Impact and implementation challenges of the Basel framework for emerging, developing and small economies.

Frait, Jan ; Tomsik, Vladimir

INTRODUCTION

In response to the global financial crisis, the Basel Committee on Banking Supervision (BCBS) introduced two sets of reforms to the international capital framework for banks. The 'Basel 2.5' package of reforms (BCBS, 2009) included measures to strengthen the trading book capital requirements under Basel II and enhance the three pillars of the Basel II framework. Basel III (BCBS, 2010b) introduced two additional capital buffers: a capital conservation buffer of 2.5% that should be applied across the board, and a countercyclical buffer, which should be applied during periods of high credit growth. In addition to the level of capital, Basel III addresses the issue of its quality through its focus on common equity. Capital loss absorption is dealt with by the point of non-viability clause that provides relevant authorities with discretion for a write-off or conversion to common shares if the bank is judged to be non-viable. The capital framework was also supplemented by a non-risk-based leverage ratio. There is also substantial strengthening of the counterparty credit risk framework. Most of Basel III relates to Pillar 1 measures (minimum capital requirements), but there are also implications for Pillar 2 (further requirements by supervisor related to risks not covered by Pillar 1) and Pillar 3 (set of disclosure requirements). Basel III also introduced two required liquidity ratios: the liquidity coverage ratio (LCR) and the net stable funding ratio (NSFR).

The Basel III as well as Basel II standards are designed primarily for large internationally operating banking institutions with a focus on their activities in BCBS member states. They set minimum regulatory standards and respect the need to set the overall standards in various jurisdictions according to local conditions. Nevertheless, the Basel standards are to a large extent reflected in the regulations in BCBS non-member countries too. The prime focus of this paper is the impact of Basel III changes to the capital and liquidity frameworks on emerging market and developing economies, and specifically also on small economies. For simplicity, we label this group as emerging market and smaller economies (EMSEs).

The EMSEs are a rather heterogeneous group. Some of the countries have their own currencies, some are members of currency unions, while others utilize the currencies of other economies. Some are members of the European Union (EU), which imposes extensive regulatory requirements designed outside the scope of EMSEs. Furthermore, it is obvious that jurisdictions within this group differ substantially and will continue to do so. Despite all these differences, the EMSEs also have some common features. Their financial markets are typically more volatile than those of large advanced economies, as is their GDP growth, which is meanwhile higher on average than that of advanced economies. They also typically experience stronger credit growth because their financial sectors are in the process of financial deepening. They usually have lower credit ratings and shallower government bond markets, which in some cases implies a lack of high-quality liquid assets (HQLA), a lack of adequate collateral and so on. And they face a whole range of home-host issues, as many globally significant financial institutions operate branches or subsidiaries in EMSEs.

The presence of branches and subsidiaries of multinational banks is the most important feature of EMSEs' banking sectors. Whether multinational banks operate in a particular jurisdiction as a branch or subsidiary creates striking differences. Nevertheless, as Mayes and Granlund (2008) explain, both ways will inevitably pose problems for the regulatory and supervisory authorities in both home and host countries. (1) As to the choice of a multinational bank regarding preferred organizational form, Cerutti et al. (2007) find that banks are more likely to operate as branches in countries that have higher taxes and lower regulatory restrictions on bank entry and on foreign branches. Subsidiary operations are preferred by banks seeking to penetrate host markets by establishing large retail operations. These findings are to a large extent confirmed by Fiechter et al. (2011). They conclude that banks with significant wholesale operations tend to prefer a more centralized branch model that provides them the flexibility to manage liquidity and credit risks globally and to serve the needs of large clients. The funding costs for the wholesale group are likely to be lower under the branch structure, given the flexibility to move funds to where they are most needed. A subsidiary structure, in contrast, constrains the banking group's ability to transfer funds across borders and hence may be less suitable for wholesale activities. However, they also find that, given the diversity of business lines and the varying objectives and stages of financial development of different countries, there is no one obvious structure that is best suited to all cases for cross-border expansion.

EMSEs are impacted by the Basel III framework in three ways. First, there is an indirect impact of the implementation of the framework at the consolidated level by jurisdictions that are home to international banks. This involves increased cross-border cooperation challenges, and it is all the more important for countries who are hosts to branches or subsidiaries that may represent a small fraction of the consolidated group but are locally systemic. Second, there is the indirect impact of the reactions of international banks to implementation. In this case, there is the concern that as international banks change their business and deleverage, their exposures to EMSEs might be unduly reduced. Finally, there is the direct impact on EMSEs of implementing the Basel III framework in their own jurisdictions. There are many papers and studies looking at the economic effects and impacts of the Basel III framework in advanced economies using macroeconomic models (see the section 'Basel III, regulatory capital and its quality'). The implied changes in volumes and prices of credit constitute the driving forces in these approaches. In contrast, studies focusing on more detailed effects stemming from changes in individual regulatory parameters often specific to emerging and small economies are scarce. This paper intends to provide a contribution in this direction.

The structure of the paper is as follows. The next section surveys the analyses on the economic effects of the Basel III framework and the potential consequences of new capital regulations in EMSEs. The section after that considers the issues associated with the introduction of macroprudential capital buffers. The following section discusses the implications of new regulation of banks' trading exposures for affiliates of multinational banks. The subsequent section turns attention to the sensitive issue of regulatory treatment of sovereign exposure from both the micro- and the macroprudential perspective. The next section explains potential risks associated with the accelerated move of the EMSEs' banks to advanced approaches to modelling financial risks. The penultimate section describes impacts of the new liquidity framework while the final section concludes.

Before proceeding to these topics, let us note that we agree on the importance and adequacy of most of the changes that are contained in Basel III. However, we deliberately focus mainly on challenges and risks, especially on the identification of possible unintended consequences of the new rules. This is not to say that the unintended consequences prevail over the intended ones. Furthermore, not all the issues discussed here are related directly to Basel III, and not all the unintended consequences are specific exclusively to the EMSE countries. Many of the proposed rules were discussed long before the establishment of the new regulatory framework. However, the creation of the new framework gives new relevance to these issues and consequences.

BASEL III, REGULATORY CAPITAL AND ITS QUALITY

The changes in the capital framework that constitute the core of Basel III are just one, though the most important, component of the post-crisis regulatory overhaul. These changes concern not only the traditional part of banking regulation, capital requirements for covering risks faced by the individual institutions that stand at the centre of the Basel II accord. Basel III also contains parts of the new economic policies labelled 'macroprudential policy' that are designed to prevent or mitigate risks of a systemic nature. This reflects recognition that liberalized financial markets have created a favourable environment for endogenous 'boom and bust' cycles. The new framework thus takes into account both micro- and macroprudential concerns and struggles to achieve robustness as a key to avoiding financial sector vulnerability. For a bank-based system, robustness will be achieved via a high loss absorbency capacity, strong liquidity and brakes on credit booms. Loss absorption should cover potential losses through sufficient provisions against loan impairment, through capital cushions, and cyclical losses through countercyclical capital buffers. There should also be a cross-section risk component consisting of capital add-ons for systemically important institutions. Strong liquidity is essential for limiting the fragility of liabilities.

