Part VI Bank Group Supervision, 25 Financial Conglomerates
- Regulation of banks — Credit risk — Prudential regulation — Supervision
25.01 The Joint Forum on Financial Conglomerates released in 1999 a paper which sets out the basis of the supervision of conglomerates,1 and the basic principles set out in this paper remain the basis for conglomerate supervision. They are that supervisors should seek to assess the capital adequacy of financial conglomerates. In so doing, measurement techniques should be designed to:
III. include a mechanism to detect and provide for the effects of double, multiple, or excessive gearing through unregulated intermediate holding companies which have participations in dependants or affiliates engaged in financial activities;
(p. 480) IV. include a mechanism to address the risks being accepted by unregulated entities within a financial conglomerate that are carrying out activities similar to the activities of entities regulated for solvency purposes (for example leasing, factoring, reinsurance);
V. address the issue of participations in regulated dependants (and in unregulated dependants covered by principle IV) and ensure the treatment of minority and majority interests is prudentially sound.
Each of these merits a word of explanation.
Double or multiple gearing
25.02 Double gearing occurs when the same capital is counted twice. It is worth noting that the simplest form of double gearing would occur if banks were permitted to invest in each other’s equity without restriction—thus, bank A could raise 100 of capital and invest it in the shares of bank B, on terms that bank B would re-invest in new shares of bank A. Within the banking system this problem is addressed by requiring banks to deduct equity holdings in other banks—thus if bank A raises 100 of new equity and invests that money in the capital of bank B, it will be required to deduct the holding in bank B from its total capital, and its position remains unchanged. Within a group, however, the position may be more complex—if a parent company raises equity and invests it in the shares of an insurance subsidiary, and that insurance subsidiary then acquires shares in a bank affiliate, then the same capital may be counted twice: once within the bank and once within the insurer. The point here is that it is only externally generated capital which can support the group as a whole, and to the extent that concerns arise as to the stability of the group, then it is the group’s access to capital which will be relevant to external counterparties. The concern here is therefore that the capital position of the regulated entities within the group may be being artificially inflated at the expense of other group members.
Debt downstreamed as equity
25.03 It may not be immediately obvious to the observer why the bank regulator should care about the parent’s capital levels. The answer may be seen from Figure 25.1.
The concern here is that investors are subscribing for what appears to be senior debt, whereas what is in fact being raised is subordinated bank capital. The reason that regulators find this concerning is that the group itself may be increasing leverage and therefore reducing its financial stability. In addition, since the group as a whole is obliged to meet the obligations to the senior noteholders, and the group management will necessarily have significant influence over the board of the bank, the bank may in practice be obliged to treat the equity which it has (p. 481) received as being in effect on the terms of the senior debt raised. Ultimately this results in the suggestion that the equity is not ‘true’ equity, and that the bank’s position is weaker than it appears.
Unregulated intermediate holding companies
25.04 An unregulated intermediate holding company may operate to break the chain of deductions. Thus, if a bank invests in another bank, the investment will be deducted from capital. However, if a bank invests in an unregulated holding company, that holding may not necessarily be deducted, particularly if it is a small holding. Thus intermediate unregulated holding companies may operate—deliberately or accidentally—to avoid the double gearing rules.
Unregulated entities engaged in financial business
25.05 Where risk has been transferred by a regulated group member to an unregulated member of the same group, the ordinary rules of risk transfer do not apply— the relevant risk is in many ways retained by the bank. This principle may be extended to the conduct of other activities. There are many activities—invoice discounting, factoring, holding assets—which may be engaged in by a bank as part of its ordinary activities, but which do not absolutely require a banking licence and may therefore equally well be engaged in by unregulated entities. As we have seen (paras 24.10 to 24.17), when such entities are subsidiaries of a bank they are required to be included in the consolidation as financial entities rather than excluded from it as commercial entities. However, within a group such activities may be conducted by entities which have no direct linkage with the bank beyond a common ultimate parent.
