RDP 2001-07: A History of Last-Resort Lending and Other Support for Troubled Financial Institutions in Australia 12. Conclusion

During the nineteenth century isolated bouts of illiquidity and broader swings in the economic cycle prompted colonial governments and the banking industry association to provide last-resort loans and other forms of support to troubled banks on several occasions. Since Federation, however, there have been just two instances where last-resort support has been provided:

1. the loans provided to the Primary Producers Bank by the Commonwealth Bank in 1931; and
2. the loans provided to three private banks by the Reserve Bank in support of those banks' efforts to provide funds to illiquid building societies in 1974 and 1979.

The third instance of lending to a troubled financial institution, the Reserve Bank's provision of a liquidity facility to the Bank of Adelaide in 1979, was not a last-resort loan but rather a means of smoothing the takeover of the Bank of Adelaide by the ANZ.

Australia's experience highlights three preconditions that are needed to ensure the effectiveness of a lender of last resort. The lender of last resort needs to be able to readily generate liquidity, be immune from competitive pressures and limit the moral hazard consequences of the provision of support.

Firstly, a lender of last resort needs to have the means to provide substantial injections of liquidity to the banking system. The Associated Banks were constrained in the liquid resources they could draw on during the 1890s as the member banks were in straightened circumstances. The colonial governments also faced budgetary constraints through that period. They, therefore, adopted measures that did not require direct expenditure. The NSW Government's declaration of private banks' notes to be legal tender was successful in making banks more liquid without the government incurring any expense. In contrast, the Victorian Government's declaration of a week-long banking holiday froze most of what liquidity remained in the banks. While the Commonwealth Bank had ample capacity to provide liquidity during the 1930s, its concern (however misguided) to avoid inflationary expansion limited its willingness to provide support. The NSW Government's inability to produce high-powered money saw its guarantee of the Government Savings Bank count for nought.

Where the public sector is able to create high-powered money this gives it an advantage in the provision of last-resort support. Nevertheless, private sector support has been effective in several instances. The Associated Banks' support of the City of Melbourne Bank in 1879, support for the City Bank of Sydney from other Sydney-based banks in 1893, the National Bank of Australasia's financing of the Queensland Deposit Bank in 1931 and bank loans to building societies during the 1970s all suggest that the market is well able to address idiosyncratic or isolated liquidity difficulties.

Once it is publicly known a lender of last resort has the resources to effectively meet system-wide demands for liquidity, this gives the lender credibility that makes it easier to nip potential crises in the bud. Bagehot argued that a lender of last resort should clearly state its willingness to lend. In such a case, rather than direct lending, all that may be required is a reassuring public statement. The Commonwealth Bank's declaration of its strength and the government's support quelled the run on the Commonwealth Savings Bank in 1931. Statements regarding Reserve Bank support for regional banks were sufficient to stem runs between 1987 and 1992. However, the Australian experience also demonstrates that a lack of clarity in lender-of-last-resort policy can exacerbate public doubts in times of financial instability. The Associated Banks' reticence to publicly state the terms under which they were prepared to provide support added to public confusion concerning their role. Similarly, the equivocal nature of the Commonwealth Bank's statements concerning the amalgamation with the Government Savings Bank weakened public confidence.

The second aid to the successful conduct of lender-of-last-resort policy is immunity from competitive pressures. While industry associations in some other countries have successfully served as lenders of last resort, in Australia the Associated Banks struggled to move beyond the narrow interests of individual members. Similarly, charges of a conflict of interest between the Commonwealth Bank's central banking responsibilities and its commercial banking operations impaired its ability to act as a lender of last resort during the 1930s.

Since the 1930s, it has been the preference of the Commonwealth Bank, and then the Reserve Bank, to provide support via the market rather than directly to individual institutions. So long as the regulatory environment constrained competition, the Reserve Bank was able to persuade banks to override strategic considerations (which might have seen the banks leave their competitors starved of funds) when considering appeals for support. As a result, the arrangements for the support of building societies and the takeover of the Bank of Adelaide worked well. As deregulation proceeded and competition within the financial sector intensified, however, this approach became more difficult to apply. St. George's criticism of the terms of the support provided to it in 1979 was one symptom of this.

The third precondition for successful lender-of-last-resort policy is an ability to limit moral hazard costs. Australia has seen remarkably few instances of moral-hazard-driven behaviour (Provincial and Suburban, the Queensland National Bank and Farrow being notable exceptions). This suggests that, overall, lender-of-last-resort policy, in combination with other banking policies, has been successful in avoiding undue support of banks' risk-taking.

