Capital levels

Recommendation 1

Set capital standards such that Australian authorised deposit-taking institution capital ratios are unquestionably strong.

Description

APRA should raise capital requirements for Australian ADIs to make ADI capital ratios unquestionably strong. A baseline target in the top quartile of internationally active banks is recommended. This principle should apply to all ADIs but is of particular importance for ADIs that pose systemic risks or access international funding markets.

The target would be aided by adopting Recommendation 4: Transparent reporting, which aims to improve the international comparability of Australian ADI capital ratios.

The Inquiry’s judgement is that, although Australian ADIs are generally well capitalised, further strengthening the banking sector would deliver significant benefits to the economy at a small cost. Evidence available to the Inquiry suggests that the largest Australian banks are not currently in the top quartile of internationally active banks. Australian ADIs should therefore be required to have higher capital levels.

The quantum of any change should take account of the effect of other recommendations, particularly Recommendation 2: Narrow mortgage risk weight differences, which aims to improve competitive neutrality of regulatory settings.

Objectives

  • Make banks less susceptible to extreme but plausible adverse events — such as asset price collapses — unexpected loan losses or offshore funding shocks, to reduce the likelihood of bank failures and promote trust and confidence in the banking sector.
  • Create a financial system that is more resilient to shocks and thus less prone to crises, which can have devastating and long-lasting effects on the economy and society.
  • Protect the Government balance sheet from risks in the financial system, to minimise the burden on taxpayers.
  • Reduce perceptions of an implicit Government guarantee for ADIs and the associated economic inefficiency.

Discussion

Problems the recommendation seeks to address

Importance of capital in the banking sector

Capital, particularly equity capital, is an essential element in both actual and perceived financial soundness, acting as a shock absorber for unexpected losses. Once equity has been exhausted, a bank is generally non-viable — and could well have been before that point. Equity capital is therefore an important determinant of how likely a bank is to fail. Capital is also a safety buffer for creditors, as it is typically exhausted before the bank defaults on its obligations. By making creditor funds relatively safer, high levels of capital assist to maintain confidence in a bank, even in times of market stress.

Making banks safer and enhancing investor confidence both contribute to reducing the likelihood of a financial crisis. Shocks will always buffet the financial system, whether they are generated domestically or overseas. Capital is one of the best protections against those shocks generating a crisis.

Although banks choose capital at levels that account for their own specific risks, this does not account for the risks the banking system poses to the broader economy.

Risks and costs of financial crises

Australia’s financial system is a vital part of the economy, providing avenues for saving, investment and funding growth. However, it also poses risks that must be managed. This includes minimising the likelihood of future financial crises, which can have significant costs for individuals, the economy, Government and taxpayers.

Australia should not underestimate the risks of financial crises

Australia’s resilience during the GFC partly reflected the strength of the financial sector, the quality of its regulatory framework and supervision, and Government’s assistance to the financial sector. However, many other important factors were also at play, including macro-economic policy, a strong Government balance sheet and Chinese resource demand. Given these supporting factors may not be present in the next crisis, Australia should not become complacent about the risks of financial crises as a result of its GFC experience. Australia is not immune to financial crises.

Australia can draw lessons from other countries’ GFC experiences, which highlighted that financial systems are vulnerable to low-probability, high-impact ‘tail events’, which can be caused by large external shocks or be generated domestically. An asset value shock of similar magnitude to those experienced by overseas banks during the GFC would cause Australian banks significant distress.

For example, the major banks currently have a leverage ratio of around 4–4½ per cent based on the ratio of Tier 1 capital to exposures, including off-balance sheet. An overall asset value shock of this size, which was within the range of shocks experienced overseas during the GFC, would be sufficient to render Australia’s major banks insolvent in the absence of further capital raising. In reality, a bank is non-viable well before insolvency, so even a smaller shock could pose a significant threat. Following its recent stress-test of the industry, APRA concluded, “… there remains more to do to confidently deliver strength in adversity”.6

Financial crises have large costs

The costs of financial crises are high, and their effects broad-ranging. Financial crises:

  • Significantly constrain households’ and businesses’ access to credit, inhibiting the ability to invest, buy a home or grow a business.
  • Potentially affect confidence in banks, which in turn may impact confidence in the operation of the payments system.
  • Are historically associated with an average rise in unemployment of seven percentage points, which in 2014 would be almost 900,000 additional Australian workers.7 Recovery from financial crises can be protracted, with high unemployment continuing for a long period.
  • Can result in large contractions in trade credit and make it more difficult for businesses, including small businesses, to access financing.
  • Can substantially reduce the wealth and savings of a generation, particularly for those with lower initial wealth. This may particularly harm those people who rely most on savings, such as those in or close to retirement.
  • Create large falls in GDP. International estimates of the GDP cost of a crisis are19–158 per cent of one year’s GDP — which for Australia would have equated to $300 billion to $2.4 trillion in 2013 — with a median of around 63 per cent of GDP ($950 billion).8
  • Erode a government’s fiscal position. The need to directly support the financial system, along with the deteriorating budget position due to the associated recession, causes a substantial increase in net government debt. According to International Monetary Fund (IMF) data, general government net debt between 2007 and 2013 as a share of 2013 GDP rose by around 40 percentage points in the United States, 50 percentage points in the United Kingdom, 55 percentage points in Portugal and more than 80 percentage points in Ireland.9

The Australian Federal Government is currently one of the strongest rated sovereigns in the world, with a AAA equivalent credit rating from all the major credit rating agencies. Significant deterioration in the fiscal position as a result of a financial crisis would be expected to threaten this rating. A reduction in Government’s credit rating is likely to lead to the banks’ credit ratings being downgraded, increasing funding costs.