Impact studies of Basel III were naturally a part of its preparation. The focus was mostly on the estimates of macroeconomic effects in advanced economies. BCBS has advocated Basel III on the basis of the Long-Term Economic Impact (LEI) Report (BCBS, 2010c), which focuses on the long-term effects while ignoring transitional costs. The study admits the higher cost of external finance (higher lending rates), nevertheless, the overall balance of cost and benefits is rather positive up to a 15% of capital adequacy ratio. This result is mainly because of the high perceived benefits of a reduced probability of a financial crisis. BCBS has also arranged for a study of transitional effects of Basel III implementation by the MAG (Macroeconomic Assessment Group). This study (MAG, 2010) assumed that banks face higher costs when funding assets by capital than by deposits or by debt. The initial effect of Basel III introduction is an increase in lending rates. However, in the long run the banks get less risky, the costs of funding go downwards, lending rates decline and supply of credit is restored. The result is that in the short run the costs are low and in the long run they disappear. When new liquidity requirements are added, the costs go a bit further, but the overall picture is not affected. This positive view of the long-term effects has been confirmed by Angelini et al. (2011), Slovik and Cournede (2011), Kashyap et al. (2010), Miles et al. (2011) and Admati et al. (2011).

Even though the impact studies and estimates of the effects were targeted at the advanced economies, they should hold for the EMSEs as well. In addition, the short-term negative effects on lending rates could be even lower here than in advanced economies since most EMSEs will not have great difficulties in complying with the new definition of capital in Basel III. This is because the banks in EMSEs generally maintain capital levels in excess of internationally agreed regulatory minimums and their capital base is typically dominated by common shares and retained earnings (labelled Core Tier 1, CT1; or Common Equity Tier 1, CET1). Nevertheless, there may be several implementation challenges. First of all, implementation of Basel III will generate a need for some capital replenishment. Reasons for this include: (i) banks in EMSEs inevitably need to issue additional capital given their relatively fast economic growth and the pivotal role played by banks; (ii) higher minimum regulatory capital requirements at the international level will likely lead banks in EMSEs to build up capital to maintain buffers against a relatively higher degree of macroeconomic and market volatility; and (iii) internationally active banks often use the sovereign credit rating of the host jurisdiction as the credit ceiling or the risk floor for all the exposures incurred by their subsidiaries.

EMSEs with lower credit ratings could thus find their banking system with higher capital levels than in advanced economies, regardless of whether banks in the latter are more exposed to other systemic risks.

The eligibility criteria for Tier 1 (common equity, retained earnings, reserves, non-redeemable preferred stock and other capital not secured by the issuer) and Tier 2 (hybrid instruments and subordinated debt) may turn out to be non-trivial. This applies in particular to the requirement that all these instruments have a 'point of non-viability clause', that is, all regulatory capital instruments should be able to absorb losses in the event that the issuing bank reaches a point of non-viability. In addition, once the clause is triggered, supervisors may face potential governance issues when conversion brings in shareholders that may not be appropriate. More broadly, the change of ownership structure may have implications for the viability of the institution going forward. It should be expected that the supervisor has the authority to replace management or to require some other change in management deemed necessary to ensure that the institution operates in a prudent manner. Basel III establishes a requirement that the terms of capital instruments must allow, at the option of the regulatory authority, for them to be written off or converted into common shares in the event that a bank is unable to support itself in the private market in the absence of such conversions. During the recent financial crisis, some troubled banks never reached the insolvency point where the subordinated instruments would have performed as capital because governments provided support to avoid the liquidation of these banks. It is clear that, if such support had not been provided, these banks would have failed and subordinated creditors would have received payment only after all depositors and senior creditors had been paid in full.

Basel III allows for certain debt instruments issued by banks to be included in additional Tier 1 and Tier 2 capital, subject to a set of inclusion criteria. Such policy will inevitably create a bias against those banks that are constrained by their market experience and credibility and thus face higher costs in issuing capital. As EMSEs generally lack local infrastructure to facilitate the issuing of structured capital instruments in domestic markets, local banks are faced with difficult, if not impossible, challenges to raise less costly capital. In addition, Basel III allows the instruments issued by subsidiaries to be included in the consolidated group's capital where the pre-specified trigger event is linked to the non-viability condition determined by both home and host authorities. In practice, there is a tendency for the internationally active banks to include their subsidiaries' issuance in the group's capital as these banks manage their risks and maximize their expected risk-adjusted returns by consolidating the group-wide assets and liabilities. Therefore, an immediate consequence of such policy is the potential for complications if the non-viability conditions determined by the home and host authorities are different. In addition, the possible conversion of structured instruments issued by subsidiaries into equity may lead to the potential dilution of shareholder holdings of both subsidiaries and their parent banks if the issuance is included in the group's capital, which will add to both supervision and management complications. All these potential complications will make capital issuance by subsidiaries a less desirable choice for both bank managers and investors.

Basel III also requires the banks to deduct from their common equity capital most of their assets with weaker loss-absorbing features such as minority interests, goodwill and deferred tax assets, as well as some investments or rights. In general, the use of innovative capital instruments and Tier 2 instruments is limited and most of the capital is composed of CET1 and reserves. While overall goodwill is the largest element of deductions for advanced economies, for emerging countries that participate in the quantitative impact study, the deduction of deferred tax assets seems to weight relatively more. In part, this may be because of provisioning and accounting legal frameworks; forward-looking provisions, for instance, frequently are not recognized by tax authorities and create deferred tax assets. For example, the deduction of deferred tax assets proposed by Basel III seems to weigh relatively more in EMSEs than in advanced economies. This may be partly because of provisioning and accounting legal frameworks, for example, forward-looking provisions frequently are not recognized by tax authorities and generate deferred tax assets. The World Bank (2013) estimates that the share of assets with less loss-absorbing characteristics to be deducted from CT1 capital is relatively small on average for banks in emerging and developing economies, except in the Latin America and Caribbean region. The proportion of CT1 capital to be deducted nevertheless varies greatly across banks. If applied immediately, the deductions combined with market risk adjustments (Basel 2.5 and trading exposures) would lower the CT1 ratio by about 1-3 percentage points on average. The overall impact on the CT1 ratio is the largest for Latin America, the Middle East and North Africa, and Europe and Central Asia.

