Description
Supervisors often advise banks that they should fund more of their loan book with customer deposits in order to become more robust to liquidity squeezes and contribute to the stability of the banking system.
Nr. 28 | 2012
Financial Stability
The declining deposit to loan ratio -
What can the banks do?
Sigbjørn Atle Berg
Staff Memo
Staff Memos present reports and documentation written by staff members and affiliates of
Norges Bank, the central bank of Norway. Views and conclusions expressed in Staff
Memos should not be taken to represent the views of Norges Bank.
© 2012 Norges Bank
The text may be quoted or referred to, provided that due acknowledgement is given to
source.
Staff Memo inneholder utredninger og dokumentasjon skrevet av Norges Banks ansatte og
andre forfattere tilknyttet Norges Bank. Synspunkter og konklusjoner i arbeidene er ikke
nødvendigvis representative for Norges Banks.
© 2012 Norges Bank
Det kan siteres fra eller henvises til dette arbeid, gitt at forfatter og Norges Bank
oppgis som kilde.
ISSN 1504-2596 (online only)
ISBN 978-82-7553-69 (online only)
1
The Declining Deposit to Loan Ratio – What Can the Banks Do?
1
Sigbjørn Atle Berg
1. Introduction
Supervisors often advise banks that they should fund more of their loan book with customer
deposits in order to become more robust to liquidity squeezes and contribute to the stability of
the banking system. The background is that banks in many countries have become
increasingly dependent on market funding. This may partly reflect larger trading books on
their balance sheets, but it mainly reflects the fact that customer deposits have not kept up
with the growth of the customer loan book.
Less deposit funding and more market funding is widely seen as negative for financial
stability. Market funding requires that the bank continually rolls over bill and bond issues and
renews borrowings from other financial institutions, in general depending on both domestic
and foreign investors. These funding sources have proved to be less stable than customer
deposits, and reliance on market funding has thus made the banks’ liquidity positions more
vulnerable to external shocks. The importance of this vulnerability was evident during the
recent financial crisis from 2008, as discussed by for instance Beltratti and Stulz (2012).
There are also some empirical studies indicating that more reliance on market funding
corresponds to higher risk exposure on the asset side of the balance sheet, see e.g. Demirgüc-
Kunt and Huizinga (2010) and Norden and Weber (2010). But is perhaps not a causal
relationship; it may rather be interpreted as different forms of risk-taking by banks who wants
an overall high risk exposure.
In this paper we shall ask whether and how the banks in aggregate can in fact reduce their
dependence on customer deposits. We do that by considering the determinants of the deposit-
to-loan ratio for the entire banking sector. Which are the factors behind the decline in this
ratio? Can these background factors be influenced by bank behaviour? We shall be looking at
the determinants of bank deposits and bank loans held by the domestic non-financial sector
(exclusive of the central government). This sector consists of households, non-profit
organisations, non-financial companies and the local governments. These groups are the core
customers that the banking sector should serve in an economy.
There is a large literature on the determinants of lending volume. A typical relationship for
credit growth includes GDP growth, the loan rate, the disposable household income and house
prices; see Hammersland and Træe (2011) for a Norwegian version. There is on the other
hand very little research on how bank deposit volumes are determined. Cohen and Kaufman
(1965) found that differences in deposit volumes between US states were mainly determined
1
Helpful comments from Henrik Andersen, Arild Lund and Ingvild Svendsen are gratefully acknowledged.
2
by household disposable income. Edminister and Merriken (1989) looked at the interest rate
sensitivity of customer deposits in the years following the removal of interest rate regulations
(Regulation Q) in the US. They essentially found that macroeconomic variables were much
more important determinants than interest rate changes, with varying levels of interest rate
sensitivity in different market environments. We are not aware of any more recent research
contributions.
We shall consider the factors determining bank deposits and bank loans at Norwegian banks,
both in the short and long term. A simple but superficial explanation is that bank loans have
become more attractive both for borrowers and lenders, and that this has made it worthwhile
for banks to fund the lending increases by borrowing from outside sectors, and in particular
the foreign sector. That is to say that the decline in the deposit to loan ratio is fully explained
by the denominator, while the deposit volume in the numerator has followed some (close to)
independent path.
A more sophisticated rephrasing of this standard argument would be to assume that both loans
and deposits are determined by the non-financial sector’s optimisation of its balance sheet.
Bank deposits are a form of financial investments that customers in the non-financial sector
make out of their total gross financial assets. The size of these assets, as well as the return on
alternative investments, should thus be important for explaining the size of bank deposits
within a standard Markowitz portfolio construction framework. Gross financial assets can be
accumulated either through savings or through borrowing, and can take many different forms.
Bank deposits represent only one kind of financial asset, and the non-financial sector is likely
to invest in alternative financial assets when these become more attractive.
But there may be more to the story than this. A financial flows model makes it clear that there
is a very basic relationship between bank loans and bank deposits. Whenever a loan is
provided, the money will be moved to a deposit account. When the money is used as payment
for buying goods or services, the payment will generally be made by transferring money from
one deposit account to another. If deposits are withdrawn to make payments in cash, the
receiver will very likely deposit it with a bank. In short, bank loans tend to create bank
deposits of the same size.
Loans from non-bank financial institutions are also likely to create new bank deposits. There
are on the other hand leakages to the chain of deposits when payments are made to an entity
outside the domestic non-financial sector. This could be payments to the central government,
to the financial sector or to the foreign sector. Notice that alternative financial investments in
the portfolio model above will in many cases involve a payment to the financial sector and
thus represent a leakage in the loan-deposit chain. As noted by Disyatat (2011), “for the
system as a whole, a substantial change in the aggregate amount of deposits suggests an
overall shift in the structure of the money market”, meaning that bank deposits are replaced
by other financial assets.
We thus propose two lines of reasoning for explaining the decline in the deposit-to-loan ratio.
One line holds that loans have become more attractive and/or that bank deposits have become
less attractive to the agents in the non-financial sector. The other line claims that structural
3
changes in the financial markets have redirected some financial flows affecting deposits
and/or loans. A combination of these two lines of argument is perhaps the most plausible
model, but below we shall at first explore them separately.
This paper will be limited to exploring the Norwegian experience by trying to identify the
factors behind the decline of the deposit-to-loan ratio at Norwegian banks. But declines in the
deposit-to-loan ratio have also taken place in many other countries, and the empirical analysis
below may thus be of some interest for their experiences as well.
In section 2 we shall have a brief look at the aggregate balance sheet of the Norwegian
banking sector, and in particular the developments in the deposit-to-loan ratio. Both deposit
and loan volumes depend mainly on decisions made by bank customers, and in section 3 we
explore the portfolio model by looking at how attractive bank loans and bank deposits have
been relative to alternatives. In section 4 we explore the financial flows model by looking into
the detailed financial accounts of the non-financial sector. Section 5 concludes.
2. The Deposit-to-Loan Ratio for Norwegian banks
Our empirical analysis will be using data for the Norwegian banking industry from 1975 to
2011. In the Norwegian context a meaningful time series of bank loans will have to include
loans from mortgage companies. We include both companies specialising on corporate loans
and companies specialising on housing and property loans. The mortgage companies have
always been important to the corporate sector, but they have also emerged as important
lenders to the household sector since covered bonds were introduced in Norway in 2007. The
mortgage companies are in general owned by banks and are integral parts of bank groups.
