Thank you for the invitation to speak here today. Bank capital regulation is on the one hand a fairly dry and technocratic subject, but on the other hand is crucial in maintaining a sound, efficient and competitive banking system. Sound and efficient banking systems are in their turn necessary to maintain a sound national economy.
Banking is unique among prudentially regulated industries in relying heavily upon global accords, rather than national rules. These accords are developed and voluntarily entered into by participating nations, with general coordination provided by the Basel Committee for Banking Supervision, which is associated with the Bank for International Settlements.
The first Basel Capital Accord became effective in 1988, after many years in development. This accord proved successful in lifting bank capital levels around the world, including Australia. Today the world’s internationally operating banks are as a group far sounder than was the case in 1988, and the first capital accord can claim some of the credit for this outcome. There is also a much more level playing field as regards regulatory capital requirements for internationally operating banks.
The world has moved on since the 1980s, and the first capital accord is becoming obsolete for sophisticated banks. From the early 1990s, such banks adopted economic equity models. These models estimate the equity required to limit the probability of financial failure for a given bank, using that bank’s actual risk position rather than the standardised assumptions used in regulatory capital models. Economic capital models have a long academic and applied history. Alfred Sloan, for example, was using the concept in the automobile industry as early as 1919. It is only in the past ten years, however, that economic capital or economic equity models have become widespread, and in fact many institutions still need to introduce or substantially upgrade their models to reach a reasonable state of the art. Over the past few years, both improved management practice and vastly improved data management technology has made effective economic capital modeling a reasonably achievable goal for most substantial financial institutions.
This brings me to the Basel Committee’s current work to bring out an updated capital agreement, the so-called Basel II accord. Basel II updates the first capital accord for banks and other deposit-taking institutions, but more importantly in a macroeconomic sense it allows sophisticated banks to use elements of their economic equity models as inputs into their regulatory capital requirements. As we will discuss, the world’s banking system will become both safer and more capital efficient, so Basel II is clearly worth the substantial effort flowing into it.
I note that the final Basel II rules and implementation timing are not yet fixed at this point, so my comments today are necessarily subject to revision over time.
This table demonstrates that Australia possesses over 200 Authorised Deposit-Taking Institutions, or ADIs. The industry is concentrated, however, with the four largest competitors holding two thirds of the assets and the banking sector holding 96% of the assets. Australia’s largest banks are very sound in world terms, all currently possessing at least AA minus debt ratings.
In the Basel II framework, large and sophisticated banks are expected to input their economic equity models into the so-called Internal Ratings Based or IRB models, to determine their regulatory capital. Smaller deposit takers, which in the Australian context includes all the credit unions and building societies, are likely to continue relying on standardised regulatory capital models. The middle sized institutions may go one way or the other, depending on their business model and risk management sophistication.
As this chart shows, the great majority of the assets and deposits subject to Basel II supervision will be supervised using IRB capital models, but the great majority of supervised institutions will be subject to the standardised approach.
Regulatory capital rules are important because ADIs maintaining a sound capital position are most unlikely to fail, and in fact should remain strong enough to perform their important credit allocation function even in challenging economic environments. ADI strength is not only important because so many people have substantial funds deposited with them, but because ADI ability and willingness to lend is a substantial input into the commercial, construction, and household sectors’ economic health. For the banking system in particular, maintaining a sound payments environment is critical in a modern economy.
In balancing APRA’s statutory safety obligations with our statutory efficiency, competition, and contestability obligations, safety is the primary consideration. For ADIs, the most important objective in setting capital rules is to ensure sufficiency. APRA’s position is far from unique on this matter; the general international approach also is based upon sufficiency.
As with all things, sufficiency in capital must be balanced by other considerations. For Basel II IRB banks, increased risk sensitivity is an important improvement. To take the main example, under the current capital accord all corporate loans carry the same regulatory capital requirement. Under Basel II, low risk loans will attract a far lower capital requirement than high risk loans.
Excessive capital requirements are more comfortable for a regulator than insufficient requirements, but they carry large economic costs. Tying up too much of the national capital stock in bank equity positions means that this money is not available to take risks and earn returns in other sectors of the economy.
Less obviously but also importantly, excess capital can lead to complacent risk management practices, and no capital position can long withstand poor risk management in a bank.
Finally, Basel II is unavoidably complicated and will require substantial conversion costs. APRA is striving to minimise these costs where possible. In addition to explicit costs, Basel II will absorb a great deal of management time and attention, which might detract from attention to more fundamental risk management issues.
Let’s now look at how Basel II will work for large banks relying on their internal risk estimates.
In such estimates, the bank will separately calculate the capital required for credit, operational, and all other risks. In each case the bank’s internal estimates are the starting point, but there is considerable scope for supervisory discretion. A bank is not allowed to use its internal capital estimates unless it receives prior approval from the supervisor, APRA in the Australian case. It would be fair to say that discussions have already started with prospective IRB banks, and they are likely to continue up to and past the conversion to Basel II.
In addition to adopting in-house risk estimates for regulatory purposes, Basel II will require a great deal more disclosure on risks and risk management from ADIs. This should in turn lead to more market discipline on banks who are straying from a sound risk/reward balance.
On our current calculations, the likely outcome for banks on the IRB approach will be a modestly lower total capital requirement, but higher marginal capital applied to specific risky behaviour. This represents the best of both worlds in efficiency and safety terms. The improved efficiency and safety outcome for IRB banks is the main justification for the international effort flowing into Basel II.