Difficulties associated with the deductions and eligibility criteria for Tier 1 and Tier 2 instruments under Basel III may create pressure on EMSEs to implement changes to their legal frameworks. In the case of the eligibility criteria for Tier 1 and Tier 2 instruments, the national legal and regulatory frameworks should be consistent to ensure that the non-viability clause is effective. While the criteria driving the trigger for conversions or write-offs can in principle be set out a priori in broad terms, allowing for better pricing of the instruments and helping to reduce market uncertainty and the legal risk to supervisors, the decision as to whether a bank can continue on its own will ultimately always be a judgement call. As a result, supervisors need to be given sufficient powers to be able to make such decisions.

EMSEs may also need to take actions to cultivate domestic markets for the issuance of structured capital instruments, including: (i) having in place legal and institutional arrangements to enable the issuance of Basel III-recognized capital instruments; (ii) defining a priori criteria for the triggering of the point of non-viability, which may help boost market confidence and acceptance for issuance of structural capital instruments; (iii) having an enabling tax environment where the capital instruments issued by banks are typically tax-deductible and are not subject to withholding tax; and (iv) designing the terms of capital instruments to make such instruments suitable for a broad range of investors. If foreign bank subsidiaries operating in EMSEs issue capital, and this is included in the consolidated group's capital, authorities may need to impose certain regulatory requirements to avoid capital being used to cover losses of parent banks while simultaneously harming the confidence in the stability of subsidiaries. Enhanced communication between home and host regulators needs to be encouraged and should cover, among other issues, regular information sharing about the performance of related parent banks and subsidiaries and related resolution plans.

MACROPRUDENTIAL CAPITAL BUFFERS

The supervisory powers to implement the mandatory capital conservation and countercyclical buffers (CCyBs) will be essential to ensure effective implementation of Basel III. The values of the buffers themselves may not present any particular difficulties for EMSEs. The supervisory powers to implement these buffers may be more problematic. With the capital conservation and counter-cyclical buffers in place, the restrictions on the distribution of profits in cases of non-compliance with these capital buffers should be automatic and imposed on banks through requirements set forth by national legislation. The buffers cannot function as desired if the supervisor does not have sufficient power to restrict the distribution of profits, or if authorities do not have the will to activate the various triggers. Many countries are still struggling to create such powers for their regulators. In many EMSEs that have not yet implemented Pillar 2 of Basel II, the development of supervisory judgement that is crucial for Basel II and Basel III may be lacking. In addition, there are complex issues related to the interaction of these buffers, other additional buffers and Pillar 2. For instance, several countries already include additional Pillar 2 capital requirements for banks that are considered systemically important or to cover idiosyncratic risks. Authorities will need to consider whether these Pillar 2 charges overlap with the Basel III buffers. Supervisors need to assess whether the Pillar 2 add-ons could safely be drawn down in times of crisis. If this is not the case, such add-ons should be considered a minimum requirement for the bank.

Similar issues arise when discussing the calibration of the CCyB. The actual implementation of a CCyB in countries with underdeveloped credit markets seems to be more complex than what was proposed by the BCBS; finding the right indicators of systemic risk and establishing a framework to calculate the adequate levels of CCyB countries is still a work in progress (Drehmann et ai, 2010). In many EMSEs, it may not be adequate to mechanistically apply the recommended methodology by measuring excessive growth on the basis of deviations from the actual approximation of the long-term trend of credit relative to GDP. (2) The framework allows flexibility for supervisors to use judgement in defining other indicators, and the BCBS has published principles that can be used to assist supervisors to identify the build-up of systemic risk (BCBS, 2010a). EMSEs have often been well ahead in terms of applying macroprudential logic in their policies (Hahm et al, 2012; Moreno, 2011). Some had already implemented countercyclical and other macroprudential measures such as loan-to-value ratio on mortgage loans, debt-to-income ratio on credit cards and personal loans, and countercyclical provisioning, and now are working on how these measures will interact with the buffer and what their cumulative effect might be. Changes in the CCyB in EMSEs may also precipitate capital flows. As the volume of liquidity in a domestic banking sector is reduced because of the introduction of the CCyB, the result could be excessive liquidity outflows to other parts of the financial system, including to financial markets abroad. Even though the potential size of the flows induced by the CCyB itself may be limited, in combination with swings in monetary policy the effect could be sizeable.

In setting the CCyB and other macroprudential tools, the supervisors should use the flexibility provided by the framework to use judgement and/or undertake a comprehensive analysis to improve the understanding of credit cycles rather than mechanically relying on credit to GDP de-trending. The principal task is for them to make their own informed judgements about the equilibrium or sustainable level of credit in the economy. Subsequently, a set of forward-looking indicators providing information on the possible materialization of systemic risk resulting from currently emerging financial imbalances has to be employed for a thorough assessment. The CCyB may help the authorities to lean against the expansionary phase of the cycle by raising the cost of credit and therefore slowing down lending if they conclude that the stock of credit has grown to excessive levels relative to the benchmarks. Nevertheless, this potentially moderating effect on the expansionary phase of the credit cycle should be viewed as a positive side benefit rather than as the primary aim of the CCyB regime. The quantitative impact of the CCyB per se may be rather weak during a credit boom. Its major contribution is to help maintain the flow of credit in the economy when the broader financial system experiences distress after a credit boom and to assist in ensuring a smooth landing for both the banking sector and the real economy. No single policy tool used in isolation can tame credit booms. Therefore, other macroprudential tools, including as sector-specific ones, that could be used to enhance banks' resilience to credit booms should explicitly be set out in Basel standards.

BASEL 2.5 AND TRADING EXPOSURES

The capital requirements of Basel 2.5 were set to increase the resilience of banks against market risk in their trading books through a significant increase in the respective capital charge (BCBS, 2009). BCBS indicated that market risk capital requirements associated with Basel 2.5 will increase by an estimated average of three- to four-fold for large internationally active banks. (3) As regards EMSEs, two issues should be highlighted. The criteria for estimating risk-weighted assets (RWA) for exposures of local subsidiaries in host countries are decided by parent banks and home country supervisors. Global banks manage their risks and estimate RWA by consolidating all their subsidiaries' assets and liabilities at their parent bank. Hence, assets held by subsidiaries result in capital charges for the group as a whole. The highest credit quality risk for a host country local bank, domestic sovereign debt, could be transformed through the process of balance sheet consolidation into a parent bank's foreign sovereign risk exposure. A foreign sovereign risk exposure denominated in foreign currency is often assigned a much higher RWA than a domestic sovereign exposure denominated and financed in local currency.