Banks have been transferring an increasingly large portion of their mortgage loans to their
subsidiary mortgage companies in order to use these loans as a basis for issuing covered
bonds. Covered bond issuance now funds the major part of mortgage loans, loans that
previously had to be funded by bank deposits or banks’ non-secured borrowing.
Chart 1 depicts the ratio between bank deposits and banking sector loans (including the loans
from the mortgage companies) held by the non-financial sector (the green line). The ratio
declined from nearly 100 per cent to around 60 per cent in the run-up to the Norwegian
banking crisis of 1990-92 when loan volumes increased rapidly. After the crisis period there
was an upswing to nearly 80 per cent coverage, until a new decline in the ratio started in 1995
and brought it down to about 50 per cent today. The only periods with an increasing ratio
have been 1976-80 and 1989-95.
In the analysis below we shall look separately at the households sector (including non-profit
organisations) and the non-financial corporate sector (including local governments), who
together constitute the domestic non-financial sector. Chart 1 includes separate deposit-to-
loan ratios for these two subsectors. There is a declining trend in the ratio is for the
households sector, while there are swings, but no clear downward trend, for the non-financial
companies. The ratio has for most of the period been lower for the corporate sector, but that
4
difference has lately disappeared and been reversed. The chart seems to indicate that the long
term downward trend has mainly to do with household behaviour.
Chart 1: Bank deposits of the domestic non-financial sector as a ratio of its loans from banks
and mortgage companies. Source: Financial Accounts, Statistics Norway.
Deposits from and loans to the non-financial sector are important components of Norwegian
banking sector’s (banks and mortgage companies) balance sheets. Simplified balance sheets at
end 1975 and end 2011 are depicted in Chart 2. The left hand panel shows banks’ assets.
Loans to the domestic non-financial sector constitute about 60 per cent of banking sector total
assets both in 1975 and 2011. These numbers clearly illustrate that loans to the domestic non-
financial sector remains the core activity of the banking sector. The remaining 40 per cent of
the total assets consist in 2011 of claims on other financial institutions (20 per cent), tradable
financial instruments (10 per cent) and other assets (10 per cent).
Chart 2: Aggregate assets (left) and liabilities (right) of all banks by end 1975 and all banks
and mortgage companies by end 2011. Source: Banking Statistics, Statistics Norway.
5
The right hand panel in Chart 2 shows the liabilities. In 2011 deposits from the domestic non-
financial sector makes for 31 per cent of total liabilities in banks and mortgage companies,
down from 70 per cent in 1975. A little more than half of those deposits are guaranteed by the
deposit insurance scheme. Around 20 per cent are deposits from other financial institutions
and the central government, a little more in 2011 than in 1975. Equity capital represents 5-6
per cent. The remaining 40 per cent are in 2011 bills and bonds and other forms of borrowing
from the markets or from the central bank. This is in contrast to 1975 when government
regulation of credit volumes was extensive, and market funding was hardly used. The shift
that has taken place on the liability side of the balance sheet is a quite dramatic decline in the
importance of customer deposits.
3. Optimising the balance sheet of the non-financial sector
The bank deposit volume held by the domestic non-financial sector is determined by decisions
made by the agents in that sector. The volume of bank loan taken by the non-financial sector
depends heavily on decisions made by the same set of agents, even if banks’ credit evaluation
standards will also be important. Our first line of reasoning for explaining the deposit-to-loan
ratio holds that the deposit-to-loan ratio is determined by optimising agents in the non-
financial sector. We thus need to look more closely at the balance sheet of that sector.
Chart 3: The share of non-financial sector deposits (left hand panel) and bank loans (right
hand panel) held by households and non-profit organisations. Source: Financial Sector
Accounts, Statistics Norway.
We shall look separately at the household sector (including non-profit organisations) and the
non-financial corporate sector (including local governments). Their balance sheets can be
extracted from the quarterly Financial Accounts, published by Statistics Norway. We start by
looking at the relative importance of the two sectors. Chart 3 shows the households’ share of
the total bank deposits and the total bank loans held by the non-financial sector. Households’
share of deposits has been declining since the late 1980’s and is currently below 60 per cent,
whereas the households’ share of bank loans has been around 60 per cent for the entire period,
50 %
55 %
60 %
65 %
70 %
75 %
80 %
1
9
7
5
Q
1
1
9
7
8
Q
1
1
9
8
1
Q
1
1
9
8
4
Q
1
1
9
8
7
Q
1
1
9
9
0
Q
1
1
9
9
3
Q
1
1
9
9
6
Q
1
1
9
9
9
Q
1
2
0
0
2
Q
1
2
0
0
5
Q
1
2
0
0
8
Q
1
2
0
1
1
Q
1
40 %
50 %
60 %
70 %
80 %
1
9
7
5
Q
1
1
9
7
8
Q
1
1
9
8
1
Q
1
1
9
8
4
Q
1
1
9
8
7
Q
1
1
9
9
0
Q
1
1
9
9
3
Q
1
1
9
9
6
Q
1
1
9
9
9
Q
1
2
0
0
2
Q
1
2
0
0
5
Q
1
2
0
0
8
Q
1
2
0
1
1
Q
1
6
with a slight upward trend. This chart provides a further indication that the declining deposit-
to-loan ratio has a lot to do with household behaviour.
Chart 4: The portion of bank deposits held in transaction accounts. Source: Financial Sector
Accounts, Statistics Norway.
The composition of bank deposits by type of account is available from 1995 only. Chart 4
shows that households have had only 30-35 per cent of their deposits in transaction accounts,
whereas the corresponding ratio for non-financial companies has remained above 60 per cent
and is currently nearly 70 per cent. This may indicate that bank deposits are mainly
investment instruments for households, whereas non-financial companies have deposits
mostly for transaction purposes.
Chart 5: The gross financial assets of households and non-profit organisations by instrument.
Source: Financial Accounts, Statistics Norway.
0 %
10 %
20 %
30 %
40 %
50 %
60 %
70 %
80 %
90 %
100 %
Transaction accounts Other deposit accounts
0 %
10 %
20 %
30 %
40 %
50 %
60 %
70 %
80 %
90 %
100 %
Transcation accounts Other deposit accounts
Households sector Corporate sector
7
The composition of households’ gross financial assets is presented in Chart 5. The share of
bank deposits has declined from 44 per cent in 1975 to 32 per cent at end 2011. The decline
started after bank loan markets had been deregulated in the mid 1980s. Prior to that loans
were effectively rationed, and households would have to save in bank deposits in order to
qualify for a loan from that bank. The alternative investments whose share of total assets has
been increasing are mainly insurance reserves and equity investments. Equity holdings have
increased from 5 per cent of households’ financial assets in 1975 to 13 per cent in 2011, and
the share of insurance reserves from 23 to 38 per cent.
Insurance reserves are essentially the customer portfolios in pension funds and life insurance
companies. These portfolios are predominantly invested in fixed income (about 70 per cent in
2011) and equity instruments (about 20 per cent in 2011). Households thus hold about 50 per
cent of their financial assets as capital market investments, as compared to 30 per cent in
1975. This shift out of bank deposits and into market instruments may partly be because
higher income levels have stimulated long term saving for retirement or because returns have
become more attractive. But to a very large extent it reflects structural changes whereby less
of the long term saving desired by households has been provided through government non-
funded pension schemes and more from privately funded schemes. Finally, new financial
instruments introduced and sold by the banking sector may have tempted customers out of
bank deposits.
