Financial Stability 1/2000
6 Credit risk
Norges Bank: Financial stability 1/2000
6.1 General credit developments
6.2 Credit risk in the household sector
6.3 Credit risk in the enterprise sector
Box - Housing market
6.1 General credit developments
In the surveillance of financial stability it is important to monitor
developments in credit markets and assess whether developments observed
may imply an increased risk of future imbalances. For the economy as a
whole, credit as a percentage of GDP provides an indication of the willingness
of the various actors to incur debt. High credit growth in relation to
GDP growth may indicate that there are growing imbalances between the
credit market and the real economy.
After expanding at a slow pace up to the end of 1996,
overall credit to the private sector and municipalities (C3) picked up
and reached a level that was considerably higher than total GDP growth
in the period to the latter half of 1999 (see Chart 6.1). Towards the
end of 1999, C3 growth contracted sharply, reflecting a marked fall in
the growth in credit in foreign currency.
Chart 6.2 shows that developments in credit in NOK and
credit in foreign currency have followed two markedly different paths
over the last two years. The domestic rise in interest rates in 1998 led
to a shift in demand towards credit in foreign currency. When interest
rates rose, it became relatively cheaper to raise loans in foreign currency,
while the Norwegian krone was expected to appreciate against other currencies.
The sharp growth in foreign currency loans slowed in 1999 as the differential
between Norwegian and foreign interest rates narrowed.
Twelve-month growth in credit from domestic sources
(C2) has picked up further in the first quarter of this year, reaching
9.1% at the end of March, compared with about 8% over the last four months
of 1999. Household debt has increased, while debt growth for enterprises
has been declining. At end-March, growth in household debt was 9.4% while
the figure for enterprises was 8.7%. Twelve-month growth in credit from
domestic sources (C2) has picked up further in the first quarter of this
year, reaching 9.1% at the end of March, compared with about 8% over the
last four months of 1999. Household debt has increased, while debt growth
for enterprises has been declining. At end-March, growth in household
debt was 9.4% while the figure for enterprises was 8.7%.
Growth in bank loans has been higher than growth in
total domestic credit (C2) in recent years (see Chart 6.3). Banks' capital
adequacy is solid (see Section 7), and will not restrain further growth
in bank lending.
Chart 6.1
Chart 6.2
Chart 6.3
6.2 Credit risk in the household sector5
At end-1999, household loan debt to domestic banks, mortgage companies
and finance companies came to NOK 573bn (see Table 6.1). This accounted
for a good half of total loans from these institutions. In addition, households
have raised loans in insurance companies and state lending institutions.
Household loan debt is primarily matched by the value of housing. Chart
6.4 shows that household housing wealth has increased at a markedly faster
pace than loan debt since 1993.
Household net financial assets have increased markedly
through the 1990s, as a result of both capital gains and positive net
lending. At the end of the fourth quarter of 1999, household net financial
assets stood at 61% of disposable income. In recent years, households
have maintained a high saving ratio and net financial investment to income
ratio in spite of the sharp rise in the value of housing and financial
assets. The trend in the US has been the opposite (see Section 3). So
far, we have avoided a situation where households finance increased consumption
by raising loans or reducing other assets in response to wealth gains.
This has contributed to maintaining a solid financial situation in the
household sector.
Although banks' loan losses on credit to the household
sector have historically been limited, this sector is of significance
to banks' credit risk. The main indicators used are household debt and
interest burden.6 House prices are also monitored,
as property is the most commonly used collateral for loans to households.
House prices are expected to continue to rise in the years ahead. A correction
in house prices will in isolation increase banks' credit risk.
Table 6.1
Chart 6.4
Chart 6.5
5Households comprise wage-earners,
pensioners, social security recipients, students, self-employed and private
non-profit institutions serving households.
6Debt burden is defined as household borrowing as a percentage
of disposable income. Interest burden is defined as households' gross
interest expenses (after tax) as a percentage of cash income (ie disposable
income before deduction of interest expenses).
