RDP 2013-15: Trends in the Funding and Lending Behaviour of Australian Banks 2. Funding Behaviour

A distinguishing feature of the Australian banking system, relative to its international peers, has been its greater use of wholesale funding, particularly offshore funding (Table 1). This reflects the influence of a range of macroeconomic, institutional and bank-specific factors over the past couple of decades. It is worthwhile noting that differences in statistical definitions play some role in explaining the Australian banking sector's low use of deposit funding. For example, the RBA treats certificates of deposit and intragroup deposits from offshore entities as wholesale debt based on an assessment of how ‘sticky’ these funding sources might be during a crisis. If they were instead treated as deposit liabilities (as they are in Canada and the United States, where they are covered by deposit insurance), this would add 10 percentage points to the deposit share of funding in Australia and reduce the (short-term) wholesale funding share by the equivalent amount.

2.1 Historical Drivers

A number of factors play some role in the Australian banking system's greater use of wholesale funding relative to some of its international peers. One of the most important was the strong growth in lending to Australian households in the two decades prior to the global financial crisis. This was driven by the household sector's willingness to increase its gearing that was facilitated by deregulation and the shift to a lower inflation environment, which decreased the cost and increased the availability, of finance. Indeed, Kent et al (2007) note that Australia is one country where these factors are likely to have been particularly strong and operating simultaneously. Banks also continued to intermediate funding for Australian corporations.

Deposit growth was somewhat slower than credit growth, although it still grew faster than nominal income (Figure 1). This was partly associated with the asset allocation decisions of households. For example, since the early 1990s, Australian households have had a relatively low and decreasing share of their wealth held in financial (for example, non-housing) assets and, within financial assets, a declining share has been held in the form of bank deposits (Ynesta 2008; Davis 2013).^[1]

The low share of household financial assets held directly as deposits might be partly explained by the fact that Australia's compulsory superannuation scheme accounts for a large share of household wealth. This is consistent with cross-country evidence that suggests higher shares of household financial wealth held in superannuation are associated with lower shares of financial assets held directly as deposits (Figure 2). At the same time, however, Australian insurers and superannuation funds hold a relatively large share of their assets in the form of deposits. In particular, the share of Australian superannuation assets invested in deposits is more than double the median asset allocation share for a number of OECD countries (see Section 2.2). This means that Australian households' direct and indirect holdings of deposits are not as low as direct holdings alone would suggest.

In this environment, the increased demand for credit was accommodated through wholesale funding, and much of this funding was obtained offshore. A large share of financial inflows to Australia were intermediated by the banking sector, because the demand for credit was largely from the household sector that did not have direct access to these markets (Debelle 2013a). Regular offshore debt issuance by the banks to support housing lending also reinforced their advantage in issuing more cheaply, and for longer tenors, than was possible for corporations that issued only infrequently.

The trend toward offshore wholesale funding allowed banks to diversify their funding sources and was supported by the ability of banks to effectively hedge foreign-currency risk. About three-quarters of funds sourced from offshore are denominated in foreign currencies, and the vast bulk of these funds are hedged back into Australian dollars.^[2] This activity was complemented by natural counterparties for the other side of the transaction. In particular, offshore issuers wanted to access Australian investors by issuing Australian dollar-denominated bonds in Australia and hedging this back into foreign currency (Ryan 2007). Furthermore, a sizeable proportion of Australia's superannuation assets are invested in foreign currency-denominated assets, some of which are hedged back into Australian dollars.

Changes in regulations and a shift in emphasis away from ‘asset’ towards ‘liability’ management also encouraged the shift towards wholesale funding from the late 1970s to the late 1980s (Battelino and McMillan 1989). For example, banks started to intermediate a greater share of lending through bill acceptances rather than through deposits, partly to avoid the Statutory Reserve Deposit requirement. In the 2000s, banks' use of wholesale funding was supported by financial innovation, particularly the parcelling and redistribution of risk through RMBS that were designed, in part, to reduce the capital required to support a given amount of credit (see below).

The size, composition and behaviour of banks operating in Australia are other factors that influence the share of wholesale funding in the banking sector. In terms of size, empirical analysis generally suggests that larger banks are more likely to use wholesale funding because they are better able to overcome the fixed costs involved, particularly with securitisation programs (see, for example, Uzun and Webb (2007), Bannier and Hänsel (2008), and Panetta and Pozzolo (2010)). Given that each of the major Australian banks was among the largest 50 banks globally in 2013, according to both total assets and Tier 1 capital, this evidence is consistent with their relatively high share of wholesale funding.