Chart 6: The gross financial assets of non-financial companies and local government by
instrument. Source: Financial Accounts, Statistics Norway.
Chart 6 similarly depicts the asset composition of non-financial companies. These time series
have a break in 1995Q4 when new definitions were introduced.
2
But we can still see a clear
pattern: The share of bank deposits declined by 8 percentage points from 1975 to 1995, and
by another 2 percentage points after 1995. The bank deposit share is currently 14 % of total
2
The main change was that loans from one non-financial company to another were included on both sides of the
balance sheet, thus blowing up both total assets and total liabilities.
8
assets. Equity holdings did on the other hand increase by 17 percentage points between 1975
and 1995, and by a further 15 percentage points after 1995. At end 2011 the equity share is 33
per cent of total assets.
Comparing charts 5 and 6 above, we notice that bank deposits have consistently been a much
larger share of total financial assets in the households sector than in the non-financial
companies. This would indicate that the sector allocation of gross financial assets in the non-
financial sector may have had an effect on the aggregate share of financial assets held as bank
deposits.
On the liability side of the balance sheet, bank loans are very important for the households
sector, with nearly 80 per cent of their total liabilities. Chart 7 shows that the increased
reliance on loans from private banks has a counterpart in less reliance on loans from
government sector lenders. The sum of loans from these two types of lenders has increased
slightly, from 84 to 92 per cent of households’ total liabilities.
Chart 7: The liabilities of households and non-profit organisations by creditor. Source:
Financial Accounts, Statistics Norway.
Chart 8: The liabilities of non-financial companies and local government by creditor. Source:
Financial Accounts, Statistics Norway.
9
Chart 8 shows the liabilities of the corporate sector. The break in the time series from 1995
reflects changes in definitions, the main implication being that more loans between companies
are included in the numbers (footnote 1). We notice that there have been only small changes
in the importance of loans from private sector banks, but a clear decline in the importance of
government lending.
The increased importance of bank loans to the non-financial sector thus partly reflects a
reduction in lending from the central government sector. Government lenders have reduced
their loans to both households and companies, and the private banks have been providing
most of these loans instead. The public sector share of lending to households and companies
have been steadily decreasing since 1981-82. This structural change in the Norwegian credit
market was a consequence of government policies to reduce the importance of lending
through government agencies. Notice that government lenders mostly did not take deposits.
This structural change in the loan market may thus be an important factor behind the decline
in the deposit-to-loan ratio at private banks.
We go on to consider how changes in the return of different investments may have changed
the incentives to make deposits or take loans. Government bond returns are available from
1990. Chart 9 shows that they have followed bank deposit rates closely since then, but with
the return differential tending towards zero. The returns on equity investments have been very
volatile, but for most of the period since 1984 well above the returns on fixed income
investments. There have been incentives to move funds into the stock market. Notice also that
the period 1989-95, when the deposit-to-loan ratio were increasing, had high real returns on
bank deposits and weak returns on the Oslo Stock Exchange.
Chart 9: Real post-tax returns (deflated by the CPI) on bank deposits, government bonds and
stocks listed on the Oslo Stock Exchange. Three years moving averages. Source: Statistics
Norway, Norges Bank and Oslo Stock Exchange.
On the liability side we first notice that the real borrowing costs of the non-financial sector
were high when the deposit-to-loan ratio was declining in 1989-95. The value of interest
10
payment tax deductibility was reduced from up to 50% down to a flat 28 per cent by the 1992
tax reform, but pre-tax interest rates have also been reduced. Chart 10 shows that the tax and
inflation adjusted cost of taking out a bank loan is currently less than 2 per cent. This is very
low by most standards. Notice, however, that the real cost of borrowing was negative in the
run-up to the 1988-92 banking crisis.
Households mostly borrow for buying homes. The profitability of borrowing to invest
depends on the house price inflation relative to the borrowing costs. Chart 10 does not take
maintenance costs into account, but the chart still strongly indicates that housing investments
have been profitable for most of the period after 1993. This followed a period during the
banking crisis of 1988-92 when these investments had been highly unprofitable.
Chart 10: Real post-tax costs of bank loans and real house price inflation (deflated by the
CPI). Three years moving averages. Source: Statistics Norway
The economic incentives to increased borrowing have thus been evident for the households’
sector during the past 20 years. Above we saw that incentives to hold bank deposits had been
weakened. In the period 1989-95 when the deposit-to-loan ratio increased, the economic
incentives to make deposits were temporarily stronger and the incentives to borrow weaker.
Eyeball econometrics seem to indicate that the portfolio model has some explanatory power.
4. Financial flows analysis
The gross financial assets of the non-financial sector can essentially be accumulated in three
different ways; by financial saving, by borrowing and by revaluation of existing assets. Chart
11 provides convincing evidence that asset accumulation in the non-financial sector mainly
comes from borrowing. The levels of total financial assets and total debt of the non-financial
sector move in parallel in the upper panel. In the lower panel we can observe the very tight
co-variation of changes in these two variables.
11
However, if we look at the households and non-financial company sectors separately, the co-
variation is less close, see Chart 12. This is in particular true for the households sector. Their
financial assets increased much faster than their debts during the 1990s, probably because
savings became important for consolidating their balance sheets after the banking crisis. The
opposite happened during most of the 2000s, when debt grew faster than assets.
Chart 11: Gross financial assets and total debt of the non-financial sector in per cent of GDP.
Levels in the upper panel and 4 quarter changes in the lower panel. Source: Financial
accounts, Statistics Norway.
Chart 12: Gross financial assets and total debt of the households sector (left hand panel) and
the non-financial corporate sector (right hand panel) in per cent of GDP. Source: Financial
accounts and national accounts, Statistics Norway.
0 %
50 %
100 %
150 %
200 %
250 %
300 %
350 %
Financial assts Total debt
40 %
50 %
60 %
70 %
80 %
90 %
100 %
110 %
1
9
7
5
Q
1
1
9
7
8
Q
1
1
9
8
1
Q
1
1
9
8
4
Q
1
1
9
8
7
Q
1
1
9
9
0
Q
1
1
9
9
3
Q
1
1
9
9
6
Q
1
1
9
9
9
Q
1
2
0
0
2
Q
1
2
0
0
5
Q
1
2
0
0
8
Q
1
2
0
1
1
Q
1
Financial assets Total debt
12
Chart 12 shows the debt and assets data in per cent of GDP. We can see that the non-financial
corporate sector has much more debt and also more financial assets than the households. The
households have now about 100 per cent of GDP on both sides of their balance sheet, with net
assets of around 10 per cent of GDP. The corporate sector has gross debts above 260 per cent
and assets above 170 per cent of GDP, with net debts at 80-90 per cent of GDP. Furthermore,
the households’ share of total assets in the non-financial sector has been declining from
around 55 per cent in 1975 to less than 40 per cent in 2011. That shift alone contributes to a
smaller fraction of assets in the non-financial sector being invested as bank deposits, confer
Chart 3.
Chart 13: The levels (left hand panel) and four quarter changes (right hand panel) of bank
deposits and bank loans to the non–financial sector. Millions NOK. Source: Statistics
Norway.
We have seen that debt creates financial assets. Similarly, bank loans create bank deposits. A
bank loan is given by making a deposit in the name of the borrower. The borrower normally
spends or invests the proceeds by transferring it to another deposit account and so on. But as
demonstrated in Chart 13 there is correlation, but not close co-variation of the two data series,
neither in levels nor in differences.