Household debt burden increases
Household debt burden has been increasing in pace with growth in disposable
income in recent years. The prospect of zero growth in employment and
moderate wage growth is expected to result in low growth in household
real disposable income this year, at about 1½%. It is still assumed
that resale home prices and housing investment will continue to rise (see
Table 6.2). Both factors suggest stronger growth in household gross loan
debt than in nominal disposable income. The household debt burden is thus
expected to increase both this year and over the next three years (see
Chart 6.6). Despite an increased debt burden, the indicator will remain
at an historically low level. However, the trend represents a break with
the trend prevailing over the last 5-6 years. If resale home prices continue
to rise at the same pace recorded over the last year, the rate of increase
will be higher than indicated in Chart 6.6.
Household interest burden increased in 1998 and 1999
as a result of the rise in interest rates in the latter half of 1998 (see
Chart 6.7). The interest burden is expected to increase to about 6.5%
this year, and then edge down.7 This is an upward
revision compared with last year's December report and primarily reflects
the technical assumptions concerning higher interest rates.
Although household debt and the interest burden have
increased, the indicators remain at a relatively low level. However, there
are some households and groups of households whose interest burden is
higher than average. These groups may constitute a loss risk for banks.
According to Statistics Norway's income and wealth survey for 1997, about
5% of households have an interest burden of more than 20%.8
Further analyses show that this group accounted for 18% of total gross
debt in the household sector.9 In view of the banks'
low, and in some cases negative, losses on loans to households, only a
small share of this debt represents a loss risk for banks, even though
an interest burden of 20% may seem high. In addition to the interest burden,
factors such as income, gross financial wealth and collateral value will
influence the risk of incurring losses.
Table 6.2
Chart 6.6
Chart 6.7
7It is assumed that money market interest
rates move in line with market expectations as indicated by forward rates in March 2000.
8The data for 1984-1997 were presented in the December report.
9 The interest burden is defined as the ratio between interest
expenses and cash income. Interest expenses depend on the interest rate
level and size of debt. For a given interest burden, a household with
high income will therefore normally have a higher debt than a household
with low income. Similarly, a household with a high interest rate burden
will normally have higher debt than a household with a low interest burden,
given that they have the same income.
Loan security in the household sector
The most common collateral for loans to households is property. Mortgage-backed
loans as a percentage of total loans outstanding in commercial banks and
savings banks increased from 30% in 1991 to a stable level of about 45%
over the last four years (see Chart 6.8). In isolation, this has increased
banks' exposure to the housing market, making them more vulnerable to
a correction in house prices.
Price changes do not necessarily represent a problem,
providing creditors allow for the possibility of a substantial change
in the value of collateral. Surveys by the Banking, Insurance and Securities
Commission show that the share of housing loans in banks with a loan to
asset ratio in excess of 80% has decreased somewhat from 1997 to 1999.
Generally, extensive use of loans with a high loan to asset ratio without
additional collateral provide little buffer against changes in house prices.
Resale home prices have risen more or less steadily
since 1993 (see Chart 6.9) The average rate of increase10
on the previous year has ranged between 6% and 12% since 1993, and was
not particularly high, at 10.0%, in 1999 (see Chart 6.10). A faster rise
in house prices was recorded in the year to end-1999, primarily reflecting
the drop in prices in the latter half of 1998. The increase in housing
starts, among other things, is expected to lead to a decline in the annual
average increase in house prices to 5% as from 2001.
Chart 6.8
Chart 6.9
Chart 6.10
10Rise in average prices one year compared
with average prices the previous year.
Somewhat higher debt burden - but credit risk remains low
An increased debt burden this year, and to some extent in the years ahead,
will in isolation augment the financial exposure of households. Against
the background of the economic outlook with low unemployment, a high saving
ratio, stable house price developments and the assumption of stable interest
rates, this will only have limited impact on banks' credit risk.
An element of uncertainty is house price developments.
There are few signs of a marked fall in house prices with an associated
increase in credit risk. However, a continued sharp rise in house prices
at the rates recorded in recent years will increase the probability of
a correction.
Housing market
During most of the post-war period, housing was a stable asset with
a steady rise in value, partly due to extensive regulation of the
housing market and the credit market. After the deregulation of
these markets early in the 1980s, house prices have shown greater
volatility.
In the long term, the supply of housing will adapt
to demand, and house prices will tend to a level determined by site
prices and building costs. However, in the short term the supply
of housing will be approximately fixed as the supply of new dwellings
per year only accounts for a small share of the total stock of housing.