As well as affecting the financial sector directly, a downgrade of the sovereign credit rating would raise Government’s borrowing costs and damage Australia’s reputation as a safe investment destination, ultimately harming the broader economy — including by raising borrowing costs for households and businesses.

Characteristics of Australia’s banking system create additional systemic risks

Historically, Australia’s growth has been assisted by the banks’ role as a conduit for foreign savings to fund domestic investment — a trend the Inquiry expects to continue. However, the benefits of offshore funding come with the risk that foreign investors will stop lending to an Australian bank.

The Inquiry recognises that Australian banks have built a reputation for prudent risk management, with low levels of proprietary trading and sound management. However, maintaining foreign investor confidence in the strength of the Australian banking system is paramount for maintaining the banks’ access to foreign funding. This goes beyond the strength of any individual bank, as Australia is a small part of the global financial system and investors may view Australian banks as a group. Many jurisdictions are still increasing capital levels to implement Basel III, a process largely complete in Australia. Over time, the relative strength of Australian ADI capital ratios may therefore decline as banks in other jurisdictions continue to increase capital.

Australia’s highly concentrated banking sector, at the core of its financial system, poses a further risk. The majority of Australian banks pursue similar business models, with broadly similar balance sheet compositions that can be expected to have a high correlation during a crisis. The major banks form part of the largest Australian financial groups and are highly interconnected with the financial sector. Hence, disruption to the functioning of one major bank could be expected to impose significant costs on the economy, particularly if it resulted in contagion to other Australian financial institutions.

Implicit guarantee

Actions taken by governments both in Australia and overseas to support their financial sectors during the GFC have reinforced perceptions of an implicit guarantee. Implicit guarantees arise when creditors believe that, if a bank were to fail, the government would step in to rescue the institution.

Implicit guarantees reduce banks’ funding costs by moving risk from private investors onto the Government balance sheet — a contingent liability for Government. As a result, the creditor takes no (or a reduced) loss, making it less risky to invest in the institution. Creditors will therefore accept a lower interest rate, which lowers funding costs for the bank and provides a competitive advantage to those institutions most affected.

Empirical studies have found that Australian ADIs, especially the largest ADIs, benefit from an implicit guarantee.10 This is also evident in the credit ratings of the major Australian banks, which all receive a two-notch credit rating uplift from credit rating agencies Standard & Poor’s and Moody’s due to expectations of Government support. Implicit guarantees create inefficiencies by:

  • Providing a funding cost advantage for banks over other corporations.
  • Giving large banks an advantage over smaller banks.
  • Weakening the market discipline provided by creditors.
  • Potentially creating moral hazard that encourages inefficiently high risk taking.11

Rationale

The Inquiry considers that these factors provide a compelling case for ensuring Australian ADIs have unquestionably strong capital ratios. The Inquiry’s judgement, based on the available evidence, is that the CET1 capital ratio of Australia’s major banks is currently not in the top quartile of internationally active banks, although it is likely to be above the median.12 Although this position does not suggest capital levels at Australian ADIs are weak, it also does not suggest they are unquestionably strong.

Perceptions of an implicit guarantee introduce a range of damaging distortions into the financial sector that reduce efficiency. They also transfer risk from the banking sector to taxpayers. In the Inquiry’s view, such factors make it appropriate to take steps to minimise implicit guarantees.

Raising capital requirements means that a larger share of bank funding would be in the form of equity — which is not perceived to have a guarantee — rather than debt. In addition, the perceived value of the guarantee for remaining debt would be lessened, as the ADI is safer and there is less chance the guarantee will be called upon. This reduces the implicit guarantee, in conjunction with Recommendation 3: Loss absorbing and recapitalisation capacity and Recommendation 5: Crisis management toolkit, which strengthen credible options to resolve an ADI with minimal recourse to public finds.

Options considered

  1. Recommended: Set capital standards such that Australian ADI capital ratios are unquestionably strong.
  2. Make no changes to capital ratio requirements.

Option costs and benefits

Summary of stakeholder submissions

Submissions from ADIs do not generally support increases to capital requirements — especially equity requirements — for a number of reasons.

They argue that increased capital, particularly equity, is unnecessary. As outlined in the Australian Bankers’ Association’s (ABA) second round submission (discussed later in this chapter), the Australian banks consider themselves to be highly capitalised relative to global peers. They argue that they are among the best capitalised banks in the world and are around or above the 75th percentile of capital ratios globally.13

The banks submit that their absolute (not only relative) capital position is very strong. Several banks note that internal stress tests show their capital position is sufficient to absorb large economic shocks, and APRA’s stress-testing has not led the regulator to raise capital requirements.14

Despite the banks’ submissions, the Inquiry notes that Standard & Poor’s classifies the major bank capital ratios as ‘adequate’ but not ‘strong’ or ‘very strong’.15

Several of the major banks point to the need to consider capital within the context of broader settings for financial stability in Australia. They argue that these broader settings and conditions make Australia a safe environment, reducing or negating the need for additional equity.16 Such conditions include conservative prudential regulation, which is stricter in a number of aspects than in other countries, and intensive and effective supervision from APRA.

