sector. The market value and rental prices of office premises have fallen
in recent years (see Chart 3.17). According to Eiendomsspar’s study
of the market for office buildings in Oslo, Asker and Bærum, space absorption
(space used per employee) for the total stock of office buildings fell for
the first time since measurement began in 1992. Available space rose from
5% in February 2002 to 10% in February 2003 (see Chart 3.18). Eiendomsspar
expects a decline in completed projects in this region through 2003. In isolation,
this will reduce growth in the supply of space.
Chart 3.17 Chart
3.18
At the end of 2002, rental prices for standard office premi-ses had fallen
by 21% since peaking in 2001. During the previous downturn, rental prices
fell by 41% from the peak in 1987 to the trough in 1993. With continued weak
developments in the Norwegian economy, available space will remain high and
lower rental prices may be expected. A rising share of property companies
may experience problems in servicing their debt. Most exposed are enterprises
with high debt and a short remaining term for rental contracts. However, the
real value of rental prices and property values are considerably lower than
at the end of the 1980s. This implies that the potential for a similar decline
is lower now than at that time.
Revision in indicators of debt and interest burden
Ratios of interest and debt to the cash surplus are important indicators of
enterprises’ debt-servicing capacity. The cash surplus is derived from
ordinary operations and from financial assets (capital income). Gross capital
income consists of interest, share dividends and other income from enterprises’
financial investments. New revised national accounts figures from Statistics
Norway for 2000 and 2001 show a sharp rise in gross capital income due to
an increase in dividends received. At the same time, enterprises’ dividend
payments have risen considerably. The figures now reflect, to a greater extent
than earlier, transfers between enterprises. The result of the revision is
a substantial downward revision of enterprises’ total debt and interest
burden in 2000 and 2001 based on gross capital income (see Charts 3.19 and
3.20).
Chart 3.19 Chart
3.20
On the basis of net capital income, which shows a more steady development than
gross capital income, we find that the debt and interest burden increased
in 2002. The debt burden is now just as high as in 1990. In 2002, the interest
burden was at its highest level since 1993. In isolation, the recent reduction
in interest rates will reduce the interest burden. The extent to which weak
borrowers gain the full benefit of the reduction in interest rates will depend,
however, on any changes in the risk premium demanded by banks.2
2 For a discussion of banks’ pricing
of risk, see the November 2002 Financial Stability report.
Developments ahead
In the November 2002 Financial Stability report, credit risk was considered
to be relatively high for internationally exposed enterprises and moderate
for enterprises in sheltered industries. The recent depreciation of the krone
will help to improve the competitiveness of export-oriented industries. Weaker
growth prospects in Norway and internationally will have the opposite effect.
All in all, credit risk associated with loans to enterprises in the exposed
sector is considered to be unchanged. Credit risk associated with loans to
commercial property companies has increased, especially for enterprises that
rent out office premises. Credit risk associated with loans to other enterprises
in the sheltered sector has also increased, but the risk remains moderate.