In terms of composition, while foreign-owned institutions generally have considerably more wholesale debt than domestic banks, it is not clear from the available data whether they represent a larger share of banking system assets in Australia than in other countries (Figure 3).^[3] Having said that, differences in the relative importance of foreign-owned banks are likely to explain some of the divergence in funding patterns between Australia and countries such as Canada and Japan where there is very little foreign bank activity. Furthermore, the nationality of foreign banks operating in Australia might have some role to play, because banks from different home countries use different expansion and funding patterns (CGFS 2010).

In terms of behaviour, both larger and smaller Australian institutions made increasing use of securitisation before the onset of the global financial crisis. The fall in the cost of securitisation, relative to other potential funding sources, over the late 1990s and early 2000s was sufficient to more than offset the fixed costs of issuance.^[4] This allowed some smaller, non-deposit-taking lenders to compete despite having neither balance sheets nor capital bases from which to fund their lending in the traditional sense. They also had the advantage of low costs given that they were unencumbered by large branch networks. The use of RMBS funding allowed the market share of these wholesale lenders to rise to over 10 per cent of housing credit from less than 2 per cent in the late 1990s, while the share of housing lending financed through securitisation rose from about 5 per cent in the mid 1990s to over 20 per cent by mid 2007.

From a risk management and stability perspective, the maturity profile of wholesale debt is an important consideration. Relative to other banking systems, a large share of Australian banks' wholesale debt is issued with an original maturity of more than one year; this helps to reduce liquidity risks by reducing the rate at which debt rolls over (Figure 4).^[5] For wholesale debt issued with original maturities of less than a year, Australian banks make relatively little use of repo funding, which tends to be very short term, and the maturity of Australian banks' offshore money market funding is generally longer than it is for many of their overseas counterparts.^[6]

Wholesale funding can also help to reduce liquidity risks by diversifying the investor base on the liability side of banks' balance sheets. It also allows banks to react more quickly to unexpected increases in loan demand on the asset side of their balance sheets. More generally, the relative transparency of the Australian banking system's activities and its strength – proxied by the quality of its assets, credit ratings, etc – would support a higher share of wholesale funding.

2.2 Recent Influences and Implications

Following the onset of the global financial crisis in 2007, there has been a marked increase in the focus on deposit funding by regulators, ratings agencies, investors, and banks' own management (Edey 2010). This has been driven by the perceived stability of deposit funding relative to short-term wholesale debt and has resulted in the share of deposit funding in Australia rising to be closer to that in many other countries. At the individual bank level, increasing deposit funding is a way to reduce liquidity risks. For the banking system as a whole, it can be argued that more deposit funding increases the resilience of the supply of credit to shocks in wholesale funding markets. To the extent that these shifts are related to Basel III and other regulatory changes, they are likely to persist. The increased focus on deposit funding has also been driven by the rapid increase in the cost of some other sources of funding; for example, spreads on RMBS rose to unprofitable levels during the financial crisis.

At the same time as the demand for deposit funding from banks has increased, there has been a significant increase in the supply of deposits from Australian households and businesses. All else equal, this has moderated the increase in the cost of deposit funding that would have otherwise occurred. The increase in the supply of deposits has been particularly marked from superannuation funds; the consolidated position of the Australian household sector has gone from investing a relatively low share of its financial assets in deposits relative to other OECD countries to having a roughly comparable share (Figure 5). A notable aspect of this increase has been the role of self-managed superannuation funds. These funds invest a significantly higher share of their assets in deposits, and other interest-bearing securities, than other superannuation funds, and have become increasingly important over the past decade.

Changes in the nature of capital inflows to Australia are also likely to be influencing the banking system's relative use of deposit funding.^[7] In recent years, a greater share of capital inflows to Australia have come from resource companies either borrowing in wholesale markets or through retained earnings, rather than the banking sector (Debelle 2013a). This implies that these companies have had a lower demand for credit intermediated by the Australian banking sector for a given level of activity. At the same time, the amount of deposits ultimately available to the banking system is not likely to have changed much because the proceeds of the corporations' wholesale debt issues still circulate in the financial system. In combination with the fall in demand by Australian banks for offshore wholesale funding, this implies an increase in the deposit share of funding, which may have moderated some of the increase in the price of deposits relative to wholesale debt.