The banks also point out that equity is their most expensive source of funding and that this cost will be passed on (at least in part) to consumers, ultimately slowing credit and GDP growth. They note that other forms of regulatory capital, such as Tier 2 capital, would provide protection from losses at a lower cost. The banks did not provide estimates of the extent to which competitive pressure would limit any rise in loan interest rates, but they did note that the sector is highly competitive.

Some smaller ADIs suggest that it would be appropriate to impose an additional capital requirement on those banks APRA designates as domestic systemically important banks (D-SIBs), to offset funding advantages from perceptions of an implicit guarantee. They argue that offsetting the funding cost advantage of the implicit guarantee would improve competitive neutrality in the banking sector.17

APRA considers Australian banks to be well capitalised, but acknowledges that overseas jurisdictions are continuing to increase capital requirements for their domestic banks. APRA’s preliminary view is that the major banks’ CET1 ratios are likely positioned broadly in the middle of the second highest quartile of internationally active banks, which is consistent with the Inquiry’s findings.

In its submission and subsequent discussions, APRA notes that stress-testing is a useful tool for assessing the riskiness of banks and their capital position. However, APRA cautions against relying on stress-testing too heavily to determine exact capital levels, given the margin for error in such exercises.18 It notes that many banks are still developing and improving their stress-test modelling as well as the critical data that underpins the models.

In addition, stress-testing exercises do not typically take into account more complex feedback loops and amplification mechanisms that can develop in practice. For example, banks are likely to respond to stress by cutting lending growth, which in turn may amplify stress and restrict economic recovery. Losses may also be more concentrated at one institution or a handful of institutions than can be assumed in the stress test. Even though a given institution may survive the average industry loss, it may be less resilient to a concentrated loss. As APRA notes in its most recent stress test, “… even though CET1 requirements were not breached, it is unlikely that Australia would have the fully-functioning banking system it would like in such an environment”.19

A number of analysts, think tanks and academics argue that, although equity funding may be expensive for the banks, increasing it does not impose large costs on the economy overall in terms of higher loan interest rates or lower GDP growth.20 They view greater use of equity funding as cheap insurance against the risks to which banking can expose depositors, Government, taxpayers and the broader economy.

Capital levels at the major banks

In the Inquiry’s judgement, capital levels at Australia’s major banks — as measured by CET1 capital — are likely to be above the global median but below the top quartile.

The Inquiry has not sought to determine the exact capital position of Australian banks on a consistent basis compared with banks in other countries. It is a very complex area, given the varied national discretions taken by different countries, including Australia. This is a task for APRA, taking into account the recommendations in this report. However, the Inquiry has sought to determine a plausible range for the current capital ratios of Australian banks for comparison with the current global distribution.

Based on the evidence available for the purposes of comparing with the global distribution, a plausible range for current Australian major bank CET1 capital ratios is 10.0–11.6 per cent (Figure 4: Adjusted average Australian major bank CET1 capital ratios). The lower bound is derived from the latest Basel Committee on Banking Supervision (BCBS) data, adjusted upwards by 0.8 percentage points to account for risk-weighted asset calculation differences based on APRA’s RCAP (Regulatory Consistency Assessment Program) data.21,22 The upper bound derives from a report submitted by the ABA which calculates capital ratios based on the Basel minimum requirements.23

Figure 4: Adjusted average Australian major bank CET1 capital ratios

Based on December 2013 global distribution

This chart shows the plausible range for current Australian major bank CET1 capital ratios against the global distribution.  The lower bound is 10.0 per cent derived from the latest BCBS data, adjusted upwards by to account for risk-weighted asset calculation differences and the upper bound is 11.6 per cent derived from the ABA report’s calculation of capital based on the Basel minimum requirements.  Other data points in the range are the capital ratio as reported by Australian rules 8.3 per cent; by BCBS 9.2 per cent; the global median 10.5 per cent; the Global 75th percentile 12.2 per cent and the ABA estimate against ‘international practice’ 12.7 per cent. 

View image enlarged

Sources: Australian Bankers’ Association 2014, Second round submission to the Financial System Inquiry, Appendix A: International comparability of capital ratios of Australia’s major banks; Australian Prudential Regulation Authority 2014, First round submission to the Financial System Inquiry, page 81; Basel Committee on Banking Supervision 2014, Basel III monitoring report September 2014, Bank for International Settlement, Basel; Inquiry calculations.

As Figure 4 shows, the available evidence suggests the major banks are not in the top quartile of CET1 capital ratios globally.

This view is supported by APRA’s assessment that the largest Australian banks are broadly in the middle of the second-top quartile of their peers for CET1 capital ratios.24

The Inquiry’s conclusion updates the observation in the Interim Report that Australia’s major banks were around the middle of the pack globally. That observation was based on data from the BCBS, which remains the most comprehensive data available for comparing capital levels across jurisdictions. However, although the BCBS data account for national differences in how capital is defined, they do not adjust for national differences in the way risk-weighted assets are calculated. Adjusting for this would move the Australian banks higher in the global distribution.

