132. The Tripartite Group has had detailed discussions on how supervisors should regard a parent institution's participation of less than 100% in a financial subsidiary:

- in relation to the responsibility of the parent for regulatory capital shortfalls in the dependant;

- for the purposes of valuing the participation of the parent in the dependant in the balance sheet of the parent;

- for the purposes of assessing capital adequacy on a group-wide basis, and of eliminating double gearing;

Particular issues that arise in considering these questions are:

- in the event of a capital shortfall, what is the responsibility of the parent company? Does this responsibility vary according to the level of the participation and the distribution of the balance of the share holdings (including as between other regulated financial institutions and non-regulated shareholders). If so, how?

- if there is excess capital in a dependant, to what extent can the parent attribute value to its holding for regulatory or risk­bearing purposes?

- the transferability of excess capital from a partly­owned dependant to its parent and from one dependant to a sister company (via the parent), reflecting any legal, taxation or regulatory restrictions on the distribution of surplus capital in the dependant to the parent and/or other shareholders.

133. Where the parent holds 100% of the share capital and voting rights in a subsidiary, then capital in that subsidiary not required there for regulatory purposes will in normal circumstances be available to the parent (and, hence, to other parts of the group), subject to any other requirements (e.g. legal, tax or foreign exchange control restrictions). Provided that excess capital, in addition to that required by the regulator of the subsidiary, is of a type which is acceptable to the regulator of the parent and there are no current or foreseeable restrictions on its transfer, it is not imprudent to allow such an excess to be regarded as available for the bearing of risks by the parent institution or by other entities in the group.

134. The position is, however, less clear-cut when external holdings exist in a dependant company. Partly­owned undertakings can be categorised in a number of ways, for example:

- Subsidiary undertakings over which control is established, either by the group owning more than 50% of the shares or the voting rights, or through a contractual or other arrangement;

- "Associated undertakings", denoting for these purposes undertakings over which the group does not have control but does have significant influence (i.e. in the sense of a group shareholding or share of the voting rights of between 20% and 50%);

- Simple minority shareholdings in undertakings over which the group has neither control nor significant influence (i.e. the group shareholding or share of voting rights is less than 20%).

136. Simple minority shareholdings (less than 20%) in undertakings over which the group has neither control nor significant influence do not present a problem. There is widespread agreement that, as a general rule, these small participations, over which the group has no significant influence and which are not consolidated for accounting purposes, should be treated in accordance with the solo entity rules for assessing the capital requirements of the "parent" undertaking. Only in very exceptional circumstances, would supervisors expect to consider proportional or full integration to be appropriate.

137. Similarly, in the case of "associated undertakings", members of the Tripartite Group are agreed that they would normally expect the group's share of the undertaking's capital (and the same proportion of the undertaking's regulatory capital requirement) to be included in any group-wide assessment of capital adequacy. In other words, the pro-rata approach is generally advocated where the parent or the group has what is deemed to be a "significant" influence, subject to questions of availability and suitability discussed in the ensuing paragraphs of this section and in the next section of this chapter. However, some Tripartite Group members would wish to add a rider to the effect that if it appeared that the group subject to integration would by itself maintain the solvency of a particular "associated undertaking", then the whole of that undertaking's regulatory capital requirement should (if greater than the amount invested by the group) be included in the group-wide assessment of capital adequacy. The proportions of the undertaking held by third parties and the identity of those shareholders may be other factors which have a bearing on the most appropriate treatment.

138. There are, however, some differences of view with regard to the treatment of subsidiary undertakings which are not wholly owned, but over which a group has effective control (either through the ownership of more than 50% of the shares or voting rights, or through a contractual or other arrangement). Some members favour the full integration of such controlling interests for the purposes of assessing capital adequacy from a group-wide perspective (i.e. the full integration of the capital and risks prevailing in the partly-owned subsidiary); others support integration on a pro-rata basis.

139. Those who argue in favour of full integration of such controlling participations with no specific safeguards being necessary make the following points:

- A "break-up" or liquidation value approach is inconsistent with the assumptions underlying the prudential framework of many supervisors. If such an approach were to be followed, for example, some bank and insurance supervisors would have to move to a mark-to-market valuation of all assets.

