Mars: Treasury Supervisor

februari 8, 2017 in Job on the spot, Vacatures by onze redactie

Location: Veghel
Type: Fulltime

Read the rest of this entry →

European banking regulation and the Eurozone crisis

maart 28, 2012 in Cash Management, Publicaties, Treasury Algemeen by onze redactie

European banking regulation and the Eurozone crisis

Ben S Bernanke: Community banking

februari 21, 2012 in Publicaties, Risico management, Treasury Algemeen by onze redactie

Ben S Bernanke- Community banking

Building CVA on top of an existing risk infrastructure

november 29, 2011 in Columns-Artikelen by Michel Dorval

The potential cost of doing business with certain counterparties is now a significant concern for anyone trading in the financial markets. In the past, the valuation of counterparty credit risk (CCR) was largely ignored, thanks to the relatively small size of derivatives exposures and the high credit rating of the counterparties involved – in general, other highly rated financial institutions. As the size of derivatives exposure increased and the credit quality of the counterparties fell in the wake of the 2008 crisis, however, the valuation of counterparty credit risk could no longer be assumed to be negligible and had to be priced in, more in particular credit value adjustment (CVA).

A CVA captures the counterparty default risk inherent in over-the-counter derivatives portfolios. In a sense, the CVA is similar to loss reserves made on loan portfolios; on the other hand CVA is a highly volatile figure that depends directly on fluctuating daily market prices.

This article explains the ‘CVA landscape’ like the drivers and the different approaches that banks can adopt in response to these drivers.  It provides practical help for banks looking to implement CVA projects effectively taken into account the implications of each of these strategies.

CVA appears in several different contexts. The original context was in managing P&L volatility arising from counterparty default risk in large OTC books. Since then, the CVA concept has been taken up by accounting standards organizations (specifically, with the development of IFRS 9 and ASC820 standards for fair value), as well as forming part of the requirements for additional regulatory capital, as put forward in the Basel III framework.

Since the current Basel II counterparty credit risk rules cover only default risk and no CVA risk, the Basel Committee on Banking Supervision introduced in the Basel III framework a new capital charge for potential mark-to-market losses associated with any deterioration in the creditworthiness of a counterparty.  These new guidelines put forward both a standardized and advanced CVA charge.

Following the 2008 credit crunch, front offices realized that better quantification, pricing, and management of their counterparty credit risk was crucial because CVA losses dominated default losses during the crisis. Some banks created specific CVA desks that managed CVA P&L and collected charges from the originating desks in return for insulating them against counterparty default losses. The total CVA book may represent a very large part of the bank’s P&L, making it important to hedge the overall CVA and so avoid CVA uncertainty having a negative impact on bank profitability.

It is up to each bank to decide the level of CVA management it will try to attain in both short and long term.

To this end, it may be helpful to categorize a bank’s CVA strategy into four broad stages, with the sophistication and cost increasing at each stage:

1. Measure: A CVA measuring capability is created to calculate and aggregate CVA risks. Accounting and risk management departments will be the principal users of this function. This stage fulfils compliance obligations under accounting and regulatory standards.

2. Advise: In addition to measuring CVA, the bank will advise its trading departments on CVA-related risks. For example, position limits may be set to include CVA, or traders may be given minimum spreads to charge on a counterparty by counterparty basis.

3. Hedge: At this level the CVA is transferred from the trading desks to a CVA desk, perhaps through a one-time charge to the trading desk. The CVA desk is then responsible for managing the CVA P&L and, for example, for hedging it through the CDS market.

4. Trade: Here the CVA desk becomes a profit centre. The bank is not only hedging its own CVA risks but is also actively taking CVA positions

The choice will depend on the size of the bank and the scope of its derivatives book; the strategic direction the bank is following; and the regulatory and accounting standards in place. So banks with only limited derivatives activity may opt to stay at a compliance level and restrict investment in CVA measurement to whatever is required to be in line with the accounting and supervisory regime in place. A larger derivatives player on the other hand will transfer CVAs from individual trading departments to a consolidated CVA desk that will hedge or even trade CVA.

Having previously invested in the capabilities necessary for calculating economic and regulatory capital, most banks will already have in place all, or at least parts, of the different elements required to build a CVA solution. Unfortunately, these elements might be (and usually are) dispersed across different departments, where they serve specific purposes. A more consolidated approach is required for CVA. These elements can be broadly grouped under the headings of data, analytics and reporting.

Most of the data will already be present in the bank as it is standard input to current platforms used to calculate market and counterparty credit risk. The challenge, however, lies in consolidating and normalizing this data so that it can be used for a centralized CVA computation.