BCBS reported capital levels

Nonetheless, the BCBS provides the only available information about the distribution of global capital ratios, offering a useful context against which to compare Australian bank capital. The latest release, from December 2013, shows the CET1 capital global median and 75th percentile both increased to 10.5 per cent and 12.2 per cent respectively over the prior six months.25 In that time, the adjusted Australian major bank CET1 capital ratios reported by the BCBS increased by a lesser amount, to 9.2 per cent on average. This highlights that many countries are still ‘catching up’ with their implementation of Basel III relative to Australia and, in a number of cases, are introducing stricter requirements than exist locally. It can be expected that the global distribution of capital levels will therefore continue to rise for some time yet.

ABA reported capital levels

The ABA submitted a report that endeavoured to adjust Australia’s major bank capital ratios for differences between the Australian framework and estimates against the Basel framework and against international practice. Against the Basel framework, the ABA report assessed the average CET1 capital ratio across the major banks to be around 11.6 per cent as at August 2014. This approach considered similar items to those in APRA’s submission, although the ABA’s estimated value within categories was higher in a number of cases.26 In the Inquiry’s view, this estimate forms a plausible upper bound on the range of adjusted Australian major bank capital ratios.

Against its measure of international practice, the ABA report estimated a higher average CET1 capital ratio of 12.7 per cent. On this basis, it concluded that the major Australian banks were at or above the 75th percentile of identified international peers in terms of CET1 capital. Although this material was useful for considering the relative strength of Australian bank capital, its accuracy was limited by issues such as:

  • The restricted number of comparison countries and banks — it used 52 banks, compared to 102 in the BCBS report.
  • Minimal or no adjustment to foreign bank capital ratios to ensure they were on the same basis as the ABA-adjusted Australian bank capital ratios, which is crucial since the report directly compares these ratios.
  • Attempting to adjust some items to ‘international practice’ where credible benchmarks are not available, rather than to the minimums set out in the Basel framework, where benchmarks are clearer.

There is no benchmark of international practice: all jurisdictions have implemented the Basel framework in a different manner, reflecting their domestic circumstances. As a result, it is highly complex to compare even two jurisdictions, let alone to compare all jurisdictions. This is reflected in Recommendation 4: Transparent reporting, which recommends using the Basel framework since a broader benchmark does not exist. In the Inquiry’s view, the ABA’s adjustments that go beyond comparisons to minimums in the Basel framework are not a plausible basis for international comparison.

Benefits of higher capital

Higher capital provides insurance against the large losses that can be caused by financial crises through reducing the likelihood of such crises. It achieves this by making individual financial institutions safer and by promoting greater investor confidence in the system. It also reduces distortions caused by perceptions of an implicit Government guarantee. The benefits of increasing capital are not linear; however, the incremental benefit will decrease as the starting level of capital rises.

All financial crises are different, and the exact benefits of avoiding any particular crisis are therefore difficult to predict. Figures for the ‘average’ experience of a crisis should not be taken as precise estimates, but instead as indicative of the likely experience, noting that the actual magnitude can be much more severe.

Despite this limitation, a safer banking system that is less prone to crises provides large benefits, which accrue to individuals, the economy, Government and taxpayers.

Benefits to individuals

Financial crises are costly to individuals, both through the direct effects of financial institution failure and falls in asset prices, and as a result of the large recessions that typically accompany such crises.

Research on the ‘average’ financial crisis finds that the unemployment rate typically rises by around seven percentage points — over three times the increase in Australia during the GFC.27 The associated economic weakness lasts around four years on average, meaning that high unemployment can be protracted. The effect is often greatest for younger generations, particularly those trying to enter the workforce for the first time.

Financial crises can substantially reduce the savings of an entire generation. In relative terms, the largest effect tends to be concentrated on those people with lower initial levels of wealth. This can have a lasting effect on society, particularly on those who are in retirement, or about to retire, and who have limited capacity to rebuild lost savings.

Benefits to the economy

In its review of 21 empirical studies, the BCBS found the median estimate of the cost of a crisis in terms of cumulative foregone output due to the economic downturn is 63 per cent of one year’s GDP.28 The estimate range is 19–158 per cent of one year’s GDP, with the BCBS noting that the maximum cost of a crisis tends to be three to five times the average cost. Haldane estimated that the cost of the GFC, a particularly severe crisis, could be at least 90 per cent of 2009 world GDP.29 More recently, the Dallas Federal Reserve estimated the cost of the GFC to the United States economy at US$6–US$14 trillion (40–90 per cent of annual GDP).30

The BCBS also estimates that financial crises occur, on average, every 20–25 years in a given country, implying a 4–5 per cent chance of a financial crisis in any given year. However, across the world, crises occur somewhere with much greater frequency, and these crises typically have spill-over effects for other countries.

Combined with this estimated probability, the median cost of a financial crisis suggests an annual expected loss of around 2½–3 per cent of GDP. In dollar terms, based on 2013 nominal GDP, this translates to an expected cost to the Australian economy of $40–$50 billion per year. If this estimate was instead based on the top of the BCBS’s range for the cost of a crisis, this figure would rise to $100–$120 billion per year.