There are two reasons for these deviations, as illustrated in the financial flows diagram (Chart
14). First, the bank deposit chain has leakages whenever payments are made from the non-
financial sector to any outside sector. These outside sectors are the financial sector, the central
government and the foreign sector. Any investments of the non-financial sector in financial
instruments other than bank deposits will involve payments to the financial sector. The
increased equity investment and insurance reserves shown in charts 5 and 6 represent such
leakages. These increased leakages have contributed to the decline of the deposit-to-loan
ratio.
Second, transfers to the non-financial sector from the outside sectors has helped increase
bank deposits, but these transfers are not correlated with bank loan volumes. Looking more
closely on these transfers we notice that borrowing from the central government has been
declining, see charts 5 and 6, thus gradually contributing less to deposit growth. Borrowing
from abroad to the non-financial sector has increased, however, which is a net contribution to
?100 000
?50 000
0
50 000
100 000
150 000
200 000
250 000
300 000
350 000
400 000
Deposit growth Loan growth
0
500 000
1 000 000
1 500 000
2 000 000
2 500 000
3 000 000
3 500 000
Deposits Loans
13
domestic bank deposits. But borrowing from abroad is nearly exclusively to the corporate
sector, whose bank deposits are only a small portion of total financial assets.
Chart 14: Deposits in the financial flows diagram
Charts 7 and 8 above illustrated the importance of bank loans as a share of total debt of the
non-financial sector. Bank loans have become very important for the households sector where
they constitute nearly 80 per cent of total debt in 2011. But they are much less important for
the corporate sector, where the share is only 18 per cent.
Chart 15: The share of mortgage loans to total private bank loans to the households sector
(left hand panel) and changes in the deposit to loan ratio for households plotted against
changes in share of loans from public sector banks in households’ total liabilities (right hand
panel). Source: Statistics Norway.
30 %
40 %
50 %
60 %
70 %
80 %
90 %
100 %
Share of mortgage loans
?9 %
?6 %
?3 %
0 %
3 %
6 %
9 %
12 %
Change D/L ratio
Change public sector banks
14
In parallel with this transfer of credit extension from public sector to private sector banks, the
composition of the household loan book with private banks have changed. Chart 15 shows
that the share of mortgage loans has increased from 40 per cent of total private bank loans to
households in 1987 to 90 per cent in 2011. This naturally reflects the gradual transfer of
mortgage loans from government lenders to private banks. In the last few years it also reflects
the introduction of new loan products where households can use real estate as collateral for
loans intended for other purposes than real estate purchases. There has also been an increasing
share of mortgage loans with no repayments for extended periods.
The right hand panel of Chart 15 shows no close co-variation between the share of public
sector loans and the deposit-to-loan ratio for households. This may indicate that the shift of
mortgage loans from public sector lenders to private banks is not a good explanation for the
year-to-year changes in the deposit-to-loan ratio. It may still be important for explaining the
long term trend. Notice in particular the temporary upward shift both in the importance of
government lending and the deposit-to-loan ratio at private banks in the early 1990s.
5. Conclusions
We have been exploring possible explanations for the strong decline in the ratio between bank
deposits held by the non-financial sector in Norway, and the bank loans taken by the same
sector. We have followed two different but related lines of reasoning. First, we have been
looking at the relative return on bank deposits and on the relative costs of bank loans in a
portfolio model approach. Second, we have tried to follow financial flows to identify
structural changes in the Norwegian markets that could have an effect on the deposit-to-loan
ratio.
We did find that real deposit rates have been low during periods with strong decline in the
deposit-to-loan ratio, pointing to the attraction of alternative investment opportunities in
capital market instruments. This has either been done directly or through the participation in
funded pension schemes. During the periods of strong declines in the ratio we also found that
the real loan rate has been low. In combination with a significant house price inflation this
have made borrowing quite attractive to households.
The return incentives are particularly clear for the households sector. That is also the sector
where the deposit-to-loan ratio has been declining since 1995. For the corporate sector the
return incentives have been less clear-cut, and the decline in the deposit-to-loan ratio has been
modest.
On the structural side, we first find that funded pension schemes have become more important
during the past 30 years. This has probably to do not only with return incentives, but also with
the growing importance of privately funded pension schemes as a supplement to the public
non-funded pension schemes. Public pension schemes have become less satisfactory both
because the upper limit on benefits have become binding for more people, and recently also
because the level of benefits have been curtailed.
15
A second structural factor is that a large part of the lending to households has been transferred
from public sector agencies to private sector banks during the past 30 years. A similar
development has taken place for the corporate sector, but at a much smaller scale. Since
public sector lenders mostly did not take deposits, this represents an obvious factor behind the
long term trend. In parallel with this development, bank loan markets were deregulated,
making deposits less important as a qualifier for bank loans. But these processes have been
slow and may not contribute much to explaining the year-to-year changes in the deposit-to-
loan ratio.
We also find a counteracting factor in the increased reliance of the corporate sector on foreign
borrowing. Such financial inflows should lead to more deposits with Norwegian banks, at
least when the borrowers operate on the domestic market. But there are two reasons why this
effect may be small: First, much of the foreign borrowing has probably been to the petroleum
sector that to a large extent operates outside the Norwegian economy. Second, the corporate
sector holds only a small share of its financial assets as bank deposits. They do that mainly for
transaction purposes, and their transaction needs will not increase as a consequence of shifting
from domestic to foreign borrowing.
We are often told that banks should increase their deposit funding and reduce their reliance on
market funding, meaning that they should engineer a rise in the deposit-to-loan ratio. While
this is clearly desirable from a financial stability point of view, it is less clear how much the
banks can actually do. There is very little they can do with some of the structural changes that
have taken place. The strongest return incentive for household behaviour probably stems from
house price inflation, that individual banks are not able to control.
The banks can certainly bid higher deposit rates to attract deposits, but the aggregate effect on
total deposits in the system may be small. Banks can restrain their sales of new financial
products of the kind that have shifted investments out of deposits and into alternative financial
instruments, but that must a concerted effort to make much difference at the macro level.
What individual banks undoubtedly can do is to restrict lending. That will certainly also
reduce deposits, but the link between loans and deposits is likely to be much weaker at the
individual bank level than it is in aggregate. And if lending growth is not restricted, one
contribution individual banks could make to financial stability would be to make more of their
market funding long term.
References:
Beltratti, Andrea and Stulz, René M. (2012), The credit crisis around the globe: Why did
some banks perform better? Journal of Financial Economics 105, 1-17.
Cohen, Bruce C. and George G. Kaufman (1965), Factors Determining Bank Deposit Growth
by State: An Empirical Analysis. Journal of Finance, 20 (1), 59-70.
16
Demirgüc-Kunt, Asli and Harry Huizinga (2010), Bank activity and funding strategies: The
impact on risk and returns. Journal of Financial Economics 98, 626-650.
Disyatat, Piti (2011), The bank lending channel revisited. Journal of money, Credit and
Banking 43 (4), 711-734.
Edminister, Robert O. and Henry E. Merriken (1989), Measuring interest rate sensitivity of
consumer depositors, Journal of Financial Services Research 2, 133-145.
Hammersland, Roger and Cathrine Bolstad Træe (2011), The financial accelerator and the
real economy. Discussion Paper 668, Statistics Norway.
Norden. Lars and Martin Weber (2010), Funding modes of German banks: structural changes
and their implications. Journal of Financial Services Research 38, 69-93.
doc_915949780.pdf
Supervisors often advise banks that they should fund more of their loan book with customer deposits in order to become more robust to liquidity squeezes and contribute to the stability of the banking system.