2.2.1 Implications for risks

Changes in the composition of funding will alter the nature of liquidity risk in the banking sector. How this occurs will depend on a number of factors, including the relative maturity of deposits versus wholesale debt and how banks respond to changes in liquidity risk.

Following the experience of the financial crisis, it is generally argued that deposits are stickier than short-term wholesale debt given switching costs and the presence of (capped) deposit guarantees or depositor preference arrangements. In Australia, for example, while most deposits are ‘at call’, they generally have quite long maturities when depositors' actual behaviour is considered.^[8] Furthermore, term deposits have been responsible for a disproportionate share of the growth in deposits over recent years and now comprise about a quarter of total bank funding. These accounts have an average initial maturity of about six months, and typically roll over a number of times before being withdrawn. In contrast, the average residual maturity of banks' short-term wholesale funding is somewhere between two and three months.

However, the advantages for liquidity of issuing longer-maturity wholesale debt and ensuring a maturity profile that does not have a large share of liabilities maturing in a short period of time should not be overlooked. The average maturity of Australian banks' long-term wholesale debt is over five years at issuance. Even though the weighted average maturity of domestic debt at issuance has shortened by about a year since the onset of the financial crisis, it remains at over five years, and the weighted average maturity of offshore wholesale debt has increased to just under seven years, driven by the issuance of relatively long-term covered bonds (Figure 6).^[9]

Furthermore, any reduction in liquidity risk arising from a greater share of deposit funding would be illusory if it were to come from investors merely substituting wholesale debt for wholesale deposits. Likewise, an increase in competition for deposits could make them less stable from an individual banks' perspective by encouraging depositors to be more price-sensitive.

The liquidity-risk advantages of increasing the share of deposit funding could also be offset to some extent if banks were to increase their credit or market risk exposures to boost their returns, given higher funding costs associated with a switch to funding from sources with lower liquidity risk (discussed below).^[10] This is more likely to happen if there is less market discipline on banks' risk-taking because there is less monitoring by investors holding banks' short-term debt (Calomiris and Kahn 1991), although Eisenbach (2013) suggests that this disciplining role of short-term wholesale debt is too weak during normal market conditions and then excessive during periods of market disruption. The incentives for short-term debt holders to monitor banks are likely to increase with the share of funding that is secured against balance sheet assets, for example through covered bond issuance (Gai et al 2013). Following the crisis, there has been some shift in investors' preferences towards secured funding, although it has been relatively muted in Australia compared with the shift for euro area banks (CGFS 2013).

More generally, changes in the nature of capital flows which contribute to an increase in the deposit share of funding might alter the aggregate amount of liquidity risk in the economy. In particular, it is possible that the corporate sector might not be as cognisant of, or well placed to undertake, liquidity risk management as the banking sector. The ability of authorities to detect and react to the build-up or realisation of these risks also differs according to which sectors have undertaken these activities.

2.2.2 Implications for funding costs

The increased emphasis on deposit funding has, and will continue to have, important implications for bank funding costs. In particular, banks' funding costs are a function of the composition of funding, the cost of the various funding sources, and hedging strategies. One of the most obvious effects of the increased demand for deposit funding, particularly term deposit funding, and the reduction in the use of short-term wholesale debt by Australian banks, has been an increase in the cost of deposits relative to those of wholesale funding (Figure 7; Edey 2010). This has contributed to an increase in banks' overall funding costs relative to the cash rate (Robertson and Rush 2013).^[11]

The shift towards term deposits, which are generally repriced less frequently, may have marginally slowed the speed with which changes in the pricing of funding instruments flow through to overall funding costs. The transmission of monetary policy to changes in bank funding costs has, however, remained little changed (see Section 3.2.1 for more discussion).

2.2.3 Implications for the economy

Over recent years, Australian banks have been funding new loans through new deposits, with little net issuance of wholesale debt (Figure 8; RBA 2013). This has resulted in a debate about whether deposit growth will place any form of constraint on lending growth should the rate of credit growth increase or domestic savings decrease (Coffey 2012).