Given these large potential costs, even a small reduction in the probability or cost of a crisis would yield significant benefits.

The Inquiry notes that the estimated benefit of avoiding crises will tend to understate the true benefit to the Australian economy, since it does not account for:

  • Reduced perceptions of implicit guarantees. Weaker perceptions of an implicit guarantee reduce Government’s contingent liability and create fewer distortions to competition and efficiency in the financial system and broader economy. In addition, because ADIs are safer, any remaining perceptions of guarantee would be reduced.
  • An economy with fewer crises is less likely to be volatile, which has welfare benefits and promotes long-term trust and confidence to support investment in the economy. In contrast, volatility undermines long-term confidence and the ability of individuals, businesses and Government to plan for the future, impairing allocative and dynamic efficiency.
Benefits to Government and taxpayers

Reducing the likelihood of financial crises would protect Government and taxpayers from the costs of giving direct support to the financial sector. It would also help prevent the deterioration of the fiscal position due to the deep recession typically associated with financial crises.

The GFC clearly demonstrated the damage that can be done to governments’ fiscal position and the associated increase in net government debt. Chart 1 shows the change in general government net debt for a number of countries between 2007 and 2013, as a share of GDP. This captures the crisis period and the protracted recession that followed in many economies. In parts of Europe, the recession and associated fiscal costs continue more than six years after the crisis began.

Chart 1: Change in general government net debt, 2007–2013

Per cent of 2013 GDP

Chart 1 shows the change in general government net debt for a number of countries between 2007 and 2013, as a share of 2013 GDP. Germany has the lowest change (11.8%), followed by Canada (17.4%) then Australia (18.6%).  The highest in order are Ireland (80.8%), Portgual (53.4%), UK (48.6%) and US (41.9%).

Source: International Monetary Fund (IMF), World Economic Outlook Database October 2014, IMF, viewed 11 November 2014.

The GFC and the associated economic downturn left the Australian federal and state governments with notably higher net debt, which has yet to peak, despite Australia’s less acute experience of the GFC. The Inquiry understands there is limited room before Australia’s AAA credit rating is threatened. Estimates suggest this would occur as Commonwealth and state debt levels approached around a 30 per cent net debt level.31

Another financial crisis like the GFC could put Australia’s AAA credit rating in jeopardy, with likely knock-on effects for the credit ratings of Australian ADIs. This would make it more difficult for banks to access offshore funding markets, and would raise their funding costs.

Cost of higher capital

Overall, the expected cost of increasing capital requirements is small. The Inquiry estimates that a one percentage point increase in capital requirements would increase the average interest rate on a loan by less than 10 basis points.32 This is the figure if the full cost is passed on to consumers with no offset in interest rates by the RBA. However, in a competitive market, the actual change in lending interest rates would be lower and the RBA may lower the cash rate if conditions warrant. The Inquiry asked APRA to review its approach to generating these estimates, and APRA confirmed this approach was reasonable and consistent with other studies.

The Inquiry’s estimated effect on loan interest rates is roughly in the middle of the range found in a number of studies. The surveyed studies find increases in loan prices for a one percentage point increase in capital ratio are 1–22 basis points.33 The studies include:

  • APRA’s regulatory impact statement for the introduction of Basel III, which estimates a 5 basis points interest rate increase on a loan with a 50 per cent risk weight.34
  • A recent Bank for International Settlements (BIS) study on the impact of Basel III, which found a 12 basis points increase in loan prices per percentage point increase in capital, falling to around 8 basis points if only considering the advanced countries.35

This low cost reflects that changing capital requirements only affect a small portion of the funding of a loan. For example, a one percentage point rise in capital requirements affects the funding cost of less than 0.5 per cent of the average loan.36 That is, the funding cost on 99.5 per cent of the loan does not increase, and the incremental cost of equity over debt is only felt on the remaining 0.5 per cent.37 Changing the cost of this small slice of a loan’s funding therefore has a correspondingly small effect on the average funding cost.

RBA staff research suggests that an interest rate increase of this magnitude would reduce real GDP by less than 0.1 percentage points, while other studies suggest the effect could be even lower.38 In addition, the effect on growth would likely be taken into account in macro-economic policy settings since the RBA considers actual lending rates when determining the cash rate.39

The Inquiry’s estimate is consistent with a range of empirical studies that have estimated the effect of capital requirement changes on the economy (Table 2: Effect on GDP of a one percentage point rise in capital ratio). Studies examining the effect on GDP estimate that a one percentage point increase in capital ratios would potentially decrease annual GDP by 0.01–0.1 per cent ($150 million to $1.5 billion in terms of 2013 GDP) per year.

Table 2: Effect on GDP of a one percentage point rise in capital ratio*
Study Effect on GDP** Notes
Miles et al (2011) 1–5bps Effect on level of GDP
BIS (2010) 3bps Estimated lower growth during transition to higher capital. After implementation period, GDP recovers to trend.
BCBS (2010) 9bps Effect on level of GDP
Barrell et al (2009) 10bps Effect on level of GDP
Riksbank (2011) 6–16bps For low and high social cost of capital respectively
*Capital ratio measured as equity to risk-weighted assets.**Note that definitions of capital vary across studies.