Basel 2.5 thus produces significant increases in the RWA for trading exposures in the financial markets of EMSEs because they usually have greater volatility and lower global credit ratings. The report by the Regional Consultative Group for the Americas (FSB, 2014) asserts that capital requirements for some countries' sovereign bonds may increase as much as sixteen-fold as a consequence of Basel 2.5. This may thus exacerbate global banks' costs of trading exposures to EMSEs when domestic risk exposures are transformed into foreign ones particularly when global credit ratings are used. This increase in RWA takes place even when the sovereign positions are registered in the books of subsidiaries established in the same country that issues the sovereign debt and the sovereign debt is denominated and funded in the currency of the issuing country. Higher capital charges will also be particularly significant for sovereign domestic debt held by large subsidiaries of global banks because their risk positions may increase the capital requirements for concentration risks. The effect will be the increase in the cost of holding sovereign debt, especially in emerging market economies in which global banks have a material presence. In some cases this could accelerate the deleveraging process of global banks from overseas exposures. (4)

Basel 2.5 through its impact on the RWA for trading exposures may thus harm liquidity and create a less level playing field in the financial markets of EMSEs. Basel 2.5 could have a negative impact on liquidity in local financial markets through the increases in the trading book RWA of the local subsidiaries of foreign banks for their risk exposures in local sovereign debt. These increases would also discourage banks from taking advantage of arbitrage opportunities in local financial markets, decreasing their efficiency. This issue is very important for the development of efficient financial markets in EMSEs where global banks and their local subsidiaries are an important source of liquidity. The requirements associated with Basel 2.5 would also promote arbitrage between the trading and banking books. In particular, banks could wish to move risk exposures from the trading to the banking book as RWA for same risk exposures could be much higher when registered in the former than in the latter (Pepe, 2013). While all exposures in the trading book must be valued at market prices, banking book exposures can be valued at amortized cost. Hence, this shift of exposures, particularly of securities, from the trading to the banking book decreases the transparency of banks' financial situation.

In some EMSEs, as well as in advanced economies, the implementation of the credit valuation adjustment (CVA) risk charge is still under consideration. The CVA capital charge computes the amount required to cover the losses arising from marking to market the counterparty risk of banks' OTC derivative portfolios. BCBS observed that two-thirds of credit risk losses suffered by banks during the financial crisis in 2008 arose from CVA losses rather than actual defaults. (5) Strong industry pushback and fears from corporate and pension funds that their business costs would increase have caused regulators to create exceptions (EU framework includes exemptions of CVA risk to trades with corporates, sovereigns and pension funds) or postpone their decision.

In order to deal with the challenges of implementing the Basel 2.5 and III capital framework, international regulators should consider issuing specific guidance for the appropriate use of local and global credit ratings and the risk assessment of sovereign exposures (denominated and funded in local currencies) in foreign subsidiaries. When consolidating parent banks and subsidiaries' balance sheets and assigning RWAs, guidance is needed on practices regarding the risk weighting, at the consolidated level, of foreign subsidiaries' risk exposures--both the currency denomination of assets/liabilities and the legal differences between a parent bank's assets/liabilities in an overseas branch from those of a subsidiary should be taken into account. In addition, the regulators should re-evaluate the circumstances in which it may be appropriate to deduct an entity from regulatory capital rather than consolidate its exposures. The latter can be particularly important for structurally separated entities that are systemically important and have limited intra-group exposures.

BASEL III AND SOVEREIGN EXPOSURES

The recommendations of the previous section should not be viewed as a call for general acceptance of zero or close to zero risk weights for sovereign exposures. There is a case for a reconsideration of current capital treatment of sovereign debt and for making the framework globally more consistent instead. The Basel III framework continues to provide preferential treatment of sovereign exposures denominated and funded in domestic currencies. All sovereign exposures of this sort are allowed, at national discretion, to keep the same risk weight regardless of their ratings. Such treatment stems from the unique and central role of government bonds in modern financial systems. These are generally regarded as risk-free (ie, highly liquid, high-quality, HQLA) assets, thanks mainly to the high credibility of the state as issuer and the high quantity of debt issued. However, the preferential treatment brings a risk of reinforcing the links between sovereign and banking sector stresses, as became apparent in the recent global financial crisis, especially the crisis of the euro area. In addition, as jurisdictions may exercise national discretion to assign a zero-risk weight for domestic sovereign exposures under the new capital and liquidity frameworks, there could be further incentives for banks to hold domestic sovereign exposures.

The build-up of sovereign exposures on the balance sheet of banks beyond some point may become a relevant concern for authorities in EMSEs. In line with the development of domestic capital markets, sovereign debt in EMSEs is increasingly being funded in domestic currencies. However, domestic bond markets in a number of EMSEs are still under development, thus limiting the ability of banks to diversify their assets. This poses several challenges, including for the implementation of the LCR (see the section 'The Basel III liquidity framework'). Addressing banks' sovereign exposures is particularly pertinent for those EMSEs whose economies are dollarized, are members of currency blocks and/or issue a significant number of sovereign bonds denominated in foreign currency.

The approaches of international banks to recognizing sovereign risk differ (FSB, 2014). Some banks that apply the standardized approach use global ratings to calculate risk-weighted assets. Other banks using internal risk-based (IRB) models do not take into account whether the sovereign exposure is denominated and funded in local currency. As a consequence, international banks may give asymmetric treatment to comparable risks based on the entity in which the asset was booked. As explained in the previous section, home countries' and host countries' sovereign exposures thus may not always receive similar treatment. In the process of balance sheet consolidation, host country sovereign exposures often lose their place in the local lowest risk-weight category, while home country sovereigns maintain it. This creates inconsistencies when the credit quality of a given counterparty is evaluated both by the subsidiary and by the parent office. Applying different treatment to sovereign exposures booked in overseas subsidiaries creates a home country bias and makes the playing field in host countries less level.

Several principles can guide the approach aimed at limiting excessive accumulation of sovereign exposures spurred by market and regulatory incentives. While the authorities must recognize that banks have legitimate motivations to hold sovereign exposures on a relatively large scale, the approach needs to: (i) differentiate among types of sovereign debt held by banks; (ii) be tailored to country-specific circumstances; (iii) consider potential unintended consequences; and (iv) be supported and informed by a robust analytical framework.

A potential menu of approaches to address sovereign risk build-up includes: (i) sovereign risk capital buffers, such as incremental capital charges on sovereign exposures under Pillar 1 or concentration charges under Pillar 2; (ii) more robust and consistent cross-jurisdictional application of the Pillar 2 supervisory review process to address risks associated with excessive concentration; (iii) a limit on the size of sovereign debt exposures, possibly as a percentage of total assets or capital; and (iv) supply-side measures to promote sovereign risk diversification, such as vehicles or funds for sovereigns to collectively pool liabilities, as well as initiatives aimed at developing domestic and regional bond markets. The authorities in individual jurisdictions should not apply the measures to sovereign exposures unilaterally, unless these are identified as a significant threat to financial stability. The changes to regulatory treatment of sovereign exposures should preferably be considered and adopted on a globally consistent basis.