Basel II is also an important issue for smaller deposit takers including credit unions and building societies. The standardised approach applying to such entities is an update on the current capital accord. It is simplistic but workable, and among other things commences the process of introducing operational and other non-credit risks to the capital calculation.
As with IRB banks, Basel II involves increased disclosure for standardised model users, and this disclosure is likely to produce positive prudential effects.
With standardised approach ADIs, APRA expects to see on average a moderately lower capital requirement than currently applies. We are comfortable with this outcome.
APRA is striving to minimise the transition costs for smaller ADIs in Basel II. We are reasonably confident that the new rules applying to smaller ADIs will not prove unduly costly in a transition sense.
Basel II is clearly an improvement on the current capital accord. Most importantly, for IRB model users the new accord links risks and capital requirements much more closely than is currently the case. This link extends to personal incentives for senior managers. Banks using economic equity models, or at least using them properly, explicitly tie the resultant capital costs into line unit profit and loss statements, and therefore to line manager bonuses and promotion prospects.
By linking risk to capital requirements, Basel II discourages imprudent business strategies. I am thinking here, for example, of the late 1980s rush into lending for speculative property in Australia. Under Basel II no IRB bank could have tolerated the very high capital requirements associated with such lending.
By better aligning regulatory capital requirements with economic risk, Basel II makes the Australian deposit taking sector, and therefore the Australian economy, safer and more efficient. With a better capital model we are both more assured of sufficiency, and less reliant upon rules of thumb that might require excessive capital. Not to put too fine a point upon it, Australians will be beneficiaries over time when Basel II is adopted in this country.
Unfortunately, Basel II is not an unalloyed good. Most importantly, we are exposed to model risk on each IRB bank. Model risk occurs when the assumptions, data, and calculations used in the model fail to match reality. We know, without any doubt, that the assumptions and data used in these models are imperfect. Our strong expectation however is that these imperfections will be less than is the case using standardised rules of thumb imposed by international conventions.
It will be important for APRA and the banks never to succumb to model myopia, believing that models drive reality, rather than the other way round.
We also know that Basel II modeling is far stronger for credit, trading, and interest rate risk than it is for operational, strategic, business, or cyclicality risks. The former group can be modeled using well established statistical and mathematical tools. The latter group has no such mathematics, or at least none that can be confidently used by bank regulators or the banks themselves.
The risks that are mathematically inconvenient also turn out to be difficult for regulators in developing safe but fair rules. It is clear that deposit takers need substantial capital to protect against, for example, adversity in the business cycle, in strategic investments, and interest rate movements. It is far less clear how we should calculate these capital requirements.
Under the current capital accord, supervisors can calculate a bank’s required capital position using year 8 maths. Under the new accord, this will remain the case for ADIs using the standardised model, but IRB calculations are founded upon stochastic calculus and other similarly advanced methods. APRA’s supervisors will need the skills and experience to keep up with IRB capital models. Unfortunately for the APRA budget such resources do not come cheaply.
Finally, any change creates cost and distraction, and Basel II is a big change. As already noted APRA will strive to minimise unnecessary costs, but our idea of necessary will in some instances differ from the ADI industry’s view.
There are material competition issues associated with Basel II. The first consideration is Australia’s stance as a competitive financial centre. The great majority of countries can effectively supervise to the current capital standard. Only a relative handful of countries, perhaps 15 to 20, can support the IRB capital model, and Australia is one of these countries. The IRB standard is expected to generate both safety and efficiency benefits, so Australia’s ability to support this standard will serve us well in maintaining an internationally competitive finance sector.
The main domestic competition issue is the split between IRB banks and standardised approach ADIs, including smaller banks. We recognise the potential for competitive disequilibrium between IRB and standardised approach users, particularly in home loans. Doubtless this will be a matter for considerable industry discussion and possibly some angst, but our calculations indicate that the larger bank’s current capital advantages will not widen materially as a result of Basel II’s introduction.
We need to ensure that Australia’s ADIs are not unduly disadvantaged in their offshore operations, or in their access to offshore capital markets and trading markets. Adopting Basel II will ensure that Australian ADIs remain in the global top tier for capital regulation.
There are a great many outstanding Basel II issues that will require attention before the new system is launched. Among other things, there has been uncertainty at to launch date for the new system.
In addition to sorting out the international rules, APRA will need to determine the specific rules applicable to Australian ADIs. Under Basel II there are over 40 items of national discretion that require resolution.
APRA is likely to take a tougher stance than the international average on many of the national discretion items, particularly for IRB banks. This continues a well established position inherited from our predecessor bank regulators at the RBA, which emphasises the importance of a strong deposit taking sector. It is no accident that our largest banks carry double A ratings, where a single A rating might be the median in other countries. Over the past decade, there have been few if any indicators, including profitability, shareholder return, and cost to income ratios, that maintaining strong deposit taking institutions has required a tradeoff against efficiency.
In summary, APRA strongly supports Basel II. We are confident that the Australian financial system will become safer and more efficient as a result of this innovation.
Although the implementation task is considerable, for both the ADIs and for APRA, as a group we are reasonably well prepared for the effort.
Basel II will not be cheap, easy, or risk free, but in the long run it will be good for the Australian deposit taking sector, and good for the Australian economy.
Thank you for your attention.