Resale home prices may thus show fairly substantial deviations from
the long-term trend in house prices, depending on developments in
demand.
From 1988 to 1993 house prices fell sharply (see
Chart 1). It is probable that the sharp price fall led to a decline
in resale home prices below a level adapted to long-term explanatory
factors. Demand showed a marked fall and the price drop was probably
reinforced by an inelastic supply, combined with the fact that the
housing market is characterised by historic price expectations.
Since 1993 resale home prices have risen more
or less continuously. Some of the increase is ascribable to a catch-up
in prices after the marked fall, in step with demand. Several factors
have contributed to boosting demand: higher disposable income among
households - partly due to increased employment and falling interest
rates, a higher supply of credit, a shift in cohabitation patterns
towards one-person households and a continued, albeit modest, population
growth. With the price fall in the early 1990s behind us to some
extent, it is not unlikely that expectations of future increases
in house prices have increased and are now contributing to rapidly
rising prices.
Chart 1 shows developments in real prices for
resale homes over the last 20 years. The shaded areas show periods
of continuously rising prices. The chart shows that the price turnarounds
in both 1987/88 and 1992/93 coincided with changes in interest rates
and unemployment. The fairly sharp rise in interest rates in the
autumn of 1998, in conjunction with deteriorating economic prospects
in the wake of the Asian crisis, only had a short-term dampening
impact on house price developments. The reasons for this were probably
that the risk of a severe downturn in the economy rapidly abated
toward the end of 1998, and general expectations of a future fall
in interest rates. Chart 2 shows that changes in expectations towards
the end of 1998 and into 1999 coincide with the changes in house
prices. This may indicate that housing demand, and thereby house
prices, were heavily influenced by shifts in expectations through
1998 and 1999.
Developments in the supply of housing, as indicated
by housing starts, have shown little response to the increase in
demand recorded since 1993. The annual increase in housing starts
has hovered around 20 000 annually over the last 4-5 years, which
corresponds to about 1% of total occupied dwellings. This is somewhat
higher than the number of housing starts in 1992-93 of around 15
000 per year, but still markedly lower than the annual average of
around 30 000 in the 1980s. The moderate level of housing starts
reflects the continued low level of resale home prices compared
with new dwellings. The price of existing detached dwellings as
a percentage of new detached homes rose for the country as a whole
from 77.5% in 1993/1994 to 84% in 1998/99. However, wide geographical
variations have been observed. In Oslo, for example, resale home
prices are now higher than prices for new dwellings.
Following the sharp price fall in the period 1988-1993,
the price rise subsequent to 1993 seems to have been a catch-up,
which is underlined by the narrowing of the difference between prices
for resale homes and new dwellings.
Over the last year, resale home prices have risen
sharply, particularly in Oslo and the other large cities. The sharp
rise in house prices is not auspicious in itself, as prices are
less suited as a source of information concerning the housing market,
ie it is more difficult to distinguish between different segments
of the housing market. Furthermore, sharp price increases, in conjunction
with historic price expectations among house buyers, may contribute
to a self-reinforcing price increase. Somewhat higher debt burden
- but credit risk remains low.
Chart 1
Chart 2
6.3 Credit risk in the enterprise sector
Improved prospects for enterprises...
Investment in the enterprise sector (private non-financial enterprises)
is projected to fall this year, which will have a dampening effect on
economic growth in the short term. Higher export prices and increased
productivity will increase profitability in the manufacturing sector.
Production and profitability in the service and retail industry will be
stimulated by stronger growth in private consumption, among other things.
This will lead to an increase in investment in 2001 and 2002.
...but the debt burden is rising sharply
Developments in the enterprise sector up to and including 1996 were characterised
by a marked increase in value added and gross operating profits, in addition
to moderate growth in total loan debt and bond debt (see Chart 6.11).
In 1997 and 1998, growth in value added was markedly lower while debt
rose sharply. During this period, wage growth was also high with an attendant
decline in gross operating profits. For 1999, preliminary data indicate
that value added in this sector increased markedly in line with that of
the mid-1990s. Combined with lower growth in wage costs than in previous
years, this led to sharp increases in gross operating profits.
