Exhibit 11.1 Banking banana skins

Source: CSFI Banking Banana Skins, 2006.

It is not just the number of publications that has increased: Basel I (1998 updated version) totals 26 pages; Basel II (Final Framework, June 2006 version) runs to 347 pages. Taking these numbers as indicative, this represents a 150-fold increase in the volume of regulatory paper issued by BCBS from Basel I to Basel II.

Given the increase in globalization of banks since the mid-1980s, bankers also face the issue of keeping up with multiple regulatory bodies and the various pronouncements they make.

In the UK, this is manifested in the path that Basel II has taken: from the BCBS (347 pages), through the European Union’s Capital Requirements Directive (255 pages) to the FSA’s General and Banks & Investment Firms Prudential Sourcebooks12 (GENPRU and BIPRU) (868 pages).

Increasing Complexity

To give an idea of the change in complexity of requirements using the Basel I versus Basel II example, Basel I is based around a capital calculation that can be summarized as:

Capital Required = Σ Risk-Weighted Assets × 8 percent

where: risk-weighted assets = exposure13 × exposure class risk weight; and exposure class risk weight is looked up on the following exhibit.

Exhibit 11.2 Existing Basel 1 risk weights*

Exposure class	Risk weight
Claims on OECD central governments and central banks denominated in national currency and funded in that currency	0%
Claims on domestic public sector entities, excluding central government, and loans guaranteed by or collateralised by securities issued by such entities	0%, 10%, 20% or 50% at discretion
Claims on banks incorporated in the OECD and claims guaranteed by OECD incorporated banks	20%
Loans fully secured by mortgage on residential property that is or will be occupied by the borrower or that is rented	50%
Claims on the private sector Claims on banks incorporated outside the OECD with a residual maturity of over one year All other assets	100%

Source: BCBS International Convergence of Capital Measurement and Capital Standards, 1998, pp. 17–18.

* International convergence of capital measurement and capital standards, Basel Committee of the Bank for International Settlements, July 1988, updated to April 1998, pages 17–18 (summarized).

By contrast, the capital calculation for corporate borrowers at the heart of the internal ratings-based (IRB) approaches under Basel II can be represented as follows:

Capital Required = Σ Risk-Weighted Assets × 8 percent Risk-Weighted Assets = EAD × K × 12.5

where:14,15

and

b = (0.11852 – 0.05478 × ln (PD))2

N(·) stands for the standard normal cumulative distribution function

G(·) stands for the inverse standard normal cumulative distribution function.

This is the IRB formula for corporate customers under Basel II; there are four other categories of credit-risk customers (see the following Exhibit 11.3), each with its own different formula.

Exhibit 11.3 Risk weights for customer classes

Corporate	Corporate SME	Retail mortgage	Qualifying revolving retail	Retail other
As above	As above, different R factor	As above, no b factor, 0.15 R factor in all cases	As above, no b factor, 0.04 R factor in all cases. (portfolio must pass earnings volatility test before using this treatment)	As above, no b factor, different R factor
Basel 2: paragraph 272	paragraph 273	paragraph 328	paragraph 329	paragraph 330

Source: BCBS: Based II International Conference of Capital Measurement and Capital Standards: A Revised Framework, June 2006 (paragraphs as indicated).

In addition, for IRB banks, there are specific calculations required (or optional treatments available) for leasing, purchased receivables, equity exposures, specialized lending, counterparty credit risk (trading book) and securitization (see Exhibit 11.4).

Exhibit 11.4 Treatments for different products

Leasing	Purchased receivables	Equity exposures	Specialised lending	Counterparty credit risk (Trading book)	Securitisation
Must be treated as loan, with collateral. EAD equal to discounted lease payments. Residual value added, weighted by remaining term	Must calculate dilution risk, unless specifically permitted to disregard	May be exempted to use 100% / 150% risk weight if criteria met. Otherwise a simple approach with 300% / 400% risk weight applies, or a model must be created, with a 200% / 300% risk weight floor	Must use specific risk weights (the “slotting criteria”) unless specifically permitted to use general IRB formula above	May calculate exposure value using internal models (“expected positive exposure”) instead of existing MTM plus add-on approach	Must use external ratings where available, or infer an equivalent internal rating if possible. If not, supervisory formula approach is available, but it is highly complex and data-dependent.
Basel 2: paragraph 523. Residual value weighting added in the EU Capital Requirements Directive (CRD) Annex VII, Part 1, paragraph 27.	paragraph 369	paragraphs 340–358	paragraphs 275–279	paragraphs 777(xi), 777(xii)	paragraphs 538–643

Source: Ibid. (paragraphs as indicated).