25.06 Where a group has neither control nor influence over a particular entity, that entity should not be included in the supervised group. It is a generally established rule of thumb that the cut-off point for a holding which confers neither control nor influence is a holding of less than 20 per cent2 in the voting equity of the relevant entity, although this presumption can be displaced if there is evidence that the holding does confer effective influence. Difficulties, however, arise where a group has a holding in an entity of between 20 per cent and 50 per cent. On ordinary corporate principles this means that the relevant entity is not a subsidiary of the group, but the holding is sufficiently large to ensure that the relevant entity can be used to further the interests of the group as a whole and, more importantly, may be perceived by the markets as sufficiently closely connected to the group that its failure would damage the creditworthiness of the group. If this is the case, then the entity will in practice have a claim on the financial resources of the group, and in assessing the stability of the group it must be taken into account.
25.07 However, an entity in which the group has a minority stake is clearly in a different position than an entity which the group wholly supports, and the position of the remaining shareholders cannot be completely disregarded in determining the appropriate regulatory approach. As a result, regulators in general deal with participations by including a pro rata share of the assets and liabilities of the regulated entity in the group supervision. As an exception to this approach, where the relevant entity has a shortfall in its regulatory capital position, the group will generally be treated as if it were solely liable to make good that shortfall. This reflects the fact that other shareholders are assumed to be more likely to refuse to support an undercapitalized group any further, whereas a regulated investor may be under a greater degree of moral pressure (including, of course, pressure from regulators) to stop the relevant entity failing.
25.08 There has been much discussion of the regulation of groups which combine financial and non-financial business, beginning with the question of whether they are desirable at all. Many jurisdictions have, or have had at some point, legislation prohibiting banks, insurance companies, and securities firms from acquiring or being acquired by each other, and this is frequently defended as being in order to reduce systemic risks.
(p. 483) 25.09 The reality is—of course—more complicated. It is generally accepted that the diversification of risk is broadly beneficial on systemic stability grounds, and there is no obvious reason why diversification between banking and insurance should not be at least as effective in this regard as diversification within different aspects of the business of banking—indeed it can plausibly be argued that a commercial bank which is grouped with an insurer is better diversified than a commercial bank which is grouped with a retail lender. However, you can take this one stage further and argue that a bank which is grouped with (say) an aerospace manufacturer is better diversified still, and it is at this point that bank and other financial supervisors become—justifiably—nervous.
25.10 There are some cases where financial groups exist within larger non-financial conglomerates. These may arise either because the non-financial conglomerate has expanded into financial business as part of its general business activities, or because a deliberate decision has been taken to expand into finance as a new business sector. Both of these are entirely legitimate business strategies, and regulators have in broad terms concluded that there is no policy reason for prohibiting non-financial corporate groups from owning financial businesses.
25.11 This does, however, give rise to a concern regarding the pressures to which a group financial firm (and in particular a group bank) may be subject. The key issue is that a bank is generally leveraged to a degree which would be unacceptable in an ordinary commercial company, and in some cases the management of a commercial company may seek to own a bank primarily for the purposes of creating an increase in group leverage. This increase may destabilize the group, and this in turn may destabilize the bank. Technically (as we have seen) a bank regulator has the power to regulate any entity which is the parent of a bank, regardless of the nature of that parent’s activities. However, taken to extremes this would result in (say) the UK Prudential Regulation Authority (PRA) seeking to exercise prudential capital supervision over (say) a Japanese car manufacturer merely because it happened to own a small bank in the UK. Such supervision would be redundant even if it were possible.
25.12 As a result, the regulatory system is generally constructed (in the EU at least) such that the regulator supervises a subgroup within the main group. That subgroup is defined as the subgroup where not less than 40 per cent of the balance sheet value of the consolidated subgroup is composed of the balance sheets of any bank, insurance company, or investment firm. Capital adequacy supervision will be imposed on that subgroup as if it were an independent group, and it will be subjected to requirements on risk concentration, intra-group limits, and internal control and risk management processes as if it were a free-standing group. This can frequently give rise to problems for such entities, since it will generally result in the financial group within a conglomerate being managed in a way which is very different from the management philosophy which applies across the (p. 484) conglomerate as a whole. However, regulators tend to be particularly concerned with such groups, and in reality the regulator will wish to be reassured that the degree of independence of the management of the financial group within the conglomerate is sufficient to protect depositors and counterparties of the financial business of that conglomerate.