The international regulators should study differences in the implementation of the Basel framework for international banks and consider developing guidance in order to achieve consistent implementation of capital standards, and avoid arbitrage and asymmetric treatment of similar exposures. Consolidation practices and home country regulation should avoid a home country bias: sovereign exposures denominated and funded in local currency at overseas subsidiaries should receive the same treatment applied by head offices to sovereign exposures of the home country sovereign denominated in their domestic currencies and funded locally. At the same time, mechanistic use of CRA global sovereign ratings should be discouraged.

ADVANCED APPROACHES TO MODELLING FINANCIAL RISKS

One of the potential consequences of the move towards higher capital requirements of Basel III is that banks in EMSEs could move to the IRB approaches without being ready and respond to higher capital requirements by not revealing and recognizing all potential risks associated with their balance sheets. The higher requirements may create an incentive for banks to move to the use of the more advanced risk measurement techniques of Basel II in the hope of saving on capital by achieving lower implicit risk weights with the same balance sheets. (6) This would create pressure on supervisors to approve such practices even if a bank is not ready (eg by citing reputational concerns). Similarly, some banks may change their stringent approach to provisioning in a way that would lead to the creation of lower provisions relative to expected losses. Such a possibility is given by the diverse accounting approaches across jurisdictions and also by the discretion in applying the accounting rules (see Bikker and Metzemakers, 2005; Angklomkliew et al, 2009; Packer and Zhu, 2012). Both directions could put somewhat arbitrary cushions against expected and unexpected losses. In addition, an incorrect or manipulative use of IRB methods and accounting rules could weaken consistency and comparability due to excessive variation in risk measurement without better management of the underlying risks.

There is a need to guard against the risk of banks moving in haste to the IRB approach under improper incentives. The IRB approaches, if applied inappropriately, may enable banks to manipulate their RWA numbers to lower risk weights. Only the gradual development of risk-sensitive approaches to regulatory capital calculations and risk management can lead to better risk capture and adequate preparation by both banks and supervisors to address possible future challenges with building internal models. Authorities have to take a conservative and prudent view on reviewing and approving IRB applications by banks. Host supervisors should actively verify and approve the models developed by parent banks, taking into account the specific features of the host country market. The supervisors need to be able to validate highly technical mathematical models as well as to find a right balance of models with qualitative features of the IRB approach.

The relevance of concerns regarding manipulation with risk weights has been confirmed in advanced economies. In recent years, bank equity analysts have frequently remarked on the difficulty of understanding differences in risk-weighted assets and coverage of impaired assets by provisions both across banks and through time. The BCBS (2013b) study on the regulatory consistency of risk-weighted assets in the banking book drawing on supervisory data from more than 100 major banks found that up to three quarters of the considerable variation across banks in average RWAs for credit risk in the banking book is consistent with the spirit of the risk-based capital framework, that is, it can be explained by differences in the composition of banks' assets. The rest of the variation is, however, driven by diversity in banks and supervisory practices. Some of this stems from supervisory choices at the national level, due either to discretion permitted under the Basel framework or deviation in national implementation from Basel standards. The differences in practices also result from banks' choices under the IRB framework, that is, varying IRB approaches used by banks, conservative adjustments to IRB parameter estimates and differences in banks' modelling choices. In some cases, variations may also reflect differences in interpretation of the Basel framework.

Therefore, priorities for progressive movement to more sophisticated approaches within the Basel framework (eg IRB approaches) should be established. Supervisory authorities should ensure robustness, reliability and transparency of prudential outcomes from the adoption of Basel standards, including the Core Principles for Effective Banking Supervision (Basel Core Principles). In this context, Pillar 3 should be seen as a tool for meeting the needs of investors and counterparties. Decisions on the pace of the implementation would need to consider particular characteristics of banks and banking systems, as well as supervisory constraints. For example, some countries have considered the adoption of a more rules-based approach to Pillar 2 requirements as a way forward in the presence of legal frameworks that significantly limit supervisory powers. When considering the capital framework for smaller and less sophisticated banks, authorities should be aware that Basel standards are designed primarily for large international banks. However, when the majority of the banking sector is owned by these large institutions, supervisory authorities in EMSEs should build relationships with the home authorities of their largest banks and, upon agreement with the home authority, participate in relevant discussions on model validation within supervisory colleges.

In addition to proper calculation of risk-weighted assets, transparent and consistent accounting is crucially important for the robustness of prudential outcomes, even though it is not directly linked to the Basel standards. The recent financial crisis highlighted that having provisions commensurate with expected losses is one of the building blocks of resilience of the banking sector in particular. Provisioning is important not only because the provisions serve as a buffer against expected loan losses, but also because they provide significant information on how banks price credit risk. There are significant differences across jurisdictions as to the factual approach to provisioning. The desired state is that all banks regulated in line with Basel standards use the International Financial Reporting Standards (IFRS). This is not the case now, even though significant progress has been made in this area. Nevertheless, even such a state would not guarantee adequate provisioning for asset impairment. Further improvements in the IFRS that would limit, among other things, existing procyclicality, are crucially needed (Frait and Komarkova, 2013).

THE BASEL III LIQUIDITY FRAMEWORK

Basel III is the first accord that attempts to set a comprehensive quantitative framework for regulating the banks' liquidity (BCBS, 2010d). The LCR is designed to improve banks' resilience to short-term liquidity shocks through holding a reserve of HQLA. The NSFR should ensure that long-term assets are funded primarily by long-term, stable funding. Previously, national practices and experiences of liquidity requirements differed. Nevertheless, many countries, including some EMSEs, had developed approaches based on the concept of banks holding stocks of liquid assets to withstand stressed periods (for examples, see CGFS, 2010).

Implementation of the new liquidity framework will be challenging for some EMSEs. Different macroeconomic and financial environments in these countries, such as the lack of availability of diversified HQLAs, the higher share of foreign currency-denominated banking assets and liabilities, and unique characteristics of depositors, would pose specific challenges for implementing the LCR appropriately. In jurisdictions with limited availability of HQLAs, concentration risk, particularly to sovereign debt (a 'Level 1' asset), can easily emerge. The LCR requires banks to hold a diversified portfolio of HQLA that can be liquidated in a stress event to cover the outflow of liabilities, but the room for diversification will be limited in many EMSEs with limited access to other HQLA than sovereign debt. Furthermore, yields on HQLAs will typically be lower than other securities, implying a cost to banks' profit margins. Greater use of 'Level 2' HQLA, which are less traded in the markets, such as corporate debt, by banks could introduce greater volatility in market risk and increase exposure to credit risk.