The savings rate is the result of decisions being made by individuals responding to macroeconomic factors such as the level of interest rates. The relationship between the savings rate, deposit growth and credit growth depends on the distribution of individuals in the economy and their preferences for different investment and saving vehicles. These relationships will also depend on whether the demand for credit is satisfied directly from the financial markets or intermediated through the banking system. As such, the relationships between the interest rates on deposits, the savings rate, preferences regarding different savings instruments, and the growth rates of deposits and credit are complex. For example, a savings rate of zero can be consistent with the banking system being fully funded by deposits given that the savings rate is calculated from the income and savings of both savers and dis-savers. The relationships are even more complicated in an open economy setting.^[12]

One way to think about these relationships is to consider how the share of deposit funding might be affected by changes in behaviour of different agents, including the official, authorised deposit-taking institution (ADI) and private non-ADI (for example, household) sectors.^[13] For example:

• If the Federal Government funds a budget deficit by issuing Commonwealth Government securities (CGS), there will be no effect on the value of deposits if those securities are purchased by the private non-ADI sector. The initial decline in the deposits held by the non-ADI sector (upon purchase of CGS) will be offset by a subsequent increase in this sector's deposits as the Government spends the proceeds of the CGS sale. In contrast, if the CGS are purchased by the ADI sector (hereafter banks), the private non-banking sector's holdings of cash and deposits may increase. For example, if banks purchase CGS using their holdings of cash, these Australian dollars re-enter the economy in the hands of the non-banking sector as the Government uses the funds obtained from the sale of CGS. The end result is that the non-banking sector's balance sheet has expanded while the banking sector's balance sheet has changed its composition and, depending on the investment preferences of households, potentially expanded.
• Changes in the banking sector's asset allocation can also affect total deposits in the economy. Consider, for example, a desire by the banking sector to increase its holdings of CGS through purchases from the household sector using cash. The household sector will place a significant share of these proceeds back with the banks as deposits. The end result of this change in preferences is that the banking sector will have a larger balance sheet (although no more credit has been provided to the private sector) with a higher share of funding sourced from deposits. Meanwhile, the household sector's balance sheet has simply changed its composition. This process will be moderated by falls in the yields on CGS relative to the deposit rates banks need to pay to ensure that households are willing to sell their CGS and invest the proceeds in deposits.

• The provision of credit by banks to the non-bank sector will ultimately generate an increase in the supply of funds back to the banking sector.^[14] Whether these funds are returned in the form of deposits or some other liability will be influenced by the returns on different liabilities of the banking sector and investors' preferences towards different types of financial instruments. An increase in interest rates on deposits relative to certificates of deposit will, for example, result in holders of certificates of deposit switching to deposits. Such a divergence in interest rates on deposits and wholesale debt of equivalent maturities has been apparent in recent years (Figure 7).^[15]

  In contrast, changes in investors' preferences between deposits and non-banking sector assets do not necessarily influence the aggregate value of deposits. Households purchasing existing shares using their deposits, for example, will not reduce the aggregate value of deposits if the proceeds of these purchases are also held as deposits by the original owner of these equities.

How the relationship between deposit and credit growth evolves over the short to medium term is difficult to predict. There is, however, no reason to think that any potential adjustment to either a rise in the demand for credit or a reduction in the supply of deposits would be disruptive from a macroeconomic perspective. In particular, the relative price of deposits and banks' wholesale debt is likely to adjust, as it has over recent years, in response to changes in the desirability of different funding sources. Furthermore, in responding to either unexpectedly high credit growth or lower deposit growth, financial institutions can partially moderate any gap between their budgeted credit and deposit growth rates by altering the interest rates on their lending, deposits and wholesale debt. More broadly, financial institutions may re-evaluate their reluctance to use additional wholesale funding in an environment where demand for credit picks up or the supply of deposit funding slows, given that this is likely to be associated with improved perceptions about the strength of the macroeconomy and financial institutions.

Another implication of the recent increase in the deposit share of funding for some institutions is that the competitive dynamics in lending markets could have become more cyclical. In periods when the volatility of financial markets is relatively low, wholesale funding costs typically fall relative to deposit prices. This provides a competitive advantage to institutions that rely on wholesale funding, rather than an expensive deposit base, and increases the number of institutions in the lending business. However, when wholesale funding conditions become more volatile, the balance of competition will swing the other way in favour of institutions with large deposit bases. Recent experience suggests, however, that the advantage of a larger transaction deposit base in times of financial market volatility may have been mitigated somewhat because the growth in deposits has been in relatively expensive forms, such as term deposits.