- From a going concern standpoint, full integration makes more sense than pro-rata integration in the assessment of capital adequacy from an overall group perspective. It recognises the majority shareholder's ability to effect the transfer of marketable assets or the granting of subordinated loans within the group (although supervisors may wish to prevent certain intra-group transactions).

- Controlling participations give the parent company a responsibility for the risks run by its subsidiary which goes further than the mere proportion of capital it has contributed and, in many cases, would extend to the totality of the risks. In the same way, a controlling participation gives the parent company a control of the own funds of a subsidiary which goes further than its contribution to its capital; indeed, it gives the parent important policy powers over the structure and restructuring of the subsidiary's own funds, subject to the legal rights of the minority shareholders.

- If there are doubts as to the availability of surplus capital in a majority-owned subsidiary for the covering of risks at parent/group level, then in reality those doubts apply both to that part of the surplus which is attributable to third parties and to that part which is attributable to the parent group.

- Full integration of controlling participations is in line with international accounting standards and, where possible, it makes sense to achieve consistency between accounting and prudential rules, especially given the certainty and guaranteed status of audited accounts. The same reasons which justify a full accounting integration of majority subsidiaries are valid from a prudential point of view.

- Full integration is consistent with the internationally agreed supervisory regime for banks. Indeed, minority interests are specifically recognised as capital at group level by bank regulators in both the Basle Accord and in the EU.

- The solo supervision of individual firms within a group would ensure that, under no circumstances, are minority interests (or any other form of excess capital) in one entity used to cover a deficit in another.

140. Other members of the Tripartite Group feel that full integration is inappropriate because there is a need to examine the attributability and to check the availability of any surplus funds. These members take the view that a surplus of an undertaking's qualifying capital over its regulatory capital requirement should not be included in the regulatory capital of its parent where there is any doubt as to the undertaking's ability to transfer that surplus to other undertakings in the group. The main reason for this view is that bringing into account at group level surpluses in subsidiaries which are not attributable to the parent, or which may prove not to be transferable to it, could give an illusory feeling of confidence about a group. These members maintain that the only value of a non-transferable surplus is that it will provide a buffer against exceptional losses in the event that they arise in the subsidiary in which the surplus is located; they say that this does not, in itself, justify inclusion in group capital which, by definition, should be attributable and available to the group as a whole.

141. On the question of attributability, the real point at issue is whether capital surpluses which are attributable to third parties should be recognised as an acceptable form of capital when supervisors assess capital adequacy from a group perspective. If they are to be so recognised, the implications are that capital surpluses attributable to third parties could be used to compensate for what might otherwise be seen as a case of double gearing between a parent and another wholly-owned subsidiary. In order to be sure of preventing the group / parent from benefiting in this way, it would appear necessary for the regulator of the parent to deduct the value of the participation from the parent's own regulatory capital base or to take other comparable measures to satisfy himself regarding the distribution of capital within the group. However, supporters of full integration would take the view that the whole of the capital surplus is indeed available to the parent/group; and that, at group level, capital surpluses attributable to third parties could also be used to cover notional deficits in unregulated entities if they are fully integrated at parent / group level.

142. Protagonists of the pro-rata approach also question whether the solo supervision of individual firms within a group would in fact prevent capital surpluses attributable to third parties being used to cover notional deficits in unregulated entities. The situation they have in mind is where full integration of a parent and a partly­owned regulated subsidiary reveals a surplus stemming from the subsidiary (see Example E in Appendix III). If a second undercapitalised unregulated subsidiary is included in the group­wide perspective, full integration could still reveal a group surplus because it would be possible for the excess in the first partly­owned subsidiary to conceal the deficit in the second. This would not be the case with pro­rata integration which, by its very nature, takes into account only that part of the surplus which is attributable to the group.