Expanding the supply of HQLA through the alternative treatment in the framework would address the problem, at least partially. The LCR framework provides three options to address an insufficient supply of 'Alternative Liquid Asset' (ALA) treatment, with different pros and cons. The first option is to use a contractual committed liquidity facility (CLF) provided by the central bank. (7) This option has the advantage of avoiding incentives for banks to change their asset portfolios. However, calibration is challenging as it must balance the potential for banks' overreliance on the facility with making the cost of the CLF unduly prohibitive. Central banks must be able to honour a CLF commitment and consider carefully how such a facility would affect monetary policy operations. Foreign currency HQLA, Option 2, can be a practical solution when banks already hold substantial numbers of such assets, but a premium must be placed on strong management of foreign exchange risk. A strong currency risk control framework, including quantitative regulatory requirements such as net foreign open position limits, should be a pre-requisite. Types of eligible assets need to be limited and haircuts must be conservative, based on historical experiences during stressed periods. Additional use of lower-quality HQLA (Level 2) with a higher haircut, Option 3, may be viable when sovereign debt is scarce but capital markets are well developed. Although this option enables banks to diversify away from sovereign bonds and alleviate pressures from their prices, the true liquidity profile of such Level 2 assets, particularly during times of stress, needs to be assessed and conservative haircuts must be set. Furthermore, supervisors must guard against Level 2 assets crowding out Level 1 assets for higher yields the former typically have.

Introducing new instruments that are included in HQLA is a further option. An option considered by some advanced economies is the covered bond, which provided a stable source of funding during the recent crisis.

However, risks must be carefully assessed when introducing new instruments. For example, covered bonds encumber assets, potentially prejudicing depositors and unsecured creditors in the event of resolution; mitigation would be to set limits on total asset encumbrance or covered bond issuance. (8)

For some EMSEs, the LCR may increase foreign currency risk if banks meet LCR shortfalls in domestic currency with foreign currency assets. Currency convertibility in the LCR framework for dollarized countries needs further guidance. It is questionable, for example, whether it is prudent to allow surplus in dollar LCR to cover a shortfall LCR in the domestic currency and vice versa. Furthermore, while Basel III does not require the LCR to be met currency by currency, monitoring and reporting of relevant currencies forms part of the LCR framework and banks may feel under pressure to meet the LCR in all individual currencies, thus increasing demand for foreign currency HQLA. This may in turn affect the price and availability of such assets.

Banks in EMSEs generally rely heavily on deposits for funding, putting a premium on applying appropriate run-off rates to deposits. (9) While this funding pattern broadly reduces the HQLA requirement, close attention needs to be paid to the breakdown of deposits to ensure that low run-off rates are suitable and reflect local conditions. This means not only must the authorities set run-off factors that are appropriate in their jurisdiction but they also must be satisfied that banks are capable of distinguishing correctly between different types of deposit liability, taking account of followings. Probabilities of funding run-off in some countries could differ substantially from those assumed in the LCR framework. Typical examples could be smaller jurisdictions where non-resident deposits or cross-border mobility of deposits is a major feature. This heightens the need for national discretion in EMSEs in calibrating run-off rates for certain types of liability. The applicable run-off rates for deposits range from 3% to 100% depending on the stability of the deposit's characteristics, making it vital for banks to categorize funding accurately to generate a meaningful LCR figure. A bank must have systems that can distinguish the relevant criteria in its deposit base, such as identifying retail and small business deposits, tracking insured deposits from uninsured funds and distinguishing operational deposits from other wholesale deposits. Where supervisors doubt the banks' operational and systems capabilities, they should impose more conservative definitions and assumptions.

Enhanced liquidity requirements could affect the way international banking groups hold liquid reserves in their different levels of group structures (CGFS, 2010). There are concerns among authorities of EMSEs that the availability of group-level liquidity to foreign subsidiaries, including deposits placed by them to parent banks, would be affected by the implementation of the LCR. These authorities are also worried that the efforts by home supervisors to improve those groups' resolvability, including the preparation of recovery and resolution plans as well as application of structural measures on bank activities, could result in banks 'compartmentalizing' their different operations, which may weaken the ownership chain and the availability of group liquidity and capital support.

The LCR implementation demands careful planning and dedicated resources. Transition to the LCR, which is relatively more sophisticated than most existing Basel methodologies, could pose a substantial challenge for many countries. Authorities in these jurisdictions may wish to consider the following issues when implementing liquidity standards. Jurisdictions must determine the scope of LCR coverage. For internationally active banks in BCBS member jurisdictions, the LCR is mandatory. For the more advanced banks in EMSEs where a similar methodology already exists, there is considerable value in implementing the LCR and applying it to banks that have material cross-border activities. For jurisdictions where an LCR-like rule does not exist and cross-border activities are minimal, the aim should be to move to the LCR framework gradually to give banks time to improve their capacity. During this transition, consideration should be given as to whether the LCR parameters are sufficiently stringent or need to be tightened as appropriate to the local context.

Jurisdictions must also assess national discretions and ALA options in the context of their own systems. A first step is to understand the availability and characteristics of liquid assets and the liquidity characteristics of banks' sources of funding. The pros and cons of the ALA options must be carefully assessed. The flexibility the LCR framework offers in terms of the ALA and national discretions should enable an orderly transition based on careful consideration of quantitative impact study (QIS) information and stringent application of criteria for ALA treatment. Nevertheless, the Basel framework provides stringent criteria and processes for jurisdictions to be qualified for the ALA treatment, including periodic self-assessment and independent peer review. EMSEs should strive to adhere to these as much as possible. It is advisable for supervisors to monitor the LCR by currency irrespective of the importance of foreign currency in banks' balance sheets. Such information allows the supervisor to identify any potential currency mismatches and to consider the liquidity risk in foreign currencies. A QIS is needed to design the LCR appropriately for EMSEs. The QIS must provide granular data, such as numbers of different types of HQLA that banks hold, or banks ability to categorize deposits based on their stability. Fluent two-way communication mechanisms with the banks, such as workshops, are recommended to ensure that banks understand the standard, and authorities understand the banks' capacities so that adjustments to local standards, criteria, haircuts and run-off rates can be made where appropriate.

Further guidance by the regulators will be important to ease transition for EMSEs. Areas for further guidance include: the use of ALAs in countries with less developed capital markets, the treatment of currency convertibility in the LCR framework for dollarized economies and the exercise of national discretion in applying run-off rates for deposits. The regulators should further encourage home supervisors to reach understandings with international banking groups and host supervisors on group-wide liquidity management. Absent such understandings, including on the provision of centrally held liquidity to subsidiaries and branches, host supervisors may be compelled to require subsidiaries and branches to retain minimum liquidity at the local level to protect their national financial stability. (10) More generally, the international regulators should place a stronger emphasis on consolidated supervision by the home supervisor while maintaining close communication with the host supervisor, and encourage a wider sense of continuing responsibility for group-wide banking operations. Market-wide solutions like those adopted in the EU should be explored and could help home supervisors avoid retrenching and becoming more inward-looking. In addition, more flexible treatment of deposits placed by foreign subsidiaries should be allowed to reflect the nature of the underlying depositors, which would reflect the reality of the business model of these foreign subsidiaries and support the continued diversification of funding.

