Unlike its moribund sibling Pillar 1, Pillar 3 of Basel II is alive and well and living in Australia.
In its implementation of Pillar 3 for Basel II, the Australian prudential regulator, APRA, has set 'minimum requirements for public disclosure' for advanced banks under its Prudential Standard - APS 300. Recently, the 'big four' Australian banks produced their APS 330 disclosures for the March 2009 quarter.
In reporting however, some banks have obviously concentrated on the word 'minimum' in the regulations, while others have focused on 'disclosure'. For example, ANZ produced a mere 7-page report, four pages of which were blank or pro-forma statements, and National Australia Bank (NAB) a 9-page report, three of which were fluff. On the other hand, Commonwealth Bank and Westpac both produced large reports (82 & 73 pages respectively) disclosing not only information on the firms' Risk Weighted Assets, upon which Minimum Regulatory Capital is to be calculated, but also quite detailed discussion on Risk Governance and Organization including sensitive topics such as 'risk appetite'. Congratulations to CommBank and Westpac! (see update below)
Now when comparing these disclosures, it must be remembered that here we are comparing apples against oranges (against cranberries and melons), since not all Risk Weighted Assets (RWA) are computed on the same basis. For example, Market Risk Assets are computed on the basis of a 99% confidence interval whereas Credit and Operational Risk are computed on the basis of a 99.9% confidence interval, or a 1 in 1,000 year loss event. However, since all of the banks are doing it wrong in the same way, we should be able to compare at least their numbers for Operational Risk.
Before looking at the computations for Operational Risk capital, it should be noted that Basel II requires that Risk Weighted Assets be multiplied by 8% to derive the required Minimum Regulatory Capital (MRC) and that Australian banks, in practice, maintain 'Total Capital' reserves at a higher 'capital ratio' that averages just over 11%. It should also be noted that the four major Australian banks (ANZ, NAB, Westpac and CommBank) are surprisingly (?) similar as regards their Total RWA, averaging some A$295 billion (with NAB highest and ANZ lowest). At 8%, this means that on average these banks must maintain some A$23 billion of regulatory capital - this is not loose change?
When considering only Operational Risk Weighted Assets the four banks are quite close, averaging some A$19.5 billion with ANZ (A$17.4 billion), NAB (A$23), Westpac (A$19.3) and Commbank (A$17.9). On average this means that these banks must maintain some A$1.5 billion of capital just to cover Operational Risks - certainly a not-insignificant 'cost of doing business'?
If we look at Operational RWA as a fraction of Total RWA, it averages some 6.6% for these banks.
But hold on a moment, under Basel II, isn't Operational Risk Capital supposed to be around 15% of Total Capital?
That after all was the basis that was used to set rules under the Basic and Standardised methods for calculating Operational Risk Capital for 'non-advanced' banks.
Have non-advanced banks been stiffed?
I believe so, they should be mad as hell, since historically it is the larger 'advanced' banks, such as NAB, that have incurred the losses that have brought down the wrath of regulators everywhere!
But that is a windmill that must remain un-tilted for the moment, there are even bigger questions raised by the Australian disclosures/ revelations on Operational Risk.
A close analysis of the detail of the published disclosures shows that while not large overall, at 6.6%, Operational Risk tends to be one of the largest risk categories reported.
For the major banks, Operational Risk tends to be the FOURTH largest risk category after Business, Mortgage and Specialised Lending. As reported, Operational Risk Weighted Assets are much greater than for example, Market Risk, Interest Rate Risk in the banking book, securitisation risk and even 'lending to banks' - has no one in Australia heard about the failure of Lehman yet?
If Operational Risk is so large, relatively, why are shareholders (and regulators for that matter) not jumping up and down demanding that overpaid bank executive get to grips with it and quickly!
As Australian banks typically estimate their 'cost of capital' at 12% this means that, even at 6.6%, Operational Risk Capital is costing an average A$188 million per year for each firm or about three quarters of a billion dollars in lost income/opportunity in the Australian banking sector.
Of course, we cannot completely eliminate Operational Risk, but surely, at these levels, bank executives should be being incentivised to reduce the overall cost to shareholders.
Operational Risk is the Heffalump in the regulatory room. No one really wants to talk about it since no one really knows what it is and no one knows how to go about reducing it.
In A.A. Milne's wonderful childrens story 'Winnie the Pooh', no one has ever seen a Heffalump, but everyone is sure that they exist because something/ someone is stealing the 'honey'. In real life, banks are losing money/honey as a result of not doing things properly but that does not mean that they have to create scary monsters, such as Operational Risk, to explain their inadequacies.
Operational Risk Capital is a benign conspiracy, regulators need a safety factor, or fluff, just in case they get it wrong and ORC is a nice candidate as no one knows what it is and banks are not going to scream too loudly since they know if it wasn't that, it would be something else.
But in a world where regulators are soon going to be multiplying risk-weighted assets by a new 'leverage ratio' wouldn't it make sense to multiple something meaningful by something sensible or we would just be going through the motions until the next crisis?
As Winnie the Pooh so eloquently put it, "People who don't think probably don't have brains; rather, they have gray fluff that's blown into their heads by mistake."
Update:
Some correspondents have noted that banks such as NAB have, in the past, disclosed more than the minimum requirement, but just not in this quarter. Likewise, others, such as Commonwealth, have been less forthcoming in the past. This is fully accepted, but it is a shame that good news is not always 'disclosed' in this context, if only with an Internet reference?
