مقررات بانک های چند ملیتی : پژوهش نظری

This paper examines national regulators’ incentives to intervene in a multinational bank’s activities and the extent to which these incentives differ with the bank’s foreign representation choice (branch or subsidiary). Shared liability leads to higher incentives for intervention than legal separation. Cross-border deposit insurance, on the other hand, yields less intervention than when regulators compensate local depositors only. Based on these results, we derive implications for multinational banks’ and regulators’ preference on foreign expansion and representation.

Multinational banking has expanded significantly as barriers to international capital flows have been progressively dismantled and entry to foreign markets has eased.2 The rapid development of multinational banks (MNBs hereafter) represents a source of new concerns for regulators. Regulation of an MNB in one country may well affect the behavior of the bank and of regulators in other countries, as shown, for example, in the rescue of Belgian–Dutch bank Fortis in 2008. This paper provides a simple framework for examining regulation of an MNB by independent national authorities and the extent to which foreign representation affects both regulatory actions and the bank’s choice of foreign expansion. Banks pursuing a full range of activities abroad can be represented in the foreign countries through branches or subsidiaries.3 Representation affects both the liability structure between the home and foreign units, and the allocation of supervisory responsibilities between regulators of the home and host countries. One can think of branches as extensions of the home bank: the two institutions share joint liability for the failure of their assets and they call upon the same deposit insurance fund. Branch-represented MNBs are usually overseen by the regulator of the home bank. Subsidiary banks are the assets of the home bank: while the subsidiary and home bank share liability for the home bank’s assets, the home bank has no liability in case of subsidiary failure. Reflecting the higher independence of subsidiaries relative to branches, national regulators have independent power over the locally incorporated units, and depositors are covered by the local deposit insurance fund in the event of bankruptcy.4 In this paper, we address two related questions. How does the liability structure between the home and foreign units and the allocation of supervisory responsibilities affect regulators’ incentives to intervene in bank units? How does regulation influence a bank’s decision regarding foreign expansion and its choice of foreign representation? We consider a setup where an MNB operates in two countries. The bank is legally incorporated in the ‘home country’ (with a home unit) and operates an additional unit in the ‘foreign country’ (foreign unit). Each unit collects deposits and invests locally in risky and illiquid projects that are uncorrelated across countries. Bank regulators have two mandates: they (fully) insure depositors and exercise prudential intervention over the unit they are in charge of. We first consider a setup in which regulators exercise their mandates with the aim of minimizing costs stemming from their deposit insurance function.5 Regulators thus face the following trade-off: early intervention results in a positive liquidation value for a unit’s assets, but leads to certain deposit insurance costs; letting a unit continue might lead to no costs for the regulator if the unit’s investment pays out, but might result in a higher deposit insurance cost if it does not. We show that there is a material difference in the likelihood of regulatory intervention in a domestic bank and an MNB represented abroad by either a branch or a subsidiary. The differences can be attributed to two effects: (i) the extent to which the regulator of a given unit of an MNB can draw upon the (residual) assets of the other unit if the unit’s assets fall short of liabilities; and (ii) the regulator’s responsibility for insuring depositors located in the other country. Where it occurs, shared liability among the MNB’s units provides higher incentives for regulatory intervention than with a domestic bank or when units are legally separate. This is because residual assets from the other unit have a higher value to the regulator when it intervenes in the unit it is in charge of than when it does not. In the former case, residual assets available in the other unit lower the regulator’s deposit insurance costs with certainty. With no intervention, however, those assets are valuable to the regulator only if the unit it is in charge of eventually fails (i.e., the residual assets are valued with a probability smaller than one). Hence, this effect, which we dub the liability effect, induces a regulator to intervene, a ‘tough response’. On the other hand, insuring depositors in both countries, as with branch representation, makes a regulator internalize the full costs of its decision. In particular, the regulator takes into consideration that intervention in one unit leaves fewer or no assets to support the other unit in case of need: this internalization effect reduces the regulator’s incentives to intervene in a unit, a ‘soft response’. These two effects are the main drives of a number of interesting implications concerning regulators’ behavior. We show that the home unit faces tougher regulation when it is part of a subsidiary MNB than when it is simply a domestic bank (i.e., regulatory decisions either coincide, or the home unit of the MNB faces intervention when the domestic bank does not). This is due to the liability effect that is at play in an MNB. On the contrary, the home unit in a branch MNB falls under softer regulation than the same unit in a subsidiary MNB (i.e., either intervention takes place in both representations or it does so with subsidiary but not with branch MNBs). This is because the liability effect is present with both representations, but the internalization effect is only at work with a branch MNB. Furthermore, the home unit of a branch MNB may face a softer or tougher regulation than a domestic bank depending on the foreign unit’s prospects since it determines which of the two effects prevails. Concerning regulation of the foreign unit, the rule for intervention in the subsidiary coincides with that in a domestic bank since the foreign regulator is affected neither by the liability nor by the internalization