Nr. 28 | 2012
Financial Stability
The declining deposit to loan ratio -
What can the banks do?
Sigbjørn Atle Berg
Staff Memo
Staff Memos present reports and documentation written by staff members and affiliates of
Norges Bank, the central bank of Norway. Views and conclusions expressed in Staff
Memos should not be taken to represent the views of Norges Bank.
© 2012 Norges Bank
The text may be quoted or referred to, provided that due acknowledgement is given to
source.
Staff Memo inneholder utredninger og dokumentasjon skrevet av Norges Banks ansatte og
andre forfattere tilknyttet Norges Bank. Synspunkter og konklusjoner i arbeidene er ikke
nødvendigvis representative for Norges Banks.
© 2012 Norges Bank
Det kan siteres fra eller henvises til dette arbeid, gitt at forfatter og Norges Bank
oppgis som kilde.
ISSN 1504-2596 (online only)
ISBN 978-82-7553-69 (online only)
1
The Declining Deposit to Loan Ratio – What Can the Banks Do?
1
Sigbjørn Atle Berg
1. Introduction
Supervisors often advise banks that they should fund more of their loan book with customer
deposits in order to become more robust to liquidity squeezes and contribute to the stability of
the banking system. The background is that banks in many countries have become
increasingly dependent on market funding. This may partly reflect larger trading books on
their balance sheets, but it mainly reflects the fact that customer deposits have not kept up
with the growth of the customer loan book.
Less deposit funding and more market funding is widely seen as negative for financial
stability. Market funding requires that the bank continually rolls over bill and bond issues and
renews borrowings from other financial institutions, in general depending on both domestic
and foreign investors. These funding sources have proved to be less stable than customer
deposits, and reliance on market funding has thus made the banks’ liquidity positions more
vulnerable to external shocks. The importance of this vulnerability was evident during the
recent financial crisis from 2008, as discussed by for instance Beltratti and Stulz (2012).
There are also some empirical studies indicating that more reliance on market funding
corresponds to higher risk exposure on the asset side of the balance sheet, see e.g. Demirgüc-
Kunt and Huizinga (2010) and Norden and Weber (2010). But is perhaps not a causal
relationship; it may rather be interpreted as different forms of risk-taking by banks who wants
an overall high risk exposure.
In this paper we shall ask whether and how the banks in aggregate can in fact reduce their
dependence on customer deposits. We do that by considering the determinants of the deposit-
to-loan ratio for the entire banking sector. Which are the factors behind the decline in this
ratio? Can these background factors be influenced by bank behaviour? We shall be looking at
the determinants of bank deposits and bank loans held by the domestic non-financial sector
(exclusive of the central government). This sector consists of households, non-profit
organisations, non-financial companies and the local governments. These groups are the core
customers that the banking sector should serve in an economy.
There is a large literature on the determinants of lending volume. A typical relationship for
credit growth includes GDP growth, the loan rate, the disposable household income and house
prices; see Hammersland and Træe (2011) for a Norwegian version. There is on the other
hand very little research on how bank deposit volumes are determined. Cohen and Kaufman
(1965) found that differences in deposit volumes between US states were mainly determined
1
Helpful comments from Henrik Andersen, Arild Lund and Ingvild Svendsen are gratefully acknowledged.
2
by household disposable income. Edminister and Merriken (1989) looked at the interest rate
sensitivity of customer deposits in the years following the removal of interest rate regulations
(Regulation Q) in the US. They essentially found that macroeconomic variables were much
more important determinants than interest rate changes, with varying levels of interest rate
sensitivity in different market environments. We are not aware of any more recent research
contributions.
We shall consider the factors determining bank deposits and bank loans at Norwegian banks,
both in the short and long term. A simple but superficial explanation is that bank loans have
become more attractive both for borrowers and lenders, and that this has made it worthwhile
for banks to fund the lending increases by borrowing from outside sectors, and in particular
the foreign sector. That is to say that the decline in the deposit to loan ratio is fully explained
by the denominator, while the deposit volume in the numerator has followed some (close to)
independent path.
A more sophisticated rephrasing of this standard argument would be to assume that both loans
and deposits are determined by the non-financial sector’s optimisation of its balance sheet.
Bank deposits are a form of financial investments that customers in the non-financial sector
make out of their total gross financial assets. The size of these assets, as well as the return on
alternative investments, should thus be important for explaining the size of bank deposits
within a standard Markowitz portfolio construction framework. Gross financial assets can be
accumulated either through savings or through borrowing, and can take many different forms.
Bank deposits represent only one kind of financial asset, and the non-financial sector is likely
to invest in alternative financial assets when these become more attractive.
But there may be more to the story than this. A financial flows model makes it clear that there
is a very basic relationship between bank loans and bank deposits. Whenever a loan is
provided, the money will be moved to a deposit account. When the money is used as payment
for buying goods or services, the payment will generally be made by transferring money from
one deposit account to another. If deposits are withdrawn to make payments in cash, the
receiver will very likely deposit it with a bank. In short, bank loans tend to create bank
deposits of the same size.
Loans from non-bank financial institutions are also likely to create new bank deposits. There
are on the other hand leakages to the chain of deposits when payments are made to an entity
outside the domestic non-financial sector. This could be payments to the central government,
to the financial sector or to the foreign sector. Notice that alternative financial investments in
the portfolio model above will in many cases involve a payment to the financial sector and
thus represent a leakage in the loan-deposit chain. As noted by Disyatat (2011), “for the
system as a whole, a substantial change in the aggregate amount of deposits suggests an
overall shift in the structure of the money market”, meaning that bank deposits are replaced
by other financial assets.
We thus propose two lines of reasoning for explaining the decline in the deposit-to-loan ratio.
One line holds that loans have become more attractive and/or that bank deposits have become
less attractive to the agents in the non-financial sector. The other line claims that structural
3
changes in the financial markets have redirected some financial flows affecting deposits
and/or loans. A combination of these two lines of argument is perhaps the most plausible
model, but below we shall at first explore them separately.
This paper will be limited to exploring the Norwegian experience by trying to identify the
factors behind the decline of the deposit-to-loan ratio at Norwegian banks. But declines in the
deposit-to-loan ratio have also taken place in many other countries, and the empirical analysis
below may thus be of some interest for their experiences as well.
In section 2 we shall have a brief look at the aggregate balance sheet of the Norwegian
banking sector, and in particular the developments in the deposit-to-loan ratio. Both deposit
and loan volumes depend mainly on decisions made by bank customers, and in section 3 we
explore the portfolio model by looking at how attractive bank loans and bank deposits have
been relative to alternatives. In section 4 we explore the financial flows model by looking into
the detailed financial accounts of the non-financial sector. Section 5 concludes.
2. The Deposit-to-Loan Ratio for Norwegian banks
Our empirical analysis will be using data for the Norwegian banking industry from 1975 to
2011. In the Norwegian context a meaningful time series of bank loans will have to include
loans from mortgage companies. We include both companies specialising on corporate loans
and companies specialising on housing and property loans. The mortgage companies have
always been important to the corporate sector, but they have also emerged as important
lenders to the household sector since covered bonds were introduced in Norway in 2007. The
mortgage companies are in general owned by banks and are integral parts of bank groups.