CONCLUSIONS

In this paper we have discussed the impact of changes to the capital and liquidity frameworks brought about by the Basel III accord on emerging market, developing and small economies (EMSEs). Even though we believe that Basel III will deliver significant benefits over the longer time horizon, our intention was to identify challenges, potential unintended consequences of the new rules and their adverse economic effects. Some of them are associated with the presence of branches and subsidiaries of multinational banks in EMSEs' banking sectors, and resulting home-host relations and conflicts.

We surveyed the analyses on the macroeconomic effects of Basel III framework and potential consequences of new capital and liquidity regulations, as well as issues associated with the introduction of macroprudential capital buffers. As to the specific areas, we looked at implications of new regulation of banks' exposures in trading book, regulatory treatment of sovereign exposures and potential risks associated with an accelerated move to advanced approaches to modelling financial risks.

We identified several key areas of concern for both banks and regulators. Potentially most pressing are increases in risk-weighted assets for exposures in trading books located at foreign affiliates of multinational banks. The effect will be the increase in the cost of holding sovereign debt, especially in emerging market economies, which could lead to partial deleveraging of global banks from overseas exposures. To mitigate such risks, both national and international regulators should promote consistent application of rules in various parts of multinational banks, especially the ones regarding local sovereign exposures denominated and funded in local currency, and assess the methods of consolidation practices regarding the risk weighting of foreign subsidiaries' exposures.

The challenges may also be generated by the need for capital replenishments and required capital deductions, especially in jurisdictions with weaker governance and less developed financial markets. The condition that regulatory capital instruments should be able to absorb losses in the event that the issuing bank reaches a point of non-viability may create governance issues, for example, when conversion brings in shareholders that may not be suitable. Coping with these issues requires strengthening legal and institutional arrangements to enable smoother issuance of capital instruments, and also encouragement of the provision of adequate supervisory powers.

Enhanced liquidity requirements may run against limited availability of truly liquid assets that at the same time are of high quality in a number of jurisdictions. In such jurisdictions, concentration risk, particularly to sovereign debt, deemed both liquid and of high quality, can easily emerge. The new regulation could also encourage groups to hold liquid reserves at the parent level. However, it may not always be clear when and how these reserves should be made available, while deposits placed at a parent bank by foreign subsidiaries could become subject to bail-in arrangements. The implementation of liquidity standards thus demands careful planning and proper assessment of alternative sources of high-quality liquid assets in the context of concrete jurisdictions. If such assets are sought abroad, special attention will have to be paid to currency mismatches and liquidity risk in foreign currencies.

Overall, when transitioning away from Basel I, EMSEs could usefully take into account some guiding principles. Capital requirements are only one part of a good supervisory framework. First, the effective implementation of the regulatory framework for capital definition, buffers and disclosures depends on sufficient powers and resources. In this sense, countries that have successfully implemented Pillar 2 and Pillar 3 of the Basel II framework would be better placed to implement Basel III as well. Compliance with Basel Core Principles for Effective Banking Supervision (BCP) should be a priority for all countries, advanced or developing. A lack of supervisory powers, capacity and independence in supervision is a greater hurdle to safe banking systems and effective supervision than solvency alone. Second, a progressive movement towards implementing elements of Basel II and III could be beneficial - these represent a higher level of requirements for both banks' risk management and supervisors' review. Implementation planning should start by building capacity to manage the process effectively. Decisions on the pace of the implementation would need to consider particular characteristics of banks and banking systems, as well as supervisory constraints. Some countries have considered the adoption of a more rules-based approach to Pillar 2 requirements as the way forward in legal frameworks where interpretation powers of supervisors may be limited. Third, there are elements of Basel III that could be implemented in Basel 1 countries, even if Basel II has not been implemented. While some requirements of Basel III regarding the denominator of the capital adequacy ratio are directly linked to the Basel II securitization framework, the enhanced definition of capital, the buffers and enhanced disclosures could be introduced without Pillar 1 of Basel II as a prerequisite. Ensuring that the capital base is of good quality and market discipline is functioning should be a supervisory goal independent of the capital regime adopted.

Acknowledgements

The authors would like to thank Rudy Araujo, Pascual O'Dogherty, Karl Cordewener, Ju Quan Tan, Jong Ku Kang, Tae Soo Kang, Samsiah Yunus, Lixing Zhang, Bryan Stirewalt, Michaela Erbenova, Christopher Wilson, Alejandro Lopez Mejia, John Aspden and Jan Kubicek. They note that the paper represents their own views and not necessarily those of the Czech National Bank. All errors and omissions remain entirely the fault of the authors. Part of the research behind this paper was supported by the Grant Agency of the Czech Republic within Project No. 13-08549S. The paper reflects the discussions of the workstream analytical and research group set up by the Basel Consultative Group of the Basel Committee on Banking Supervision.

REFERENCES

Admati, A, DeMarzo, P, Hellwig, M and Pfleiderer, P. 2011: Fallacies, irrelevant facts, and myths in the discussion of capital regulation: Why bank equity is not expensive. Stanford Graduate School of Business research paper no. 2065, Palo Alto, CA.

Aiyar, S and Jain-Chandra, S. 2012: The domestic credit supply response to international bank deleveraging: Is Asia different? IMF working paper 12/258, International Monetary Fund: Washington DC.

Angelini, P, Clerc, L, Curdia, V, Gambacorta, L, Gerali, A, Locarno, A, Motto, R, Roeger, W, Van den Heuvel, S and Vlcek, J. 2011: BASEL III: Long-term impact on economic performance and fluctuations. BIS working papers no. 338, Bank for International Settlements: Basel.

Angklomkliew, S, George, J and Packer, F.. 2009: Issues and developments in loan loss provisioning: The case of Asia. BIS Quarterly Review, Bank for International Settlements, Basel, http://www.bis .org/publ/qtrpdf/r_qt0912h.pdf.

Basel Committee on Banking Supervision. 2009: Guidelines for computing capital for incremental risk in the trading book. Bank for International Settlements: Basel.

Basel Committee on Banking Supervision. 2010a: Guidance for national authorities operating the countercyclical buffer. Bank for International Settlements: Basel.

Basel Committee on Banking Supervision. 2010b: Basel III: A global regulatory framework for more resilient banks and banking systems. Bank for International Settlements: Basel.

Basel Committee on Banking Supervision. 2010c: An assessment of the long-term impact of stronger capital and liquidity requirements. (LEI Report), Bank for International Settlements: Basel.

Basel Committee on Banking Supervision. 2010d: Basel III: International framework for liquidity risk measurement, standards and monitoring. Bank for International Settlements: Basel.