143. As for the availability aspect, there are a number of circumstances in which the whole or part of a surplus may not be available to the group. Distributions of profit, dividend taxation, foreign exchange controls, withholding taxes, covenants given to suppliers of debt finance, legal restrictions, general creditors' rights, solo regulatory requirements (such as the required ratio between tier 1 and tier 2 capital and any restrictions on capital withdrawals) and the relationship with minority interests are examples of such circumstances. Moreover, in the case of capital surpluses which have been generated by the issue of subordinated debt in a subsidiary, for example (see Example H in Appendix III), it could be considered inappropriate to regard any surplus capital generated by that debt as being attributable to the group as a whole (unless the debt is also subordinated to the liabilities of the parent). It is arguable that non­transferable resources in other group companies will not help the parent company to support a troubled dependant; that they will not prevent the troubled entity being wound up; and that they will be of no use to the parent if its own capital becomes depleted (e.g. if it has to write off its investments in troubled subsidiaries).

144. Many of the impediments to the availability (i.e. the free movement) of capital surpluses can apply equally to that part of a surplus which is attributable to the parent and that part which is attributable to a minority shareholder. However, protagonists of the pro-rata approach tend to view the very fact that part of the surplus is attributable to a third party as an impediment to free movement in itself.

145. Example F in Appendix III shows that, where a partly­owned subsidiary has a capital surplus, full integration can be viewed as the least conservative of the approaches. A parent increases its stake in a partly-owned subsidiary with a capital surplus, inducing its regulator to switch from pro­rata to full integration (because it is now regarded as the controlling shareholder). Although the parent can be viewed as having taken on increased risks, the assessment of capital adequacy on a group basis shows an improved capital surplus because full account has been taken of the subsidiary's excess capital. Protagonists of full integration, however, maintain that this is an accurate reflection of the position from a going concern perspective because the parent's increased stake in the subsidiary provides it with an overriding influence over the subsidiary's excess capital and the whole of the excess should therefore be regarded as available to the group.

146. However, full integration need not necessarily always benefit the parent. If the subsidiary in Example F were to have a capital deficit, full integration would have the effect of upstreaming the full deficit to group level, probably giving rise to a deficit at that level. In other words, it holds the controlling shareholder responsible for making good the whole of the deficit. Pro­rata integration, on the other hand, might still reveal a surplus at group level.

147. Some members of the Tripartite Group draw attention to what they regard as a significant flaw associated with full integration. They maintain that if the capital surpluses of partly-owned subsidiaries are integrated in full at group level, then it is possible to conceal instances of double gearing. The basis of this argument is that the full integration of a capital surplus at solo level in a (partly­owned) subsidiary could be used to conceal double gearing between a parent and another (wholly­owned) subsidiary (see Example D in Appendix III). Faced with evidence of double gearing between the parent and the wholly­owned subsidiary, the supervisor of the parent would invariably instruct the parent to inject more capital or to reduce its risks. However, the element of double gearing is not evident if the partly­owned subsidiary is integrated in full and action on the part of the supervisor does not appear to be justified. In these circumstances, the parent / group can be said to be benefiting ­ albeit indirectly ­ from the full extent of the capital surplus in the second subsidiary. In order to prevent this occurring, it would appear necessary for the regulator of the parent to deduct the value of the participation from the parent's own regulatory capital base or to take other comparable measures to satisfy himself regarding the distribution of capital within the group. However, supporters of full integration would take the view that the whole of the capital surplus is indeed available to the parent / group.

148. Some members of the Tripartite Group have pointed out that it is possible to combine the full integration of deficits with the pro­rata attribution of excess capital. Such an asymmetric approach would be consistent to the extent that it would always produce the most prudent result, rather than mechanically consistent in selecting the same approach for every situation regardless of the result achieved. It would ensure that the controlling parent at least had the means to make good a deficit in a subsidiary even if, in the event, it was not required to do so. On some occasions, the full integration of deficits would clearly overstate the extent to which the parent is ultimately liable for a subsidiary's deficit, but supporters of the asymmetric approach see this as infinitely preferable to the pro­rata approach, which obviously understates the parent's obligations in cases where risks of contagion are prevalent. It is also suggested that it would be comparatively rare for majority shareholders to allow a subsidiary to fail just because additional funds were not forthcoming from minority shareholders. Again, the need for a group-wide perspective, and regard for supervision of a conglomerate as a single economic unit, rather than a collection of individual entities, becomes apparent. To some extent, it might be argued that supervisors should be seeking to ensure that the solvency position of a financial conglomerate is better than the sum of the capital requirements of its individual entities.