Sources: Miles, D, Yang, J, Marcheggiano, G 2011, Optimal bank capital, MPC Unit Discussion Paper No. 31., Bank of England, London; Macroeconomic Assessment Group 2010, Assessing the macroeconomic impact of the transition to stronger capital and liquidity requirements, Bank for International Settlement, Basel; Basel Committee on Banking Supervision (BCBS) 2010, An assessment of the long-term economic impact of strong capital and liquidity requirements, Bank for International Settlement, Basel, page 25; Barrell, R, Davis, E, Fic, T, Holland, D, Kirby, S and Liadze, I 2009, Optimal regulation of bank capital and liquidity: how to calibrate new international standards, Financial Services Authority Occasional Paper 38, London; Sveriges Riksbank 2011, Appropriate capital ratio in major Swedish banks — an economic analysis, Sveriges Riksbank, Stockholm, page 31.

Reducing perceptions of an implicit Government guarantee reduces Government’s contingent liability. This benefit is not factored into the cost estimates above. The United Kingdom’s Independent Commission on Banking report estimated that around half the cost of its proposal to increase capital was offset by a reduction in the implicit guarantee.40

Box 7: The cost of raising capital requirements

To examine the potential effect on loan prices, this box provides a stylised example of a one percentage point increase in capital requirements. For simplicity, it does not account for a number of complications such as tax.

The effect on pricing is primarily driven by the proportion of funding that changes from debt to equity, the cost of the debt funding that is being replaced, and the cost of the new equity funding in terms of shareholder-required return on equity (ROE).

In Figure 5, a bank has a $100 portfolio of loans (the asset), which is funded by a mixture of debt and equity (the liabilities).

Figure 5: Example of raising capital requirements

To examine the potential effect on loan prices, this box provides a stylised example of a one percentage point increase in capital requirements on a simple balance sheet. The figures are all outlined in Box X.

View image enlarged

Originally, the bank has a capital ratio of 8 per cent, with an average risk weight of 50 per cent. The bank therefore:

  • Has $50 in risk-weighted assets ($100 x 50 per cent risk weight).
  • Uses $4 of equity funding ($50 x 8 per cent capital requirement) and $96 of debt funding.
  • Has a weighted average funding cost of 4.15 per cent given a cost of equity (target ROE) of 15 per cent, and an interest rate on debt funding of 3.7 per cent.

Increasing the capital ratio by one percentage point requires an additional $0.50 of equity funding ($100 x 1 percentage point capital increase x 50 per cent risk weight). The additional cost is the cost of the new equity less the cost of the debt it replaces (15 per cent — 3.7 per cent) x $0.50, or $0.06. To retain the same ROE, the bank charges an additional 6 basis points on the loan.

However, with greater equity, the bank would be safer so the risk premium built into both the cost of debt and investors’ required ROE should fall. In addition, competition may limit the extent to which the bank decides to increase prices for customers. These factors would reduce the increase in loan price.

This is an indicative example that illustrates how capital increases affect pricing. Although the identified price increase should not be interpreted as a precise change that would occur, it gives the Inquiry confidence that the change in loan pricing due to a one percentage point rise in capital ratios would be less than 10 basis points.

Conclusion

The Inquiry’s judgement is that, although Australian ADIs are generally well capitalised, strengthening the banking sector would deliver a net benefit to taxpayers and the broader economy. Evidence available to the Inquiry suggests the largest Australian banks are not currently in the top quartile of internationally active banks. Australian ADIs should therefore be required to have higher capital levels.

Unquestionably strong capital positions would deliver benefits by providing greater insurance against future financial crises and the associated harm to individuals, the economy, Government and taxpayers. Moreover, the cost of strong capital positions— the ‘insurance premium’ to reduce the risk of financial crises — is low. This cost would be reduced by competition in the market, including the effect of the recommendations in this report. The Inquiry estimates that a one percentage point increase in capital ratios would, absent the benefits of competition, increase lending interest rates by less than 10 basis points, which could reduce GDP by 0.01–0.1 per cent.

Although the benefits of higher capital are inherently difficult to quantify in a single number, to provide a net benefit to the economy, an additional percentage point of capital would only need to reduce the probability or severity of a crisis by 1 in 25 to 1 in 30.41

In addition, the RBA sets monetary policy, taking into account actual lending rates, and — to the extent that higher capital would affect GDP or inflation — can change the cash rate to at least partially offset the cost.42 Weighed against the risk of widespread unemployment, many households losing their savings, several years of economic recession and a large deterioration in the fiscal position, the Inquiry views this as a small cost.

The Inquiry recognises that the benefits of additional capital are likely to diminish the higher the starting level is. For example, moving from 2 per cent to 3 per cent capital is likely to have a larger effect on stability than going from 15 per cent to 16 per cent and, at some point, adding additional capital will not provide sufficient benefit to justify the added cost. However, in the Inquiry’s judgement, capital levels at Australian ADIs are below this point and there are clear benefits to additional capital.