Pushing Banks Toward a Precipice?

One of the major consequences of Basel II is that if a bank wishes to adopt the more advanced options available for credit risk assessment under Basel II, they are required to be able to model a number of ­quantitative ­measures of risk, specifically:

• Probability of default (PD);
• Loss-given default (LGD);
• Exposure at default (EAD); and
• Maturity (M).

Given the “use test”16 requirements under Basel II, there is then a requirement for banks to incorporate these models into the decision-­making processes of the bank, with the consequence being that the ­output of the models assumes a significant weight in the management of the firm.

Relying blindly on risk models to make decisions for a bank cannot be correct—the potential costs of type I and type II model errors17 are too significant to allow banks to leave credit decisions to models alone, particularly in the commercial and corporate banking markets.

Chorafas (2000) states “…analytical models are no substitute for sound judgement….”18 This is logically correct. Models can only react, in the prescribed way they have been built, to the inputs that they have been designed to assess.

By contrast, an experienced lender, who is well informed about the market, can make assessments that a model cannot, by virtue of the unconstrained nature of his or her thought processes.

For this reason, it is important that both banks and regulators recognize, respect and attribute proper weight to the value of expertise and experience in making credit decisions.

Regulators do recognize the risks of over-reliance on models. John Tiner, Chief Executive of the Financial Services Authority in the UK recently said:

…I am concerned about the realities of ever more complex financial instruments, accounting rules and prudential capital rules … There is no let-up in the stream of increasingly complex ­instruments emerging from firms. We should be in no doubt that this is generally a good thing—innovation feeds customer choice and competition. But as the ranks of quants and rocket scientists in financial firms continue to grow, we must not lose sight of the limitations of complex and sophisticated financial modeling techniques.19 (italics added for emphasis)

There is another risk in the drive toward risk modeling, which comes from the increasing tendency of regulators to look for consistency in risk models across banks.

The more regulators encourage consistency, the more they are in ­danger of building up a high level of systemic risk. This arises from the potential for “herd mentality” whereby models that are too similar may react in the same way to an unforeseen circumstance.

Regulators should certainly understand the differences between banks’ models and approaches, but they should be encouraging differences in bank strategy and approach, not eradicating them.

Much Ado About Nothing?

By the time Basel II reaches its final implementation date of January 1, 200820 (for advanced IRB in the European Union) it will be four years past its initial implementation deadline21 and will have taken a decade to develop and implement. In spite of this, in the run-up to the initial implementation date in Europe (January 1, 2007), a number of key areas remained unclear, including:

• The way “downturn”22 LGD and EAD figures are to be calculated, justified and implemented;
• The way that Pillar 2 will work in practice;
• “Home/host”23 regulator issues; and
• The impact and resolution of procyclical effects.
• These are all issues that create real concern within the banking ­industry, particularly in terms of how they impact on the concerns listed at the start of this chapter.

A further major concern in respect of Basel II relates to the costs of initial implementation and ongoing compliance. Pricewaterhouse­Coopers undertook a study on behalf of the European Union in 2004 where they estimated that it would cost Euro 20bn − Euro 30bn to implement the CRD across Europe.24

The implementation cost estimated by PwC is a staggering sum of money. One then considers:

• the impact of capital floors (in place for the first three years of Basel II at least);
• product level restrictions (such as the LGD floor for retail mortgages);
• the numerous and additive requirements for banks to ­demonstrate “conservatism” in their PDs;
• The requirement to use “downturn” LGDs and EADs; and
• the as-yet-unknown impact of Pillar 2 add-ons.

The accumulated impact of the aforementioned points calls into ­question the ability of banks to realize capital reductions as a result of Basel II. Banks have numerous stakeholders to satisfy, some of whom have their own view on the adequacy of capital resources (e.g., rating agencies). However, given the aforementioned, there is a question as to when quantifiable benefits from Basel II will find their way to the bottom line for banks.

Basel II Is Not Alone

The previous examples have focused on Basel II. It is important, however, to recognize that there are several other regulatory change programs that are currently underway or which have been completed in the last few years, including:

Sarbanes-Oxley;25

The adoption of IFRS standards;26

Anti-money-laundering regulations contained within, for example, the US Patriot Act;27 and

The Markets in Financial Instruments Directive (MiFID).