At end-1999, the loan debt of private non-financial
enterprises came to NOK 369bn distributed on domestic banks, mortgage
companies and finance companies (see Table 6.3). Enterprises have also
raised loans in state lending institutions, foreign markets, and to some
extent in bond markets.
Growth in total debt in the enterprise sector is expected
to be 7% this year. This implies a marked decline in new enterprise debt
compared with 1999. Against the background of a more favourable real economic
outlook for the enterprise sector in the years ahead, debt growth is expected
to level off at 7% in the years ahead. The estimate is somewhat higher
than six months ago. Gross operating profits in the sector are expected
to show slightly slower growth than debt growth, which implies a higher
debt and interest burden for enterprises in the years ahead.
Experience shows that the enterprise sector is the main
source of loan losses for banks. In order to shed light on financial exposure
we look at enterprises' debt and interest burden11,
and risk classification of individual enterprises based on reported accounting
data.
The debt burden in the enterprise sector (see Chart
6.12) increased from 340% in 1997 to 570% in 1999. The debt burden is
now at the same level as at the end of the 1980s, which primarily reflects
the weak trend in enterprises' gross operating results for 1997 and 1998,
and the high growth in debt from 1997 to date. In spite of the positive
trend in gross operating results in 1999, the debt burden is still rising
as a result of higher interest expenses in 1999. Our calculations show
that the debt burden will probably continue to increase up to 2003, to
a level of more than 600%. This represents a substantial upward revision
compared with our last report where the debt burden was projected to level
off just below 500%. The reason for the upward revision is an extensive
revision of historical data, which now show a weaker trend in enterprises'
gross operating results in 1998, and higher debt growth in 1999 than estimated
in the last report. Debt growth in the enterprise sector has also been
revised upwards for 2001 to 2003 as a result of higher investment in the
sector than assumed six months ago.
The interest burden in the enterprise sector (see Chart
6.13) has increased from a little less than 20% in 1997 to almost 35%
in 1999. According to our calculations, the interest rate burden will
be somewhat lower this year compared with 1999, followed by an increase
to 36% in 2003. The interest burden has been revised upwards substantially
compared with our last report, in which the interest burden was projected
at a little less than 30% for the whole projection period.
Chart 6.11
Table 6.3
Chart 6.12
Chart 6.13
11Enterprises' interest burden, ie interest
expenses as a percentage of cash surplus, provides a measure of the share of
enterprises' current surplus that is used for debt interest payments. The
debt ratio for interprises, ie the ratio between interest-bearing debt and
cash surplus less interest expenses, provides us with a measure of debt in
relation to the share of the surplus used for principal payments after
interest expenses have been paid.
Effects of changes in interest rates
Charts 6.14 and 6.15 show the first-round effects on the interest and
debt burden of a 1 percentage point increase in lending rates and a 1
percentage point reduction in relation to the baseline scenario for the
projection period as a whole. We have not assessed the extent to which
enterprises will change their behavioural pattern in response to a change
in interest rates.A change in lending rates of 1 percentage point compared
with the baseline scenario will influence the interest and debt burden.
An increase in interest rates will increase enterprises' interest burden
by almost 5 percentage points compared with the baseline scenario, bringing
the interest burden to a good 40%. The debt burden will also increase,
albeit to a lesser extent than the interest burden. Higher interest expenses
will reduce the share of cash surplus remaining after interest payments,
and the debt burden will then be a little more than 650% in 2003.
Chart 6.14
Chart 6.15
Loan collateral in the enterprise sector
The main collateral underlying loans to enterprises is commercial property,
in addition to collateral in the form of inventory and other real capital.
Commercial property is a less homogenous stock than dwellings, and is
thus less liquid as collateral than housing. This makes it difficult
to assess the aggregate value of commercial property. Developments in
rent for office premises provide some indication, however.
In the past few years, rents for office premises in
the large cities have been fairly stable (see Chart 6.16). In Oslo, rental
prices for new buildings in the centre and prestigious buildings have
edged up, while other segments have remained stable. In the first quarter
of 2000, all segments in Oslo have shown a weak upward price trend, which
represents a break with the trend in recent years.