25.13 In addition to the issues raised where a non-financial entity owns a financial entity, a further stage of complexity arises where different types of financial business are combined within the same group. The European approach to this is to permit such combinations but to require group supervision to be applied across them. This means that mixed activity groups containing insurance, banking, and investment banking activities along with—generally—other types of regulated financial activities are permitted.
25.14 Some specific issues arise with respect to such groups as a result of their inclusion of insurance companies. Regulators have spent many years grappling with the issues which arise where securities trading business is done within banks, and these two disciplines are now sufficiently interrelated that the preparation of consolidated accounts for a group which contains both banking and investment firm subsidiaries presents no very great difficulties, and such consolidation can (in theory) be accomplished on a ‘line by line’ basis. In general, however, an insurance company is prohibited from engaging in any business other than insurance business and ancillary activities. Consequently regulators have not generally had to deal with the issues involved in combining insurance and banking activities within the same entity and in practice supervision of these groups relied for many years on a firm-by-firm approach based on the regulator satisfying itself that the individual firms concerned had sufficient capital as measured by their appropriate standards.
25.15 There is, however, an interesting discontinuity in the regulation of the two firm types, in that a bank which owns an insurance company will (in general) be required to deduct its holding in the insurance company from capital on a solo basis. This means that it is effectively impossible for a bank which is at the head of a bank group to own a large insurance company directly. This problem can, however, be solved through the use of an unregulated holding company which holds the shares of both the bank and the insurer. Conversely, an insurance company which is the parent of a bank is not required to deduct the value of its stake in the bank from its capital, and structures where an insurance company is at the head of a mixed activity group are more common.
25.16 The vast majority of jurisdictions (and the EU) regulate such entities in different ways under different rules and through different supervisors, and the preparation (p. 485) of consolidated regulatory accounts is broadly outside the scope of any directly applicable piece of regulation. The EU has sought to address this through the financial groups directive regime, which amended the banking, investment, and insurance regulatory regimes at the EU level in such a way as to provide in each of them for interlocking components designed to require consolidated reports at a group level reflecting the different provisions.
25.17 The basis of the supervision of financial conglomerates is broadly the same as the basis for the supervision of banking groups. The parent company of a conglomerate is very likely to be unregulated and outside the scope of the powers of the relevant regulator. Consequently, group capital requirements are imposed by requiring the regulated bank members of the group to ensure that the group as a whole maintains sufficient capital (calculated applying the basis which would apply to the relevant bank) to pass the relevant capital adequacy test if it were applied to it. This means in practice that if the group as a whole has insufficient capital, the regulated entity is itself required to have sufficient capital to make up the shortfall. Thus for a bank which is a member of a financial conglomerate, the practical effect of these rules is that it is required to maintain a level of capital which is the higher of:
If the group does not satisfy this requirement then the bank’s licence may be withdrawn, or administrative limitations may be placed on its ability to conduct regulated business until the position is regularized.
25.18 The basis of this approach is the rule of recognition for financial conglomerates. For these purposes, any regulator may commence the process of determining whether a particular group is a financial conglomerate, or in some cases a member of the group may require the relevant authority to make that determination. The competent authority that would be co-ordinator will take the lead in establishing whether a group is a financial conglomerate once the process has been started. This process will normally involve discussions between the financial conglomerate and the competent authorities concerned.
• the group contains at least one bank, investment firm, or insurance company (a ‘regulated entity’);
• either the group is headed by a regulated entity, or regulated entities comprise more than 40 per cent of the balance sheet total of the group as a whole;
(p. 486) • the group contains both an insurance company and at least one of a bank or an investment firm, and the balance sheet totals of each should be more than 10 per cent of the balance sheet total of the financial sector entities in the group in total, or EUR 6bn.
Any subgroup of a group which satisfies these criteria will itself be a financial conglomerate. As a result, a very large group may contain more than one financial conglomerate as well as possibly being itself a financial conglomerate.