Banks have been transferring an increasingly large portion of their mortgage loans to their
subsidiary mortgage companies in order to use these loans as a basis for issuing covered
bonds. Covered bond issuance now funds the major part of mortgage loans, loans that
previously had to be funded by bank deposits or banks’ non-secured borrowing.
Chart 1 depicts the ratio between bank deposits and banking sector loans (including the loans
from the mortgage companies) held by the non-financial sector (the green line). The ratio
declined from nearly 100 per cent to around 60 per cent in the run-up to the Norwegian
banking crisis of 1990-92 when loan volumes increased rapidly. After the crisis period there
was an upswing to nearly 80 per cent coverage, until a new decline in the ratio started in 1995
and brought it down to about 50 per cent today. The only periods with an increasing ratio
have been 1976-80 and 1989-95.
In the analysis below we shall look separately at the households sector (including non-profit
organisations) and the non-financial corporate sector (including local governments), who
together constitute the domestic non-financial sector. Chart 1 includes separate deposit-to-
loan ratios for these two subsectors. There is a declining trend in the ratio is for the
households sector, while there are swings, but no clear downward trend, for the non-financial
companies. The ratio has for most of the period been lower for the corporate sector, but that
4
difference has lately disappeared and been reversed. The chart seems to indicate that the long
term downward trend has mainly to do with household behaviour.
Chart 1: Bank deposits of the domestic non-financial sector as a ratio of its loans from banks
and mortgage companies. Source: Financial Accounts, Statistics Norway.
Deposits from and loans to the non-financial sector are important components of Norwegian
banking sector’s (banks and mortgage companies) balance sheets. Simplified balance sheets at
end 1975 and end 2011 are depicted in Chart 2. The left hand panel shows banks’ assets.
Loans to the domestic non-financial sector constitute about 60 per cent of banking sector total
assets both in 1975 and 2011. These numbers clearly illustrate that loans to the domestic non-
financial sector remains the core activity of the banking sector. The remaining 40 per cent of
the total assets consist in 2011 of claims on other financial institutions (20 per cent), tradable
financial instruments (10 per cent) and other assets (10 per cent).
Chart 2: Aggregate assets (left) and liabilities (right) of all banks by end 1975 and all banks
and mortgage companies by end 2011. Source: Banking Statistics, Statistics Norway.
5
The right hand panel in Chart 2 shows the liabilities. In 2011 deposits from the domestic non-
financial sector makes for 31 per cent of total liabilities in banks and mortgage companies,
down from 70 per cent in 1975. A little more than half of those deposits are guaranteed by the
deposit insurance scheme. Around 20 per cent are deposits from other financial institutions
and the central government, a little more in 2011 than in 1975. Equity capital represents 5-6
per cent. The remaining 40 per cent are in 2011 bills and bonds and other forms of borrowing
from the markets or from the central bank. This is in contrast to 1975 when government
regulation of credit volumes was extensive, and market funding was hardly used. The shift
that has taken place on the liability side of the balance sheet is a quite dramatic decline in the
importance of customer deposits.
3. Optimising the balance sheet of the non-financial sector
The bank deposit volume held by the domestic non-financial sector is determined by decisions
made by the agents in that sector. The volume of bank loan taken by the non-financial sector
depends heavily on decisions made by the same set of agents, even if banks’ credit evaluation
standards will also be important. Our first line of reasoning for explaining the deposit-to-loan
ratio holds that the deposit-to-loan ratio is determined by optimising agents in the non-
financial sector. We thus need to look more closely at the balance sheet of that sector.
Chart 3: The share of non-financial sector deposits (left hand panel) and bank loans (right
hand panel) held by households and non-profit organisations. Source: Financial Sector
Accounts, Statistics Norway.
We shall look separately at the household sector (including non-profit organisations) and the
non-financial corporate sector (including local governments). Their balance sheets can be
extracted from the quarterly Financial Accounts, published by Statistics Norway. We start by
looking at the relative importance of the two sectors. Chart 3 shows the households’ share of
the total bank deposits and the total bank loans held by the non-financial sector. Households’
share of deposits has been declining since the late 1980’s and is currently below 60 per cent,
whereas the households’ share of bank loans has been around 60 per cent for the entire period,
50 %
55 %
60 %
65 %
70 %
75 %
80 %
1
9
7
5
Q
1
1
9
7
8
Q
1
1
9
8
1
Q
1
1
9
8
4
Q
1
1
9
8
7
Q
1
1
9
9
0
Q
1
1
9
9
3
Q
1
1
9
9
6
Q
1
1
9
9
9
Q
1
2
0
0
2
Q
1
2
0
0
5
Q
1
2
0
0
8
Q
1
2
0
1
1
Q
1
40 %
50 %
60 %
70 %
80 %
1
9
7
5
Q
1
1
9
7
8
Q
1
1
9
8
1
Q
1
1
9
8
4
Q
1
1
9
8
7
Q
1
1
9
9
0
Q
1
1
9
9
3
Q
1
1
9
9
6
Q
1
1
9
9
9
Q
1
2
0
0
2
Q
1
2
0
0
5
Q
1
2
0
0
8
Q
1
2
0
1
1
Q
1
6
with a slight upward trend. This chart provides a further indication that the declining deposit-
to-loan ratio has a lot to do with household behaviour.
Chart 4: The portion of bank deposits held in transaction accounts. Source: Financial Sector
Accounts, Statistics Norway.
The composition of bank deposits by type of account is available from 1995 only. Chart 4
shows that households have had only 30-35 per cent of their deposits in transaction accounts,
whereas the corresponding ratio for non-financial companies has remained above 60 per cent
and is currently nearly 70 per cent. This may indicate that bank deposits are mainly
investment instruments for households, whereas non-financial companies have deposits
mostly for transaction purposes.
Chart 5: The gross financial assets of households and non-profit organisations by instrument.
Source: Financial Accounts, Statistics Norway.
0 %
10 %
20 %
30 %
40 %
50 %
60 %
70 %
80 %
90 %
100 %
Transaction accounts Other deposit accounts
0 %
10 %
20 %
30 %
40 %
50 %
60 %
70 %
80 %
90 %
100 %
Transcation accounts Other deposit accounts
Households sector Corporate sector
7
The composition of households’ gross financial assets is presented in Chart 5. The share of
bank deposits has declined from 44 per cent in 1975 to 32 per cent at end 2011. The decline
started after bank loan markets had been deregulated in the mid 1980s. Prior to that loans
were effectively rationed, and households would have to save in bank deposits in order to
qualify for a loan from that bank. The alternative investments whose share of total assets has
been increasing are mainly insurance reserves and equity investments. Equity holdings have
increased from 5 per cent of households’ financial assets in 1975 to 13 per cent in 2011, and
the share of insurance reserves from 23 to 38 per cent.
Insurance reserves are essentially the customer portfolios in pension funds and life insurance
companies. These portfolios are predominantly invested in fixed income (about 70 per cent in
2011) and equity instruments (about 20 per cent in 2011). Households thus hold about 50 per
cent of their financial assets as capital market investments, as compared to 30 per cent in
1975. This shift out of bank deposits and into market instruments may partly be because
higher income levels have stimulated long term saving for retirement or because returns have
become more attractive. But to a very large extent it reflects structural changes whereby less
of the long term saving desired by households has been provided through government non-
funded pension schemes and more from privately funded schemes. Finally, new financial
instruments introduced and sold by the banking sector may have tempted customers out of
bank deposits.