149. In summary, the logic of full integration is that the conglomerate is viewed as a single economic unit with full account being taken of all capital which is both available to cover risks and recognised by the respective regulators. The viability of the group in the long run is assumed (the "going concern" perspective) and it is expected that the controlling shareholder will inject fresh capital if and when necessary for business reasons. Full integration supplies the regulator with a picture of the group as a whole; any minority interests are part of that overall perspective and need to be included if the supervisor is to be fully informed as to the capital situated in a group. What full integration does not do automatically is to determine whether or not there is adequate distribution of the capital in relation to the risks run by the group. For its part, pro­rata integration takes cognisance of capital which is attributable to minority shareholders (to the extent that it covers risks in the entity in which it is situated), but also intrinsically recognises that such capital is not attributable to the group / parent and may not be available to cover losses elsewhere in the group. Pro­rata integration provides a perspective of the capital position of the parent company as a separate legal entity. As such, however, it does not give an accurate reflection of the situation when a subsidiary has a capital deficit and the parent's responsibility for making good that deficit exceeds its pro­rata share.

Conclusions

150. The Tripartite Group agreed that, of the three main techniques considered (building­block prudential, risk­based aggregation, risk­based deduction), the building block prudential technique was inclined towards the full integration approach to majority shareholdings, simply by virtue of its reliance on consolidated accounts. Both forms of risk­based aggregation and risk­based deduction, on the other hand, tended to point towards the pro­rata approach to majority shareholdings. That said, all three approaches are in principle capable of being adapted to accommodate either the full or pro rata integration approaches to majority shareholdings.

151. Most members of the Group agreed that in situations where a deficit occurs in a subsidiary, as a general rule the full extent of that deficit should be attributed to a parent with a controlling participation or to a parent which has provided a guarantee to the subsidiary. Only in very exceptional circumstances ­ where the relationship between a parent and a majority­owned subsidiary can be distinguished from other parent / subsidiary relationships and the risk of contagion is considered to be minimal ­ would pro­rata integration be regarded as sufficient for the purposes of assessing overall group solvency. Equally, however, it should not be implied that deficits in subsidiaries will be tolerated by supervisors other than in the very short term; by their very nature, deficits need to be remedied quickly or the subsidiary put into insolvency proceedings. In any event, it will probably be rare for group financial returns to incorporate unremedied deficits in regulated subsidiaries if solo supervision is working properly, unless the group as a whole is at risk.

152. Where capital in a partly­owned subsidiary exceeds the regulatory requirements of the solo supervisor, the Tripartite Group agreed that, in principle, it was legitimate for such excess to be used to cover risks at the parent / group level, but subject to certain conditions.

153. The Tripartite Group was however unable to agree on the conditions for the use of excess capital in this way. Some argued that, consistent with the philosophy of regarding the group as a single economic unit, all excess capital, including that attributable to minority shareholders, could legitimately be used to cover risks at parent / group level, subject only to the effective solo supervision of individual entities and a check on adequate distribution of capital and risks within the group (which solo supervision should, of course, ensure). They do not believe that a further check on the availability and transferability of the excesses is necessary because they take the view that, if the supervisors have any doubts about the free movement of capital around the group (i.e. any doubts about the ability of the group to operate as a cohesive unit), they should not allow the financial conglomerate to exist in that particular shape and form. Others argued that a pro-rata recognition ­ i.e. recognising only the share of the surplus attributable to group shareholders ­ was the more prudent approach and that, even then, any amount thus recognised must be genuinely available and transferable. Notwithstanding the influence a controlling shareholder may be able to bring to bear, supporters of the pro­rata approach do not believe that supervisors should regard capital which is attributable to minority shareholders as being available to the group as a whole.