25.20 Because the definition of a financial conglomerate is highly flexible, there is often scope for debate about whether a group is a financial conglomerate and whether it has, by reason of an acquisition, become one or not. As a result, the rules on financial conglomerates do not apply as from the moment when a group actually becomes a conglomerate, but from the point at which a relevant regulator serves a notice that it considers the group to be a conglomerate. This raises the theoretical (and in some cases the actual) risk that of two identical groups, one may be treated as a conglomerate and the other not.
25.21 Once a financial conglomerate has become a financial conglomerate and subject to supervision in accordance with the relevant directive, since we are only here concerned with groups which contain banks this means that the relevant group will be supervised in accordance with the rules for bank consolidated supervision. It will only cease to be subject to this regime for this purpose if the total financial services sector activity in the group drops below 35 per cent of balance sheet total, or if the size of the smallest sector in the conglomerate drops below 8 per cent of balance sheet total or EUR 5bn. However, once a conglomerate has dropped below the original 40 per cent and 10 per cent levels, it will cease to be a financial conglomerate in any event three years after that.
25.22 For this purpose the comparison of balance sheets can produce some unusual results—for example, an insurance company will generally carry assets under management on its own balance sheet, whereas a bank will generally hold such assets in unconsolidated funds. It is therefore important to apply the percentage intelligently and in consultation with the regulator, since the analysis is likely to require judgement as well as simple mathematics. Further, arrangements of this kind by definition must confer a remarkably wide degree of discretion on national regulators to flex the terms of the decision as to which groups will be caught by it.
25.23 The Joint Forum paper3 classified the existing approaches to conglomerate regulation into the ‘building block prudential method’, the ‘risk-based aggregation (p. 487) method’, and the ‘risk-based deduction method’. The building block prudential method essentially compares the fully consolidated capital of the conglomerate (as derived from its published financial accounts) to the sum of the regulatory capital requirements for each group member. The risk-based aggregation method sums the solo capital requirements for each regulated firm, and compares it with the capital of the group’s parent. The risk-based deduction method takes the balance sheet of each company within the group and looks through to the net assets of each related company, making use of unconsolidated regulatory data. Under this method, the book value of each participation in a dependant company is replaced in the participating company’s balance sheet by the relevant share of the dependant’s capital surplus or deficit, and the adequacy of the capital of the group is measured against this yardstick.
25.24 The fact that the Joint Forum was unable to recommend a single method reflects the fact that the effectiveness of these methods will vary according to the structure of the conglomerate. In particular, accounting requirements are unlikely to work well with groups headed by insurance companies, since such entities did not historically mark their investments in subsidiaries to market on a regular basis. Equally, the solvency requirement imposed on an insurance subsidiary may well be significantly less than the carrying value of the stake, but this will not reflect the double-counting involved in an insurer investing in the capital of other group entities.
25.26 This method calculates capital adequacy using accounting consolidation and is generally applied to bank groups. It operates by calculating the capital requirements which apply to each individual entity within the group and comparing them with accounting capital, and is broadly comparable to the building block prudential method. Since it is a good bet that each individual entity is compliant with its own minimum capital requirement (if it were not it would probably have been closed down by the relevant regulator), it may seem that the conglomerate capital resources requirement will necessarily always be satisfied, since if each entity has capital which exceeds its applicable requirement, then the total of the capital of the group as a whole must necessarily exceed the total of the capital requirements of the group as a whole. This would in fact be true if no group member had an equity interest in any other group member. However, in a group, by definition, at least one group member must have an equity interest in at least one other, and more commonly all (or almost all) of the capital of the subsidiaries in the group will be owned by the parent company of the group. There (p. 488) will therefore be some double-counting of capital within the group and the requirement in practice is that the group as a whole has sufficient capital, excluding intra-group investment, to support its balance sheet. For this purpose the capital of the group will constitute the capital of the parent company of the group plus any outside capital raised by any other member of the group.