Chart 6: The gross financial assets of non-financial companies and local government by
instrument. Source: Financial Accounts, Statistics Norway.
Chart 6 similarly depicts the asset composition of non-financial companies. These time series
have a break in 1995Q4 when new definitions were introduced.
2
But we can still see a clear
pattern: The share of bank deposits declined by 8 percentage points from 1975 to 1995, and
by another 2 percentage points after 1995. The bank deposit share is currently 14 % of total
2
The main change was that loans from one non-financial company to another were included on both sides of the
balance sheet, thus blowing up both total assets and total liabilities.
8
assets. Equity holdings did on the other hand increase by 17 percentage points between 1975
and 1995, and by a further 15 percentage points after 1995. At end 2011 the equity share is 33
per cent of total assets.
Comparing charts 5 and 6 above, we notice that bank deposits have consistently been a much
larger share of total financial assets in the households sector than in the non-financial
companies. This would indicate that the sector allocation of gross financial assets in the non-
financial sector may have had an effect on the aggregate share of financial assets held as bank
deposits.
On the liability side of the balance sheet, bank loans are very important for the households
sector, with nearly 80 per cent of their total liabilities. Chart 7 shows that the increased
reliance on loans from private banks has a counterpart in less reliance on loans from
government sector lenders. The sum of loans from these two types of lenders has increased
slightly, from 84 to 92 per cent of households’ total liabilities.
Chart 7: The liabilities of households and non-profit organisations by creditor. Source:
Financial Accounts, Statistics Norway.
Chart 8: The liabilities of non-financial companies and local government by creditor. Source:
Financial Accounts, Statistics Norway.
9
Chart 8 shows the liabilities of the corporate sector. The break in the time series from 1995
reflects changes in definitions, the main implication being that more loans between companies
are included in the numbers (footnote 1). We notice that there have been only small changes
in the importance of loans from private sector banks, but a clear decline in the importance of
government lending.
The increased importance of bank loans to the non-financial sector thus partly reflects a
reduction in lending from the central government sector. Government lenders have reduced
their loans to both households and companies, and the private banks have been providing
most of these loans instead. The public sector share of lending to households and companies
have been steadily decreasing since 1981-82. This structural change in the Norwegian credit
market was a consequence of government policies to reduce the importance of lending
through government agencies. Notice that government lenders mostly did not take deposits.
This structural change in the loan market may thus be an important factor behind the decline
in the deposit-to-loan ratio at private banks.
We go on to consider how changes in the return of different investments may have changed
the incentives to make deposits or take loans. Government bond returns are available from
1990. Chart 9 shows that they have followed bank deposit rates closely since then, but with
the return differential tending towards zero. The returns on equity investments have been very
volatile, but for most of the period since 1984 well above the returns on fixed income
investments. There have been incentives to move funds into the stock market. Notice also that
the period 1989-95, when the deposit-to-loan ratio were increasing, had high real returns on
bank deposits and weak returns on the Oslo Stock Exchange.
Chart 9: Real post-tax returns (deflated by the CPI) on bank deposits, government bonds and
stocks listed on the Oslo Stock Exchange. Three years moving averages. Source: Statistics
Norway, Norges Bank and Oslo Stock Exchange.
On the liability side we first notice that the real borrowing costs of the non-financial sector
were high when the deposit-to-loan ratio was declining in 1989-95. The value of interest
10
payment tax deductibility was reduced from up to 50% down to a flat 28 per cent by the 1992
tax reform, but pre-tax interest rates have also been reduced. Chart 10 shows that the tax and
inflation adjusted cost of taking out a bank loan is currently less than 2 per cent. This is very
low by most standards. Notice, however, that the real cost of borrowing was negative in the
run-up to the 1988-92 banking crisis.
Households mostly borrow for buying homes. The profitability of borrowing to invest
depends on the house price inflation relative to the borrowing costs. Chart 10 does not take
maintenance costs into account, but the chart still strongly indicates that housing investments
have been profitable for most of the period after 1993. This followed a period during the
banking crisis of 1988-92 when these investments had been highly unprofitable.
Chart 10: Real post-tax costs of bank loans and real house price inflation (deflated by the
CPI). Three years moving averages. Source: Statistics Norway
The economic incentives to increased borrowing have thus been evident for the households’
sector during the past 20 years. Above we saw that incentives to hold bank deposits had been
weakened. In the period 1989-95 when the deposit-to-loan ratio increased, the economic
incentives to make deposits were temporarily stronger and the incentives to borrow weaker.
Eyeball econometrics seem to indicate that the portfolio model has some explanatory power.
4. Financial flows analysis
The gross financial assets of the non-financial sector can essentially be accumulated in three
different ways; by financial saving, by borrowing and by revaluation of existing assets. Chart
11 provides convincing evidence that asset accumulation in the non-financial sector mainly
comes from borrowing. The levels of total financial assets and total debt of the non-financial
sector move in parallel in the upper panel. In the lower panel we can observe the very tight
co-variation of changes in these two variables.
11
However, if we look at the households and non-financial company sectors separately, the co-
variation is less close, see Chart 12. This is in particular true for the households sector. Their
financial assets increased much faster than their debts during the 1990s, probably because
savings became important for consolidating their balance sheets after the banking crisis. The
opposite happened during most of the 2000s, when debt grew faster than assets.
Chart 11: Gross financial assets and total debt of the non-financial sector in per cent of GDP.
Levels in the upper panel and 4 quarter changes in the lower panel. Source: Financial
accounts, Statistics Norway.
Chart 12: Gross financial assets and total debt of the households sector (left hand panel) and
the non-financial corporate sector (right hand panel) in per cent of GDP. Source: Financial
accounts and national accounts, Statistics Norway.
0 %
50 %
100 %
150 %
200 %
250 %
300 %
350 %
Financial assts Total debt
40 %
50 %
60 %
70 %
80 %
90 %
100 %
110 %
1
9
7
5
Q
1
1
9
7
8
Q
1
1
9
8
1
Q
1
1
9
8
4
Q
1
1
9
8
7
Q
1
1
9
9
0
Q
1
1
9
9
3
Q
1
1
9
9
6
Q
1
1
9
9
9
Q
1
2
0
0
2
Q
1
2
0
0
5
Q
1
2
0
0
8
Q
1
2
0
1
1
Q
1
Financial assets Total debt
12
Chart 12 shows the debt and assets data in per cent of GDP. We can see that the non-financial
corporate sector has much more debt and also more financial assets than the households. The
households have now about 100 per cent of GDP on both sides of their balance sheet, with net
assets of around 10 per cent of GDP. The corporate sector has gross debts above 260 per cent
and assets above 170 per cent of GDP, with net debts at 80-90 per cent of GDP. Furthermore,
the households’ share of total assets in the non-financial sector has been declining from
around 55 per cent in 1975 to less than 40 per cent in 2011. That shift alone contributes to a
smaller fraction of assets in the non-financial sector being invested as bank deposits, confer
Chart 3.
Chart 13: The levels (left hand panel) and four quarter changes (right hand panel) of bank
deposits and bank loans to the non–financial sector. Millions NOK. Source: Statistics
Norway.
We have seen that debt creates financial assets. Similarly, bank loans create bank deposits. A
bank loan is given by making a deposit in the name of the borrower. The borrower normally
spends or invests the proceeds by transferring it to another deposit account and so on. But as
demonstrated in Chart 13 there is correlation, but not close co-variation of the two data series,
neither in levels nor in differences.