The securities data is usually available from the front-office trade capture and pricing systems and may already have been consolidated into risk management systems to calculate Value-at-Risk (VaR) for market risk or Potential Future Exposure (PFE) for credit risk.

The static data required is generally the same as that used by limit management solutions. Market data, such as yield curves, equity prices, FX rates and volatilities, can be sourced from trading and risk management systems.

Credit risk data, such as loss given default or recovery rates, ratings and probability of defaults, is entered into systems for the calculation of economic or regulatory capital, particularly if the bank is already using an internal ratings approach for regulatory capital.

In terms of analytics a complete CVA solution could potentially cover different functions like beside an EPE engine also components like calibration, wrong way risk and calculation of sensitivities to support hedging.

In terms of methodology, the requirements of CVA overlap with those required to estimate PFE in many respects. CVA also entails a simulation of the future evolution of market data, deal pricing on future dates along these paths and aggregation, while incorporating the effects of netting and collateral agreements. Many of the challenges are the same: the performance of portfolio simulation, portfolio netting and collateral agreements modeling; a need for aging and reinvestment strategies; provision for the rapid pricing of complex structured derivatives, and so on.

The common use of PFE is to compute exposures that feed into limit management systems and regulatory capital calculations, where the bank has approval for the Internal Model Method (IMM)[1]. Calibration for PFE is therefore performed principally on historical market data to capture through-the-cycle risk. Regulatory requirements specify three years of past history, with an additional calibration over a period of significant stress for the bank[2]. The same calibration could arguably be applied to compute CVA in the context of risk management. If, on the other hand, the purpose is to calculate CVA for trading and hedging or fair value accounting, then calibration needs to be implied from current market data (also referred to as risk-neutral pricing).

The CVA function is generally also tasked with checking for wrong way risk. Wrong way risk occurs when the exposure to the counterparty increases at the same time as the counterparty’s credit quality deteriorates. This correlation between exposure and credit quality is difficult to express as a model. Academics and practitioners have proposed various models for wrong way risk, but while these papers explain interesting relationships they are not general enough; hence, a common practice does not yet exist for detecting wrong way risk.

The CVA function also needs to support pre-trade CVA inquiries. This may be handled in different ways: approximated and delivered as guidance or by means of an exact computation. The latter needs to be performed rapidly, but it is worth noting that only the simulation of the new deal and aggregation with the previous deals in the same netting set need to be performed. All other values can be reused from a larger overnight batch.

While CVA is computed on a netting set by netting set basis and the CVA contribution of different netting sets is additive, CVA must also be allocated back to the transaction level. The contribution of a granular level to the total CVA can be based on different mathematical definitions (marginal, incremental, component, etc.). Note that this is generally different from the additional CVA that the counterparty would be charged at the time the deal was done.

Hedging requires a vast number of sensitivities covering credit risk, other underlying market variables, volatilities, and correlations. While the credit calculations may be quite inexpensive to calculate, most other sensitivities will require multiple Monte Carlo simulations to be run. The efficient generation of Monte Carlo-based sensitivities is therefore critical to this process.

The following table summarizes the links between the four broad stages and the functionalities discussed.

Any decision on reporting capabilities must be made with user profiles in mind. A CVA desk or Risk Management team would tend to be primarily interested in aggregated CVA figures. They would, however, also need drilldown capabilities to support the validation of the figures and to enable them to answer requests from the trading desks to approve new deals.

Aggregated views usually follow counterparty and instrument hierarchies and must show the netting and collateral agreements that are in place in a transparent manner. Traders are focused on their desk’s activity and are primarily interested in pricing (which also means knowing how much to add as CVA), in current CVA amounts, in details of any defined limits, in receiving guidance about which counterparties are favored or to be avoided and in whether or not a new transaction will ‘pass’.

Technical requirements depend on the size of the bank’s OTC derivatives operation and the scope of its CVA strategy. The number of prices that need to be simulated and the consequent amount of data that needs to be handled depends on the number of Monte Carlo paths, the number of simulation time steps and the number of transactions.  If CVA is calculated purely for compliance with accounting and supervisory requirements, then a regular daily batch simulation is sufficient. Should the bank require fast, intra-day simulations to quantify the impact of new deals, then it needs to bring the CVA computation closer to the front office. A solution to deliver this capability could mean having separate engines and data stores, fed with overnight results from a centralized CVA computation, but additionally allowing quick incremental CVA calculations and re-aggregations on netting set basis.

In conclusion, existing risk infrastructures will be a good starting point on which to build a CVA capability. What this article has aimed to show, however, is where the gaps may be and where additional work will need to be done, in line with the stance adopted by each bank towards CVA, as compliance necessity or potential profit generator.