25.27 The financial resources of a financial conglomerate are broadly calculated in accordance with the sectoral requirements which apply. However, where a conglomerate includes both insurance and non-insurance business (or in some cases investment businesses), the eligibility rules may be different. In particular, the gearing rules (which prescribe the proportion of capital which may be composed of any given type of capital) vary within business types. The issue is therefore whether capital which would be ineligible for a bank may count towards the group capital requirement. The rule here is that once the group capital requirement has been determined, it is then divided up into percentages based on the balance sheet size of the individual components, and the capital rules which apply to the appropriate component are used in respect of it. Thus, if 30 per cent of a conglomerate (measured by balance sheet size) is in the insurance sector, the insurance sector rules may be used in order to determine whether any particular element of group capital counts towards that requirement or not.
25.28 This method calculates capital adequacy using a deduction and aggregation approach. This is generally applied to insurance groups. It is broadly comparable with the risk-based aggregation method.
25.29 Under this approach, the capital requirement for the group is the sum of the capital requirements for each group member. The capital resources which are compared with this figure are the capital of the parent company plus, for each other group entity, its total capital less the book value of the parent’s investment in it (which will usually be zero plus any retained profits not reflected in the parent’s valuation of its holding).
25.31 Under this method the capital requirement is the sum of the capital requirement of the group parent and, for each subsidiary, the higher of either its solo capital requirement or the book value of the parent’s investment in it. What this approach effectively achieves is to go through the parent balance sheet, replacing each holding with the actual asset value of the entity concerned. Thus, where (p. 489) the parent is carrying a holding in a subsidiary at a value which is less than that subsidiary’s capital requirement, the parent must in effect hold capital equal to the difference between the two over and above its holding in the subsidiary itself.
25.33 It should be noted that the financial conglomerates regime permits the consolidated supervision of groups containing entities which have no capital ties between some of the members. This can arise, for example, where the group is deemed to exist because different EU members are ultimately owned by a non-EU non-financial parent company. This situation would arise if, for example, a Japanese motor company owned two small banks in the UK in circumstances where banking as a whole formed a trivial part of the overall balance sheet of the parent. In such a case Method 4 is the only practical method which can be applied due to the absence of a single financial parent company.
25.34 One of the more important elements of bank group supervision is the application of limits on the size of individual risk concentrations and of intra-group transactions to the group itself. However, the application of such restrictions within financial conglomerates is exceptionally difficult, not least because these rules are not harmonized across the insurance and banking sectors. Where a group contains both banking and insurance business, the relevant rules are applied to the functional subgroup—thus the banks within a mixed group are subject to limits on concentration and intra-group transfers as a whole as if they formed a banking group headed by the ultimate holding company. However, insurance firms within the group are not within this requirement. Technically the directive gives national supervisors the power to impose the rules of the dominant sector across the group as a whole—however, this has generally proved impossible in practice.
25.35 It should be noted that asset management activity may for this purpose be classed either as investment or as insurance at the election of the group. Asset managers will be deemed to fall within the banking and investment group, but if this were an absolute rule it could have the effect that insurance groups with large independent asset management businesses could find themselves inadvertently falling into the conglomerate regime. As a result, a group may make an election to treat its asset management activity generally as insurance business if it so desires.
25.36 The directives permit groups to be exempted from the requirements of the conglomerates directive if they are subject to equivalent supervision in their home jurisdiction. The EU has concluded that the regimes which are formally equivalent for this purpose are the Swiss regimes (both the Swiss federal Banking Commission and the Federal Office of Private Insurance) and, in the US, the consolidated supervision rules of the Federal Reserve and the Office of the Comptroller of the Currency, the New York State Banking Department, and the Securities and Exchange Commission. Note, however, that groups headed by a US insurance company would not satisfy the equivalence test since US insurance regulation was not held to be compliant with the EU model.4
25.37 Where a group is headed by an entity which is not subject to equivalent rules, the relevant EU supervisor is, under the directive, given broad discretion to impose whatever measures he sees fit for the purpose of ensuring appropriate supervision. In practice this has tended to involve the imposition of high stand-alone capital ratios on the relevant regulated entity. However, regulators have the power (if they so choose) to impose on such entities the full weight of consolidated supervision at the group level.
3 See para 25.01.