There are two reasons for these deviations, as illustrated in the financial flows diagram (Chart
14). First, the bank deposit chain has leakages whenever payments are made from the non-
financial sector to any outside sector. These outside sectors are the financial sector, the central
government and the foreign sector. Any investments of the non-financial sector in financial
instruments other than bank deposits will involve payments to the financial sector. The
increased equity investment and insurance reserves shown in charts 5 and 6 represent such
leakages. These increased leakages have contributed to the decline of the deposit-to-loan
ratio.
Second, transfers to the non-financial sector from the outside sectors has helped increase
bank deposits, but these transfers are not correlated with bank loan volumes. Looking more
closely on these transfers we notice that borrowing from the central government has been
declining, see charts 5 and 6, thus gradually contributing less to deposit growth. Borrowing
from abroad to the non-financial sector has increased, however, which is a net contribution to
?100 000
?50 000
0
50 000
100 000
150 000
200 000
250 000
300 000
350 000
400 000
Deposit growth Loan growth
0
500 000
1 000 000
1 500 000
2 000 000
2 500 000
3 000 000
3 500 000
Deposits Loans
13
domestic bank deposits. But borrowing from abroad is nearly exclusively to the corporate
sector, whose bank deposits are only a small portion of total financial assets.
Chart 14: Deposits in the financial flows diagram
Charts 7 and 8 above illustrated the importance of bank loans as a share of total debt of the
non-financial sector. Bank loans have become very important for the households sector where
they constitute nearly 80 per cent of total debt in 2011. But they are much less important for
the corporate sector, where the share is only 18 per cent.
Chart 15: The share of mortgage loans to total private bank loans to the households sector
(left hand panel) and changes in the deposit to loan ratio for households plotted against
changes in share of loans from public sector banks in households’ total liabilities (right hand
panel). Source: Statistics Norway.
30 %
40 %
50 %
60 %
70 %
80 %
90 %
100 %
Share of mortgage loans
?9 %
?6 %
?3 %
0 %
3 %
6 %
9 %
12 %
Change D/L ratio
Change public sector banks
14
In parallel with this transfer of credit extension from public sector to private sector banks, the
composition of the household loan book with private banks have changed. Chart 15 shows
that the share of mortgage loans has increased from 40 per cent of total private bank loans to
households in 1987 to 90 per cent in 2011. This naturally reflects the gradual transfer of
mortgage loans from government lenders to private banks. In the last few years it also reflects
the introduction of new loan products where households can use real estate as collateral for
loans intended for other purposes than real estate purchases. There has also been an increasing
share of mortgage loans with no repayments for extended periods.
The right hand panel of Chart 15 shows no close co-variation between the share of public
sector loans and the deposit-to-loan ratio for households. This may indicate that the shift of
mortgage loans from public sector lenders to private banks is not a good explanation for the
year-to-year changes in the deposit-to-loan ratio. It may still be important for explaining the
long term trend. Notice in particular the temporary upward shift both in the importance of
government lending and the deposit-to-loan ratio at private banks in the early 1990s.
5. Conclusions
We have been exploring possible explanations for the strong decline in the ratio between bank
deposits held by the non-financial sector in Norway, and the bank loans taken by the same
sector. We have followed two different but related lines of reasoning. First, we have been
looking at the relative return on bank deposits and on the relative costs of bank loans in a
portfolio model approach. Second, we have tried to follow financial flows to identify
structural changes in the Norwegian markets that could have an effect on the deposit-to-loan
ratio.
We did find that real deposit rates have been low during periods with strong decline in the
deposit-to-loan ratio, pointing to the attraction of alternative investment opportunities in
capital market instruments. This has either been done directly or through the participation in
funded pension schemes. During the periods of strong declines in the ratio we also found that
the real loan rate has been low. In combination with a significant house price inflation this
have made borrowing quite attractive to households.
The return incentives are particularly clear for the households sector. That is also the sector
where the deposit-to-loan ratio has been declining since 1995. For the corporate sector the
return incentives have been less clear-cut, and the decline in the deposit-to-loan ratio has been
modest.
On the structural side, we first find that funded pension schemes have become more important
during the past 30 years. This has probably to do not only with return incentives, but also with
the growing importance of privately funded pension schemes as a supplement to the public
non-funded pension schemes. Public pension schemes have become less satisfactory both
because the upper limit on benefits have become binding for more people, and recently also
because the level of benefits have been curtailed.
15
A second structural factor is that a large part of the lending to households has been transferred
from public sector agencies to private sector banks during the past 30 years. A similar
development has taken place for the corporate sector, but at a much smaller scale. Since
public sector lenders mostly did not take deposits, this represents an obvious factor behind the
long term trend. In parallel with this development, bank loan markets were deregulated,
making deposits less important as a qualifier for bank loans. But these processes have been
slow and may not contribute much to explaining the year-to-year changes in the deposit-to-
loan ratio.
We also find a counteracting factor in the increased reliance of the corporate sector on foreign
borrowing. Such financial inflows should lead to more deposits with Norwegian banks, at
least when the borrowers operate on the domestic market. But there are two reasons why this
effect may be small: First, much of the foreign borrowing has probably been to the petroleum
sector that to a large extent operates outside the Norwegian economy. Second, the corporate
sector holds only a small share of its financial assets as bank deposits. They do that mainly for
transaction purposes, and their transaction needs will not increase as a consequence of shifting
from domestic to foreign borrowing.
We are often told that banks should increase their deposit funding and reduce their reliance on
market funding, meaning that they should engineer a rise in the deposit-to-loan ratio. While
this is clearly desirable from a financial stability point of view, it is less clear how much the
banks can actually do. There is very little they can do with some of the structural changes that
have taken place. The strongest return incentive for household behaviour probably stems from
house price inflation, that individual banks are not able to control.
The banks can certainly bid higher deposit rates to attract deposits, but the aggregate effect on
total deposits in the system may be small. Banks can restrain their sales of new financial
products of the kind that have shifted investments out of deposits and into alternative financial
instruments, but that must a concerted effort to make much difference at the macro level.
What individual banks undoubtedly can do is to restrict lending. That will certainly also
reduce deposits, but the link between loans and deposits is likely to be much weaker at the
individual bank level than it is in aggregate. And if lending growth is not restricted, one
contribution individual banks could make to financial stability would be to make more of their
market funding long term.
References:
Beltratti, Andrea and Stulz, René M. (2012), The credit crisis around the globe: Why did
some banks perform better? Journal of Financial Economics 105, 1-17.
Cohen, Bruce C. and George G. Kaufman (1965), Factors Determining Bank Deposit Growth
by State: An Empirical Analysis. Journal of Finance, 20 (1), 59-70.
16
Demirgüc-Kunt, Asli and Harry Huizinga (2010), Bank activity and funding strategies: The
impact on risk and returns. Journal of Financial Economics 98, 626-650.
Disyatat, Piti (2011), The bank lending channel revisited. Journal of money, Credit and
Banking 43 (4), 711-734.
Edminister, Robert O. and Henry E. Merriken (1989), Measuring interest rate sensitivity of
consumer depositors, Journal of Financial Services Research 2, 133-145.
Hammersland, Roger and Cathrine Bolstad Træe (2011), The financial accelerator and the
real economy. Discussion Paper 668, Statistics Norway.
Norden. Lars and Martin Weber (2010), Funding modes of German banks: structural changes
and their implications. Journal of Financial Services Research 38, 69-93.
doc_915949780.pdf