Each of the aforementioned represents a significant expenditure of time and resource on a regulatory, and therefore mandatory, program of work. Each is complex, and in some cases they have not been well coordinated between the relevant regulatory authorities.

It is the combination of increasing volume and complexity represented by the regulatory programs listed earlier (plus Basel II) that causes concerns among bankers.

This is well summed up by Hank Paulson, the US Treasury Secretary, who stated in one of his first speeches after taking office that “. . . the regulatory ‘pendulum’ had swung too far in response to the Enron and WorldCom corporate scandals . . .”28 and “the challenge before us now is how to achieve the right regulatory balance to enable us to be competitive in today’s world.”29

Mr Paulson’s wording about “the right regulatory balance” neatly encapsulates the basis of the rest of this chapter—how the development and implementation of regulations might be improved.

Improving the Process

There would seem to be three high-level principles that regulators and banks should follow to ensure that future regulatory initiatives are implemented as smoothly as possible.

Principle 1: Regulations Should Be Based on Principles Not Rules

Regulations should be driven by mitigation of clearly defined risks that impact the regulators’ key areas of responsibility: protecting depositors and the integrity of the financial system.

Too many of the discussions that have gone on in recent years in respect of the consultations and negotiations around Basel II have been to do with reviews of detail in lengthy documents rather than about substantive issues of how to reduce or manage the risks about which regulators are legitimately concerned.

Both regulators and banks must recognize that there are significant implications involved with moving toward principles-based regulation. To give just two examples, one from each “side”:

• Regulators will need to spend more time reviewing banks’ operations, policies and processes to ensure that principles are being adhered to—this means that the cost of regulation will increase, as regulators will need both more people and people with a different skill set;
• Banks will have to accept that it will be possible to face enforcement action for breach of a principle, rather than a rule.

For these reasons, among others, principles will not be the appropriate solution to all regulatory requirements. In particular, where the possible enforcement action includes the prospect of individual censure (fines or imprisonment) for senior management, clarity of boundaries will be required.

Principles-based regulation would require a change in mindset on the part of both parties and would necessitate more of a relationship approach between regulators and banks. This is reflected in the second principle.

Principle 2: Regulators and Bankers Need to Understand That We Are in This Together

It is normal for banks and regulators to have differing opinions on a range of issues—if they did not it would be a cause for concern. However, it is important that regulators trust that banks will not deliberately operate in an unsafe manner, and it is equally important the banks trust that ­regulators will not impose inappropriately restrictive rules on them.

It is in the best interest of banks to behave in a responsible manner, given the potential for significant reputational damage if they do not.

It is in the best interests of regulators to set and interpret regulations in a practical, pragmatic manner. This will help to avoid unintended ­consequences such as stifling innovation or restricting competition.

An example of a difference of opinion between regulators and banks to illustrate this point is a recent discussion on whether there is a difference between a bank’s support for a branch of the bank overseas, and support for an overseas subsidiary.

• Regulators maintain that branch status puts a different legal requirement on the bank to support a branch, as opposed to a subsidiary (and from a strictly legal perspective, the regulators are right).
• Banks say that the reputational impact of allowing any related entity to fail is so catastrophic that the niceties of legal structure are irrelevant to the bank when assessing whether to support an overseas part of the bank or not. From a practical perspective, the banks’ point is compelling.

The difference of opinion will remain. What is needed is for banks and regulators to find a way to agree on rules, and interpretations of rules, that do not drive banks, for example, to constitute their overseas ­operations in a particular legal form merely because they believe that their choice will result in a more or less stringent treatment.

Principle 3: Regulators and Banks Need to Communicate More Clearly

Regulators in general (and the FSA in particular) have been making real changes in the way they operate the process of communicating with industry in respect of changes arising out of Basel II.

It should be noted, however, that from the start the debate on Basel II has taken place at a technical level. This came about in part because it was technical people (e.g., portfolio managers and economic capital experts), rather than senior management with a more strategic focus, who led the debate on Basel II. This is true of both the banks and regulators.

In line with the first two principles, the communication process could be further improved by focusing the discussions on the risks that regulators are concerned about, coming up with principles (where that is appropriate) and sharing with banks, before regulation is drafted, what the risk is and why new regulation is required.

Banks understand the risks that regulators are concerned about. Closer, earlier dialogue may well enable new regulations to be avoided by agreeing to change practices.