Implementation considerations

Determining the appropriate level of capital to ensure Australian ADI capital ratios are unquestionably strong necessarily involves judgement. In the Inquiry’s view, if requirements are set such that ADI capital ratios are positioned in the top quartile of internationally active banks, this will achieve the goal of ensuring they are, and are perceived to be, unquestionably strong.

The optimal level of capital

A body of empirical work estimates the ‘optimal’ bank equity ratio for specific countries; that is, the level at which the net benefit to the economy is maximised. To the Inquiry’s knowledge, such a study has not been undertaken for Australia. Studies from other countries typically find the optimal level of equity capital ratios is 10–20 per cent of risk-weighted assets.43

Current minimum CET1 requirements for Australian banks, including CET1 buffers, are 8 per cent for D-SIBs and 7 per cent for others. Even after adjusting these to account for differences to the Basel framework — as outlined above — Australia’s requirements are at the lower end of the range of international estimates of the capital ratio that maximises net benefits to the economy.

Of course, each system is different and there is no guarantee that what is appropriate for another country will be right for Australia. However, considering Australia’s characteristics and circumstances, the ranges found in these studies support the idea that higher bank capital ratios would have a net benefit in Australia.

Further details

Unquestionably strong levels of capital would be beneficial for all ADIs. It may be argued that only the largest, most systemically important ADIs should be held to such a standard. However, in the Inquiry’s view, the failure of an ADI would have adverse consequences for its customers and the economy, and has the potential to undermine confidence and trust in the system. As such, the Inquiry judges that this standard should apply to all ADIs. In addition, holding different parts of the banking system to substantially different standards would introduce an unwelcome distortion to the competitive neutrality of regulatory settings.

The Inquiry recommends that increases in capital ratios from current levels should primarily take the form of increases in CET1, as the highest quality form of capital providing the greatest level of protection against a bank failing. However, APRA should use its discretion regarding whether part of such change should be through Tier 1 capital or total capital requirements. Appropriate transition periods should be used to limit the costs of transitioning to higher capital.

In implementing this requirement, the interaction between this recommendation and the effects of Recommendation 2: Narrow mortgage risk weight differences should be taken into account. In addition, the Inquiry notes a higher capital base for all ADIs may reduce the need for future changes to the D-SIB buffer.


6 Byres, W 2014, Seeking strength in adversity: lessons from APRA’s 2014 stress test on Australia’s largest banks, AB+F Randstad Leaders Lecture Series, 7 November, Sydney.

7 Reinhart, C and Rogoff, K 2009, This time is different: eight centuries of financial folly, Princeton University Press, Princeton, page 224.

8 Basel Committee on Banking Supervision 2010, An assessment of the long-term economic impact of stronger capital and liquidity requirements, Bank for International Settlements, Basel, page 10.

9 International Monetary Fund (IMF), World Economic Outlook Database October 2014, IMF, viewed 11 November 2014.

10 For example, International Monetary Fund (IMF) 2012, Australia: Financial System Stability Assessment, IMF Country Report No. 12/308, IMF, Washington, DC.

11 International Monetary Fund (IMF) 2014, Global Financial Stability Report April 2014, IMF, Washington, DC; Independent Commission on Banking 2011, Final Report, Independent Commission on Banking, London, page 101.

12 On a broader measure of capital, which includes CET1, Additional Tier 1 and Tier 2 capital, Australian major banks are ranked lower reflecting their proportionally greater use of CET1.

13 Australian Bankers’ Association 2014, Second round submission to the Financial System Inquiry, page 36.

14 For example, Westpac 2014, Second round submission to the Financial System Inquiry, page 71.

15 Standard & Poor’s 2014, The Top 100 Rated Banks: Will 2014 Mark A Turning Point in Capital Cushioning?, Standard & Poor’s.

16 ANZ 2014, Second round submission to the Financial System Inquiry, page 4.

17 Customer Owned Banking Association 2014, Second round submission to the Financial System Inquiry, page 15.

18 Australian Prudential Regulation Authority 2014, Second round submission to the Financial System Inquiry, page 50.

19 Byres, W 2014, Seeking strength in adversity: lessons from APRA’s 2014 stress test on Australia’s largest banks, AB+F Randstad Leaders Lecture Series, 7 November, Sydney.

20 For example, Admati, A and Hellwig, M 2013, The Bankers’ New Clothes, Princeton University Press, Princeton. At the extreme, some academics argue that changes in capital levels will not affect bank funding costs at all under certain conditions.

21 Australian major banks’ position is contained in a non-public BCBS report and was provided to the Inquiry by APRA. The adjustment for risk-weighting calculation differences are in Australian Prudential Regulation Authority 2014, First round submission to the Financial system Inquiry, page 81.

22 Using the estimates from the ABA report, excluding those related to capital definitions (which the BCBS data adjusts for), the lower bound on the range would be 10.5 per cent.

23 Australian Bankers’ Association (ABA) 2014, Second round submission to the Financial System Inquiry, Appendix A: International comparability of capital ratios of Australia’s major banks. The ABA report was commissioned from PricewaterhouseCoopers. It uses bank data for March and June 2014, while the BCBS global distribution is as at December 2013. Between December 2013 and March/June 2014 the major banks increased CET1 capital ratios by an average of around 0.5 percentage points. A stricter comparison of the major banks to the global distribution could take this into account and suggest an upper bound of only 11.1 per cent.