In the very long run, there is also potential for there to be more intermediation of savings and lending through financial markets or superannuation funds (Davis 2013). This might, for example, arise if there are further increases in the relative cost of deposits versus wholesale debt or through growth in the size of superannuation assets relative to the rest of the financial system. How intermediation evolves will be a function of many factors, including: the value depositors place on features specific to deposits, such as their low risk due to deposit insurance, access to the payments system etc; the type of superannuation funds that experience the most rapid asset growth, given differences in their asset allocations; and whether some superannuation funds are willing to take on more of the risks traditionally borne by the banking sector by becoming the generator of assets rather than the purchaser.

Footnotes

To some degree, the relatively high share of household wealth invested in housing assets might reflect the effects of geography and urbanisation on house prices (Ellis and Andrews 2001). [1]

Banks hedge foreign currency-denominated funding to protect themselves from the risk that a depreciation of the Australian dollar will increase the Australian dollar value of their foreign currency-denominated debt and reduce the value of their equity. This is encouraged through APRA's prudential standards. For larger amounts of wholesale debt, it is also normal practice for banks to match the maturity of their hedges to their borrowings. For a broader discussion of foreign currency hedging by Australian entities see Rush, Sadeghian and Wright (2013). [2]

Foreign banks include both branches and subsidiaries. Note that branches do not have to hold any equity in Australia. [3]

The use of RMBS was also facilitated by improvements in technology, the fall in nominal interest rates (which made banks' deposit funding less of an advantage) and, from an investor's perspective, a relative lack of other high-quality fixed income securities. [4]

Within wholesale debt issued with maturities of longer than one year, Australian banks have a similar maturity profile to their overseas counterparts. [5]

Some Australian banks use a small proportion of (offshore) short-term wholesale debt to exploit arbitrage opportunities; some offshore short-term assets are funded through offshore short-term liabilities to earn additional interest with no maturity mismatch involved. [6]

Changes in the willingness of the Australian banking system to use different types of funding will also influence the nature of capital flows to Australia. [7]

Although in a crisis, such behaviour by depositors may change considerably. [8]

A number of factors underpinned the decline in the residual maturity of domestic long-term debt. These include the low level of new issuance combined with the shorter maturity of debt issued during the financial crisis. [9]

Recent history provides examples of banks with high shares of deposit funding taking on significant levels of credit and market risk to generate returns. The heavy use of deposits relative to wholesale debt can also mean that losses are concentrated domestically following any failure, rather than being spread across sophisticated investors around the world. [10]

All else equal, by reducing banks' liquidity risk, this compositional change should have reduced the cost of wholesale funding and/or the amount of liquid assets that need to be held. [11]

In particular, while capital inflows do not alter the amount of domestic currency in circulation, they could alter the supply of different types of assets if the portfolio preferences of non-resident purchasers of Australian dollars differ from those of the resident sellers of these assets. [12]

The ‘Formation Tables’ framework was used extensively in the 1970s and 1980s to monitor changes in private non-bank Australian dollar-denominated deposits and the money supply, and to understand how these were affected by the activities of the Government, central bank and private banks (Valentine 1984). These tables provide an accounting-based reconciliation of these entities' balance sheets. However, care must be taken with the interpretation of formation tables because they do not provide insights into the causes of money or credit growth. They also became less useful following the floating of the Australian dollar and as a growing body of literature cast doubt over the reliability of the relationship between monetary aggregates and real economic activity. New theoretical models place more emphasis on the information content of credit, rather than money, for economic activity. [13]

This discussion focuses on how changes in deposits facilitate bank lending rather than on banks ‘funding’ loans through endogenously creating deposits on their own balance sheet. The latter situation of a bank creating both loans and deposits simultaneously is a possibility in the case where the bank has adequate liquidity and capital. Furthermore, increasing the amount of loans and deposits simultaneously needs to be facilitated by offering the non-bank sector increasingly attractive rates on deposits relative to loans, reducing the bank's profitability until further expansion is no longer attractive and curtailing the demand for credit. [14]

This divergence in the deposit and wholesale funding rates of the major banks is also influenced by competition for deposits from banks with lower credit ratings. That is, in equating the marginal cost of deposits and (lower-rated) wholesale debt, these institutions can increase the cost of deposits for other, higher-rated, institutions. [15]