What Is Next for Risk Regulation?

Looking even further forward, when looking at what is next for the regulation of credit risk, let’s start with a question: which is the odd one out?

Financial markets are: global

Financial institutions are: global

Basel II was designed to be: global

Regulators are: local

Over the last decade, the financial services landscape has evolved, becoming more globally integrated. In order for regulators to keep up with the banks, markets, and regulatory regime that Basel II represents, it seems logical that regulators must start to act and think globally. There is a real chance that they will fail in their duties to depositors and financial stability if they do not.

It is indisputable that acting globally raises issues for regulators. The most important of the issues is probably legislative—many regulators are required by law to act in the best interests of local depositors and for the protection of the local financial system.

This is a point that is at the heart of the “home/host” issue, and which affects another major issue with regulation: consistency. Both regulators and bankers, in their respective assessment of the goal of regulation, noted the need for a level playing field. This is currently being hampered by a lack of global consistency in the implementation of Basel II.

The 1.06 multiplier for IRB risk weights in the EU CRD is an example of this. The BCBS calculated a “scaling factor” of 1.06 following the decision to calibrate Basel II to unexpected losses only, and based on data from QIS3. Crucially, the BCBS has never explicitly changed the Basel formula to incorporate the multiplier. The EU did explicitly change its risk weight formula (see formula on p. 116), opening the way for EU banks to be required to hold more capital than non-EU banks, whose regulators adopt the formula as published by Basel.

In order to act globally, regulators may need to seek changes in their legislative responsibilities. This may not be popular in some jurisdictions, especially those where protectionist instincts are strongest; however, the increasingly global nature of the banks they are regulating makes it ­necessary for regulators to take a more global view in order that they fullfill their local obligations. If a regulator is unaware of an impending problem in another jurisdiction, the problem could end up catching them unawares at home.

Even without immediate change in legislation, regulators can make the most of the freedoms they do have, working within their existing legislative responsibilities, to develop, foster and increase cooperation with their peers in other regulatory organizations.

The EU, and in particular the Committee of European Banking Supervisors (CEBS), have been working on improving consistency and transparency of regulation within the EU. This is demonstrated by the inclusion of Article 12930 (requiring home, or “consolidating,” ­supervisors to manage the approval process for IRB and AMA approaches) in the Capital Requirements Directive and the publication of CEBS’ Consultation Paper 09.31

Notwithstanding the aforementioned, one senior regulator’s attitude has been related as: “we will listen to what (another regulator) has to say, but it is our decision and we reserve the right to require more or different documentation as evidence of compliance.” This is indicative of the issues in trying to get regulators to work together.

Basel III

Assuming that there will be a Basel III, what should it look like?

The three-pillar structure is a good framework for a regulatory regime and should be retained. To make a real advance on Basel II, the revised framework should be constructed as follows.

Pillar 1: Minimum Capital Determined by Banks’ Internal Risk Models

Many banks already model varying risk types at a fairly granular level. Where economic capital models, or their equivalent, are in use there is often a high degree of senior management support and understanding, which aligns well with the existing Basel II “use test.”

Pillar 2: Single Supervisory Review

A single, coordinated process, covering all relevant jurisdictions. The ­concept behind Pillar 2 (supervisory review of Pillar 1 calculations and compliance, along with a review of banks’ assessments of the role of ­capital in managing other risks) is sound. What would significantly improve it would be to make it a single assessment, binding on all supervisors who are involved with the bank. This does not, however, mean creating a single assessment that includes all of the varying requirements from every regulator.

Pillar 3: Enhancements to Disclosures

Enhancement in this case does not necessarily mean more disclosure—but getting greater harmonization of disclosures so that the market can make meaningful comparisons between banks (without compelling banks to disclose commercially sensitive information) would be a real improvement to Pillar 3.

Conclusion

Notwithstanding the issues regarding complexity and costs, Basel II has done much to improve communications between banks and regulators. The framework that Basel II puts in place allows for a more open and granular discussion between banks and their regulators. While there is still considerable room for improvement, it may be that this more open communication process around regulation and risk management will prove to be the most tangible benefit of the whole process.

Thanks

I would like to acknowledge the assistance and support I have received from colleagues past and present, in particular Colin Jennings (LloydsTSB) and Jonathan Gray (Royal Bank of Scotland), for their comments on the draft, and David Schraa (International Institute of Finance) for input on the IIF’s current work. As always, any errors or omissions are the responsibility of the author alone.

---