24 Byres, W 2014, Seeking strength in adversity: lessons from APRA’s 2014 stress test on Australia’s largest banks, AB+F Randstad Leaders Lecture Series, 7 November, Sydney.

25 Basel Committee on Banking Supervision 2014, Basel III Monitoring Report, September 2014, Bank for International Settlement, Basel.

26 APRA’s approach is outlined in APRA 2014, First round submission to the Financial System Inquiry, page 81.

27 Reinhart, C and Rogoff, K 2009, This time is different: eight centuries of financial folly, Princeton University Press, Princeton.

28 Basel Committee on Banking Supervision 2010, An assessment of the long-term economic impact of stronger capital and liquidity requirements, Bank for International Settlements, Basel, page 10.

29 Haldane, A 2010, The $100 billion dollar question, speech at the Institute of Regulation & Risk North Asia, 30 March, Hong Kong, Table 1, page 16.

30 Atkinson, T, Luttrell, D and Rosenblum, H 2013, How bad was it? The costs and consequences of the 2007–09 financial crisis, Federal Reserve Bank of Dallas Staff Papers No. 20, Federal Reserve Bank of Dallas, Dallas, page 1.

31 Standard & Poor’s 2014, Ratings on Australia affirmed at ‘AAA/A-1+’ on monetary and fiscal flexibility; outlook remains stable, media release, 29 July.

32 The precise quantum of additional capital necessary to place Australian ADIs in the top quartile of global peers is left to APRA to determine — the one percentage point increase here is for indicative purposes only.

33 See, for example, Barrell, R, Davis, E, Fic, T, Holland, D, Kirby, S and Liadze, I 2009, Optimal regulation of bank capital and liquidity: how to calibrate new international standards, Financial Services Authority Occasional Paper 38, London; Elliott, D 2009, Quantifying the effects on lending of increasing capital requirements, Center for Financial Stability; Kashyap, A, Hanson, S and Stein, J 2011, ‘A macroprudential approach to financial regulation’, Journal of Economic Perspectives, vol 25, no. 1; Miles, D, Yang, J, Marcheggiano, G 2011, Optimal bank capital, MPC Unit Discussion Paper No. 31., Bank of England, London.

34 Australian Prudential Regulation Authority (APRA) 2012, Implementing Basel III capital reforms in Australia, APRA, Sydney, page 15.

35 Cohen, B and Scatigna, M 2014, Bank and capital requirements: channels of adjustment, Bank for International Settlements, Basel, page 17.

36 The proportion of funding affected for a given loan is the change in capital requirement multiplied by the risk weight on that loan. The average risk weight of the major banks is currently less than 45 per cent.

37 Because higher capital makes the ADI safer, the funding cost of the 99.5 per cent of the loan may actually decrease to the extent that the risk premium demanded by debt and equity holders falls.

38 Lawson, J and Rees, D 2008, A sectoral model of the Australian economy, Reserve Bank of Australia Research Discussion Paper 2008-01, estimates that an unexpected 25 basis points increase in the cash rate reduces real GDP below its baseline by just more than 0.2 percentage points. A smaller estimate is provided in Jääskelä, J and Nimark, K 2008, A medium-scale open economy model of Australia, Reserve Bank of Australia Research Discussion Paper 2008-07 and Dungey, M and Pagan, A 2009, Extending a SVAR Model of the Australian Economy, Economic Record, vol. 85 no. 268.

39 For example, Battellino, R 2009, Some comments on bank funding, remarks to 22nd Australasian Finance and Banking Conference, 16 December, Sydney; Hansard 2009, Reference: Reserve Bank of Australia annual report 2008, House of Representatives Standing Committee on Economics, 20 February, Canberra.

40 Independent Commission on Banking 2011, Final Report, Independent Commission on Banking, London, page 141.

41 As the expected average effect of a crisis is 2½–3 per cent of GDP per year, to justify a cost of capital of 0.1 per cent of GDP would require a reduction of 1 in 25 (0.1/2.5) to 1 in 30 (0.1/3) in the probability or severity of the crisis.

42 For example, Battellino, R 2009, Some comments on bank funding, remarks to 22nd Australasian Finance and Banking Conference, 16 December, Sydney; Hansard 2009, Reference: Reserve Bank of Australia annual report 2008, House of Representatives Standing Committee on Economics, 20 February, Canberra.

43 See Miles, D, Yang, J, Marcheggiano, G 2011, Optimal bank capital, MPC Unit Discussion Paper No. 31., , Bank of England, London; Basel Committee on Banking Supervision 2010, An assessment of the long-term economic impact of strong capital and liquidity requirements, Bank for International Settlements, Basel; Barrell, R, Davis, E, Fic, T, Holland, D, Kirby, S and Liadze, I 2009, Optimal regulation of bank capital and liquidity: how to calibrate new international standards, Financial Services Authority Occasional Paper 38, London; Sveriges Riksbank 2011, Appropriate capital ratio in major Swedish banks— an economic analysis, Sveriges Riksbank, Stockholm; Independent Commission on Banking 2011, Final Report, Independent Commission on Banking, London.