In three instances, an inability to monitor or limit potential moral hazard behaviour restricted the provision of last-resort support. The Associated Banks had little ability to monitor or penalise imprudent behaviour among its members. The limitations on the Commonwealth Bank's ability to monitor the operations of the private banks reinforced its bias against providing extensive support during the 1930s. Since the Reserve Bank had no regulatory responsibility for building societies, it confined its lending to institutions it directly supervised and relied on market discipline in providing funds to building societies during the 1970s.

In Australia, penalty rates have rarely been applied to last-resort loans. Instead, more direct constraints on banks' risk-taking have been applied. The conditions accompanying the government's loans to the Bank of NSW in the 1820s applied a very direct brake to the bank's risk-taking. The imposition of prudential standards on banks supported the Reserve Bank's public statements of support for regional banks in the late 1980s and early 1990s.

Another constraint on moral hazard behaviour is the restriction of last-resort support to solvent institutions. This requires distinguishing between illiquidity and insolvency. The Associated Banks did successfully discriminate between illiquid and insolvent institutions in handling the three requests for assistance it received in 1879. During the 1890s, however, broader considerations concerning the Associated Banks’ ability to provide support largely overrode distinctions between illiquidity and insolvency. During the 1930s, the Commonwealth Bank relied heavily on banks' ability to provide suitable collateral for last-resort loans to separate illiquidity and insolvency. In providing support to the Primary Producers Bank and the Government Savings Bank, the Commonwealth Bank was prepared to lend against all assets that were conventionally regarded as good collateral. The Commonwealth Bank did not, however, attempt to look beyond the banks' current condition to form some assessment of their pre-panic value.

Some, such as Mishkin (1998), posit that central banks should adopt a rules-based approach to combat moral hazard. At no time, however, have fixed rules been used to determine lender-of-last resort policy in Australia.^[87] During the twentieth century, the central bank has preferred to make the provision of last-resort support contingent on conditions in financial markets and the economy. For example, the Commonwealth Bank argued that such considerations contributed to its decision not to support the Federal Deposit Bank in 1931. While stopping short of complete ambiguity, such an approach does engender some uncertainty as to the exact conditions under which support would be provided to individual institutions, thus reducing moral hazard. Judging the systemic impact of potential bank failure, however, adds another element of difficulty to implementing lender-of-last-resort policy. The Victorian Government, for example, thought that the failure of the Commercial Bank of Australia in 1893 would have little systemic impact.

The provision of support to banks (and other deposit-taking institutions) has often been motivated by considerations other than financial stability. While the loans provided to the Bank of NSW in the 1820s were made out of concern for the effect of the bank's closure on the colony's economy, the need to foster competition was also an important consideration. Government involvement in savings banking was based on social policy considerations (the encouragement and protection of low-income depositors) and as a means of supporting regional development. The Queensland Government's introduction of a government note issue in 1893 was partly motivated by a desire to tap a cheap source of funding. State government support of building societies aimed to foster housing construction. While contributing to those objectives, the provision of such support muddies public perception of the riskiness of banks and risks fostering moral hazard behaviour. This is illustrated by the failure of the Farrow Group of building societies and the State Banks of Victoria and South Australia during the early 1990s.

The need to balance the provision of relief for sound institutions experiencing liquidity problems with moral hazard costs of supporting unviable institutions suggests that the ability to ensure smooth exit of troubled financial institutions is an important adjunct to the lender of last resort. This was borne out by the effective containment of spillover effects from the closure of the Bank of Adelaide.

The failures that have occurred, however, reinforce the notion that the systemic label should not be overapplied. The banks that failed in the 1840s were fundamentally insolvent. The banking system was well able to continue functioning while they failed. While liquidity shortages forced most banks to suspend payment during the 1890s, the banking system as a whole continued to function. Similarly, the systemic consequences of the failure of the Primary Producers Bank and the Federal Deposit Bank in 1931 and the Bank of Adelaide in 1979 were limited. Although the failure of the Farrow Group did spill over to other non-bank financial institutions and smaller banks that had not long converted from building society status, overall financial contagion was well contained. On several occasions, market-based solutions have developed to ease the banking difficulties. During the 1890s, secondary markets in frozen bank deposits developed. Such markets also sprang up during the Government Savings Bank's suspension of payment in 1931.

Footnote

The automaticity of lending under the LGS convention followed from the convention's role as an instrument of monetary policy, rather than lender-of-last-resort considerations. [87]