Basel Committee on Banking Supervision. 2013a: Report on the regulatory consistency of risk-weighted assets for market risk. Bank for International Settlements: Basel.

Basel Committee on Banking Supervision. 2013b: Report on the regulatory consistency of risk-weighted assets in the banking book. Bank for International Settlements: Basel.

Basel Committee on Banking Supervision. 2013c: The regulatory framework: Balancing risk sensitivity, simplicity and comparability. BIS discussion paper, Bank for International Settlements: Basel.

Bikker, J and Metzemakers, P. 2005: Bank provisioning behaviour and procyclicality. Journal of International Financial Markets, Institutions and Money 15(2): 141-157.

Calzolari, G and Loranth, G. 2001: On the regulation of multinational banks. Rivista di Politico Economica 91(4): 275-304.

Calzolari, G and Loranth, G. 2005: Regulation of multinational banks: A theoretical inquiry. ECB working paper series no. 431, European Central Bank: Frankfurt am Main.

Cerutti, E, Dell'Ariccia, G and Perfa, MSM. 2007: How banks go abroad: Branches or subsidiaries? Journal of Banking and Finance 31 (6): 1669-1692.

CGFS. 2010: Funding patterns and liquidity management of internationally active banks, Committee on the Global Financial System papers no. 39, Bank for International Settlements: Basel.

De Haas, R and Van Lelyveld, I. 2014: Multinational banks and the global financial crisis: Weathering the perfect storm? EBRD working paper no. 135, European Bank for Reconstruction and Development: London.

Drehmann, M, Borio, C, Gambacorta, L, Jimenez, G and Trucharte, C. 2010: Countercyclical capital buffers: Exploring options. BIS working papers no. 317, Bank for International Settlements: Basel.

European Banking Authority. 2013: Summary report on comparability and pro-cyclicality of the IRB approach. EBA Report, European Banking Authority: London.

Feyen, E, Kibuuka, K and Oetker-Robe, I. 2012: European bank deleveraging: Implications for emerging market countries. World Bank economic premise, no. 79, World Bank: Washington DC.

Fiechter, J, Otker-Robe, I, Ilyina, A, Hsu, M., Santos, A and Surti, J. 2011: Subsidiaries or branches: Does one aize fit all? IMF staff discussion paper, International Monetary Fund: Washington DC.

Financial Stability Board. 2014: Regional consultative group for the Americas report on the effect on host countries of balance sheet consolidation and risk management practices by global banks. Financial Stability Board, Bank for International Settlements: Basel.

Frait, J and Komarkova, Z. 2013: Loan loss provisioning in selected European banking sectors: Do banks really behave in a procyclical way? Czech Journal of Economics and Finance 63 (4): 308-326.

Gersl, A and Seidler, J. 2011: Excessive credit growth as an indicator of financial (in)stability and its use in macroprudential policy. Financial stability report 2010/2011, Czech National Bank: Prague.

Hahm, JH, Mishkin, F, Shin, HS and Shin, K. 2012: Macroprudential policies in open emerging economies. NBER working paper no. 17780, National Bureau of Economic Research: Cambridge, MA.

Herman, K and Rai, K. 2010: The global credit crunch and foreign banks' lending to emerging markets: Why did Latin America fare better? IMF working paper 10/102, International Monetary Fund: Washington DC.

Kashyap, A, Stein, J and Hanson, S. 2010: An analysis of the impact of 'substantially heightened' capital requirements on large financial institutions. Working paper, Harvard Business School, Boston, MA.

MAG. 2010: Assessing the macroeconomic impact of the transition to stronger capital and liquidity requirements, Macroeconomic Assessment Group established by Financial Stability Board and the Basel Committee on Banking Supervision, Bank for International Settlements: Basel.

Mayes, D and Granlund, P. 2008: Polar approaches to regulating cross-border banking: Nordic and antipodean routes to financial stability. European Business Organization Law Review 9(01): 63-95.

Miles, D, Yang, J and Marcheggiano, G. 2011: Optimal bank capital. External MPC Unit discussion paper no. 31, Bank of England: London.

Moreno, R. 2011: Policymaking from a 'macroprudential' perspective in emerging market economies. BIS working paper no. 336, Bank for International Settlements: Basel.

Packer, F and Zhu, H. 2012: Loan loss provisioning practices of Asian banks. BIS working paper no. 375, Bank for International Settlements: Basel.

Pepe, G. 2013: Basel 2.5: Potential benefits and unintended consequences. Questioni di Economia e Finanza no 154, Banca D'Italia: Rome.

Slovik, P and Cournede, B. 2011: Macroeconomic impact of Basel III. OECD economics department working papers no. 844, Organization for Economic Co-operation and Development: Paris.

World Bank. 2013: Global Financial Development Report 2013: Rethinking the role of the state in finance. Box 2.5 Impact of the Basel III Implementation in Developing Economies. World Bank: Washington DC.

(1) The issues regarding the functioning and regulation of multinational banks are discussed in Calzolari and Loranth (2001). How regulatory intervention depends on the liability structure and insurance arrangements for non-local depositors is investigated in Calzolari and Loranth (2005).

(2) See, for example, Gersl and Seidler (2011). The BCBS itself pointed out that aggregate private sector credit-to-GDP gap might not be a good indicator for all jurisdictions (BCBS, 2010a).

(3) Press release from 10 June 2010 'Adjustments to the Basel II market risk framework announced by the Basel Committee'.

(4) The issue of deleveraging is covered, for example, in Aiyar and Jain-Chandra (2012), Feyen et al. (2012) and Herman and Rai (2010).

(5) 'During the financial crisis, however, roughly two-thirds of losses attributed to counter-party credit risk were due to CVA losses and only about one-third were due to actual defaults', Basel Committee on Banking Supervision, Press release, 1 June 2011.

(6) Such risk has been studied intensively by regulatory authorities in recent years. (See, eg, BCBS, 2013a, 2013b, 2013c; EBA, 2013).

(7) Australia and South Africa have introduced this option.

(8) Covered bonds have been introduced in countries such as Australia, Belgium and Italy, among others.

(9) For example, banks in Malaysia, the Philippines and Saudi Arabia enjoy very high levels of deposits, over 80% of their total funding.

(10) There is some evidence that during the last crisis, parent institutions were not, in a number of cases, a particular source of strength for their affiliates (De Haas and Van Lelyveld, 2014).

JAN FRAIT [1,2] & VLADIMIR TOMSIK [2,3]

[1] University of Finance and Administration, Estonska 500, Prague, CZ 10100, Czech Republic. E-mail: [email protected]

[2] Czech National Bank, Na Prikope 28, Prague, CZ 11503, Czech Republic. E-mail: [email protected]

[3] Newton College, Rasinova 2, Brno, CZ 50200, Czech Republic.