Best regards,

Michel Dorval
Business Specialist Credit and Market Risk EMEA Region
Michel.dorval@thomsonreuters.com

&

Ulrich Schanz
Senior Functional Architect
Ulrich.schanz@thomsonreuters.com

Thomson Reuters

[1] BIS , Basel II : International Convergence of Capital Measurement and Capital Standards – A Revised Framework – Comprehensive Version – June 2006
[2] BIS , Basel III : A global regulatory framework for more resilient banks and banking systems – December 2010 (rev June 2011)

Basel III: Financial Stability Board (FSB)

oktober 27, 2011 in Columns-Artikelen by George Lekatis

The Financial Stability Board (FSB) is seeking comments on its Consultative Document on Effective Resolution of Systemically Important Financial Institutions. This Consultative Document contains a comprehensive package of proposed policy measures to improve the capacity of authorities to resolve systemically important financial institutions (SIFIs) without systemic disruption and without exposing the taxpayer to the risk of loss, and a time line for their implementation.

Resolution powers and tools

1. Key Attributes of Effective Resolution Regimes.

This sets out the powers and tools that all jurisdictions’ regimes for resolution of financial institutions should have to be effective, including for resolution of cross-border SIFIs.

2. Bail-in within Resolution.

This sets out proposed essential elements of a bail-in regime to enable creditor-financed recapitalisation of financial institutions.

Cross-border arrangements

3. Institution-specific Cross-border Cooperation Agreements.

This sets out proposed minimum common elements of institution-specific cooperation agreements amongst relevant resolution authorities to facilitate resolution of a cross-border firm.

Planning for resolution

4. Resolvability Assessments.

This sets out a proposed framework to be used for assessing the resolvability of a SIFI, taking into account the structure of the firm and the resolution regimes of the jurisdictions within which it operates, and which will inform Recovery and Resolution Plans.

5. Recovery and Resolution Plans (RRPs).

This sets out a proposed framework and contents of RRPs, which will be mandatory for global SIFIs (G-SIFIs).

Removing obstacles to resolvability

6. Measures to Improve Resolvability.

This seeks comment on actions to remove obstacles to resolution arising from complex firm structures and business practices, in particular obstacles that arise from fragmented information systems, intra-group transactions, reliance on service providers and global payment operations.

Discussion notes

To help inform its final recommendations, the FSB is also seeking comment on the two following notes for discussion.

These two notes reflect the preliminary views of the FSB and are being published as part of the Consultative Document to facilitate public discussion of the issues:

7. Creditor hierarchy, depositor preference and depositor protection in resolution.

This sets out the policy issues surrounding whether or not greater convergence across jurisdictions in the ranking of creditors’ claims, in particular in the treatment of deposit claims, is desirable.

8. Conditions for imposing temporary stays.

This note discusses the possible conditions under which a temporary suspension of contractual early termination rights should apply to support implementation of certain resolution tools.

The FSB invites comment on all above documents and the questions raised in the Consultative Document by Friday, 2 September 2011. Responses should be sent to the following e-mail address: fsb@bis.org.

Responses will be published on the FSB’s website unless respondents expressly request otherwise.

The FSB will revise the documents, including taking into account comments received, and will submit them, as part of its overall recommendations to address moral hazard posed by SIFIs, to the G20 Leaders Summit in Cannes on 3-4 November 2011.

Overview

The disorderly collapse of Lehman Brothers in September 2008 provided a sharp and painful lesson of the costs to the financial system and the global economy of the absence of powers and tools for dealing with the failure of a SIFI.

Lehman Brothers was the last SIFI allowed to fail during the last financial crisis.

All other SIFIs at risk were supported by public capital injections, asset or liability guarantees, or exceptional liquidity measures undertaken by central banks.

While this was necessary for economic and financial stability reasons, public bail-outs placed taxpayer funds at unacceptable risks and has increased moral hazard in a very significant way.

A recent stock-take undertaken by the Basel Committee on Banking Supervision (BCBS) of progress in implementing its Recommendations on Cross-border Bank Resolution of March 2010 shows that while some jurisdictions have enacted or are considering legislative changes, many jurisdictions continue to lack important resolution tools.

The report underlines the need to accelerate reforms of domestic resolution regimes and tools and of frameworks for cross-border enforcement of resolution actions.

At their Summits in Pittsburgh, Toronto and Seoul, the G20 Leaders asked the FSB to set out more effective arrangements for resolution of SIFIs.

The Annexes to this document set out the proposed policy recommendations, which are described in the following pages.

They comprise four key building blocks:

- Strengthened national resolution regimes that give a designated resolution authority a broad range of powers and tools to resolve a financial institution that is no longer viable and there is no reasonable prospect of it becoming so.