The industry has reported that new operators with a
higher willingness to pay and a short time horizon, particularly in the
IT industry, on the rental market in Oslo have given rise to a need for
greater contractual flexibility, which may prompt lessors to accept shorter
leases. This may lead to higher prices, particularly for choice office
premises, but also engender greater uncertainty as to future rental income.
A number of operators expect a moderate increase in
prices in the period ahead, because of limited market capa-city and virtually
no speculative building. On the other hand, prices in the less attractive
segments seem to be edging down. In the longer term, prices are expected
to fall as a result of an increase in the supply of commercial property
in the years ahead while demand remains stable.
Chart 6.16
Higher credit risk in some industries
Norwegian banks account for the bulk of property management enterprises'
long-term debt. The banks are therefore fairly vulnerable to developments
in this industry. Table 6.4 shows that this industry accounted for more
than 40% of banks' long-term debt to enterprises at the end of 1999. Enterprises
in the shipping/pipeline transport are primarily financed by foreign financial
institutions and through bond issues. Even though they account for a large
share of long-term debt in the enterprise sector, they only represent
a small share of loans outstanding in Norwegian banks.
According to our last report, credit risk in parts of
the enterprise sector probably increased in 1998 (see Economic Bulletin
1999/4).12 Developments were particularly negative in the property management
industry (ie sale and rent of property) and shipping/pipeline transport.
Table 6.5 shows that long-term debt in property management enterprises
with the highest risk exposure increased sharply in 1998 (the figures
for risk developments in 1999 are not yet available). Banks still increased
their loans to this industry in 1999. The same developments have been
observed in the shipping/pipeline transport and other industries (primarily
primary industries, construction, power and water).
Although we can assume that a large share of the new
loans to property management enterprises has been extended to enterprises
with low risk exposure, the banks' overall credit risk exposure to this
industry has probably increased. Loans to this industry already account
for a very large share of banks' total loans, and this is an industry
where changes tend to occur rapidly. If developments take a turn for the
worse in this industry, banks may be hard hit. Excessive exposure to this
industry may thus involve a substantial risk.
So far, the increase in credit risk has had little impact
on net non-performing loans (ie gross non-performing loans less specified
loss provisions) as shown in Chart 6.17. For commercial banks, net non-performing
loans increased from NOK 5.9 to 6.1bn in 1999, while for savings banks
net non-performing loans increased from NOK 3.3 to 3.6bn. For banks as
a whole non-performing loans increased from NOK 9.2 to 9.7bn.
Table 6.4
Table 6.5
Chart 6.17
12 See Financial Sector Outlook 1999/2.
Outlook for the enterprise sector
The high debt burden means that from a financial viewpoint, enterprises
are less robust in the event of severe disturbances, either in the form
of weaker demand or an increase in interest rates.This situation may mean
that an increasing number of enterprises will experience problems in servicing
their debt in the years ahead.
On the other hand, the outlook for the Norwegian economy
is more favourable than anticipated earlier. The forecasts for aggregate
demand and output this year and next are somewhat higher. Household consumption
has picked up and growth prospects for our trading partners have improved.
This has probably reduced the likelihood of serious financial problems
for the enterprise sector as a whole compared with the situation six months
ago.
On balance, credit risk in parts of the enterprise sector
is still deemed to be high in view of the marked increase in new debt
in recent years.
The pronounced increase in banks' long-term loans to
the enterprise sector in 1999 is not necessarily of concern in itself.
A potential source of concern is that the bulk of new loans have been
extended to industries that include a large number of high-risk enterprises
and to the industries where banks already have the highest exposure.
As mentioned, banks are highly sensitive to developments
in the property management industry. A decline in the level of activity
and a marked fall in prices could create problems for banks. However,
there are few signs that this scenario will occur in the property market
in coming years.
There may be several reasons why the increase in credit
risk in parts of the enterprise sector has not translated into a marked
increase in non-performing loans. First, it may take some time before
an increase in credit risk leads to loan defaults. Second, an enterprise
with earning problems may avoid default by having its owners inject new
equity capital. Third, it is usually easier to negotiate payment extensions
and refinancing during an economic upturn than during a downturn.
© 2000 Norges Bank
webmaster@norges-bank.no