effects. Furthermore, as regulation of the subsidiary then coincides with that of a domestic bank, it follows that a subsidiary faces softer regulation than a branch if the home unit’s prospects are good (i.e., its probability of failure is low), but faces tougher regulation if those prospects are bad (i.e., its probability of failure is high). The different regulatory regimes a bank faces are often said to be relevant to the choice of foreign representation, even if these are certainly not the sole drivers of the decision (see Houpt, 1999, Calzolari and Loranth, 2003 and Focarelli and Pozzolo, 2006). By comparing the likelihood of intervention under foreign expansion and the type of foreign representation, we can examine the extent to which regulation alters bank’s profit and thereby the bank’s choice of whether to expand abroad, and if so the choice of foreign representation. In our model, in the absence of regulation, a domestic bank would always choose to expand abroad with a subsidiary, as projects in the two units are uncorrelated and the home unit is shielded from the subsidiary’s losses. However, this intuition does not carry over when the bank is subject to regulation. In particular, we show that a subsidiary structure is only preferred if the home unit has a sufficiently high probability of success. The reason is that the presence of a subsidiary makes the home regulator tougher on the home unit, hence the subsidiary’s additional profit may not compensate for the lower expected profit of the home unit. The bank might also prefer a branch structure to the subsidiary structure when the home unit has a sufficiently high failure probability project. This again may seem counterintuitive at first glance, as given the joint liability of units for losses, by adopting a branch structure, the bank ceteris paribus increases the risk of losing profit compared to a subsidiary structure. We show that the reason for a branch structure being preferred is that the bank can elicit softer regulation.6 In addition to costs, regulators may also be concerned about bank profits as a consequence of successful lobbying or because profits affect the financial stability of the local banking sector. As one may expect, profit concerns shift regulatory decisions towards softer behavior. More interestingly, we find that when regulators care about bank profits in their jurisdiction the liability effect can make the regulator softer, contrary to the case of cost-minimizing regulators. We also show that in this case a domestic bank can no longer hope for softer regulation when it expands with a branch abroad. The implication of this is that the subsidiary structure will become more desirable from the bank’s point of view. The present work is part of a growing literature on the regulation of MNBs. Calzolari and Loranth (2003) provide a general introduction to the issue. 7Holthausen and Rønde, 2003, Acharya, 2003, Dell’Ariccia and Marquez, 2006 and Dalen and Olsen, 2003 address the problem of divergent interests and lack of coordination between national regulators. Our new insight is that divergence of interests (and decisions) between regulators might result from the representation choice of the MNB. Unlike earlier work, we examine in detail the interplay between the MNB’s liability structure and the allocation of supervisory functions and their effect on prudential supervision and representation choice. The question of whether the form of MNB representation affects regulation has also been raised by Harr and Rønde, 2004 and Loranth and Morrison, 2007.8 These papers focus on optimal capital requirements, while our contribution instead focuses on regulators’ incentives to take disciplinary actions. In an early contribution, Repullo (2001) addresses the problem of the domestic regulator’s limited information (for supervision and deposit insurance) and offers some conclusions concerning the incentives that lead to cross-border takeovers. Our paper differs from these in studying the effects on regulation of the form of MNB representation and in endogenizing the MNB’s choice of representation. Our modelling choice of bank supervision is most closely aligned with that of Mailath and Mester (1994), who also consider a positive theory of regulatory intervention. In their paper, a (domestic) bank invests sequentially in two projects, and regulatory intervention may prevent the financing of the second project. The (single) regulator anticipates future bank’s asset choice (over safe or risky assets) in case it is permitted to remain open and, understanding that the initial investment decision will affect the regulator’s policy, the bank modifies its first period risk-taking behavior. Considering both a cost-minimizing and welfare-maximizing regulator, this paper shows some analogies with our analysis. In particular, the authors show that the regulator might leave open a bank with negative expected value, if the returns of the second project will allow the regulator to reduce the reimbursement costs of the first-period project. Thus, in their paper, too, regulatory decisions might be driven by the possibility of getting some equity out of the second project. However, the main driver for this effect is different in the two papers. In their paper, it is generated by a dynamic interaction between the bank and the regulator, and an essential ingredient is the commitment problem on the side of the regulator to punish the bank after having chosen a risky (negative net present value) project. This also explains why the issue still remains when the regulator maximizes welfare. In our paper, however, the liability effect arises because of shared liability between two (geographically) separate units. This effect is present as long as units are separately supervised and even when regulators maximize welfare. However, the effect vanishes with a single regulator maximizing welfare in two countries. Finally, our analysis of an MNB with subsidiary representation brings in strategic interaction among independent national regulators, which is clearly absent from Mailath and Mester (1994). The rest of the paper is organized as follows. Section 2 presents the base model. Section 3 analyzes regulators’ incentives to intervene under the two representations. Section 4 discusses banks’ and regulators’ preferences over foreign expansion and representation. Section 5 extends the base model to regulators that have an interest in the MNB’s profit. Section 6 concludes. Proofs are in Appendix A.