- Cross-border cooperation arrangements in the form of bilateral or multilateral institution-specific cooperation agreements, underpinned by national law, that will enable resolution authorities to act collectively to resolve cross-border firms in a more orderly, less costly way.

- Improved resolution planning by firms and authorities based on ex ante resolvability assessments that should inform the preparation of RRPs and that may, if necessary, require changes to individual firm structures and business practices to make them more effectively resolvable.

- Measures to remove obstacles to resolution arising from fragmented information systems, intra-group transactions, reliance on service providers and the provision of global payment services.

Legislation or regulatory changes will be required in many jurisdictions to implement these recommended measures.

Moreover, ensuring that financial institutions are resolvable under the current resolution regimes will require a reorientation of the supervision of SIFIs.

The FSB is also publishing two discussion notes for public comment on the pros and cons of greater convergence in creditors’ hierarchy, depositor preference and depositor protection in resolution and on the possible introduction of a brief stay on contractual early termination rights upon entry into resolution to support the implementation of resolution measures.

The measures to improve resolution regimes and tools set out in this consultative document represent a “bookend” to the FSB’s policy framework for addressing the systemic and moral hazard risks associated with SIFIs that are “too-big, too-complex and too-interconnected-tofail”.

The other “bookend” of the FSB’s policy framework is a requirement that global SIFIs (G-SIFIs) hold additional loss absorption capacity, as set out for banks in a separate consultative document from the BCBS released today.

The framework also comprises requirements for more intensive and effective supervisory oversight of SIFIs, as set out in the FSB’s November 2010 report report, and improvements to financial market infrastructures (FMIs) both to strengthen their robustness and reduce counterparty exposures, so as to reduce systemic contagion from a SIFI failure.

Many countries entered this crisis without a proper resolution regime, and no country had a regime that could cope with failing SIFIs.

Where effective resolution tools existed, these did not address the cross border dimension or obstacles stemming from within firms themselves.

This meant that proper market discipline was not in place in the years preceding the crisis and made the handling of the crisis more difficult.

The G20 called on the FSB to propose actions to address these challenges.

These proposed policy recommendation are offered for public consultation ahead of finalising the recommendations for the G20 Leaders in November.

Their effective implementation would entail changes in laws and regulation, supervisory practice and cross-border cooperation as well as within firms.

Proposed policy recommendations

Effective resolution regimes

A national resolution regime should provide the authorities with the tools to intervene safely and quickly to ensure the continued performance of the firm’s systemically important functions.

It should ensure prompt payout or transfer of insured deposits and prompt access to transactions accounts as well as to segregated client funds, wherever they are located.

It should enable the transfer or sale of viable portions of the firm while apportioning losses, including to unsecured and uninsured creditors, in a manner that is fair and predictable and so avoids panic or destabilisation of financial markets.

Need for a special national resolution regime for financial institutions

Corporate liquidation procedures are not well suited to deal with the failure of major banks and other financial institutions.

Such procedures freeze an institution’s balance sheet, typically in multiple jurisdictions, preventing access to the funds needed to manage its positions and to the assets and funds to which counterparties have claims.

This rapidly destroys the value of the SIFI’s balance sheet assets, including from fire sales, the tying up of liquidity and multiple, prolonged legal proceedings.

A resolution regime is therefore needed that is better tailored to the problems posed by the balance sheets and activities of major financial institutions than are corporate liquidation procedures.

An effective national resolution regime should provide a broad range of options to resolve a financial institution that is no longer viable.

It needs a designated administrative authority with a statutory mandate to promote financial stability in the exercise of its resolution powers.

This resolution authority should have the expertise, resources, capacity and operational independence consistent with their statutory responsibilities to exercise those powers, including for large and complex institutions such as SIFIs.

And just as is the case for supervisors, the law should provide for legal protection against lawsuits for actions or omission made while discharging their duties in good faith. It should be able to act with the necessary speed.

In those jurisdictions where a court order is required, it should consider any possible delay in its resolution planning process.

If more than one authority has responsibilities in the domestic resolution process, their respective powers and cooperation mechanism should be clear, and a lead authority should be identified to coordinate the resolution process of a group with multiple entities in the jurisdiction.

Statutory financial stability objectives

A resolution authority should have the powers and tools to meet the following key objectives:

- to preserve those of the SIFI’s operations that provide vital services to the financial system and the wider economy, which would cause system-wide damage if lost;

- to avoid unnecessary loss in value of financial assets and contagion (direct and indirect) to other parts of the financial system; and

- to ensure that losses are borne by those with whom the risks properly reside – first shareholders, and unsecured and uninsured creditors – rather than taxpayers.

A resolution regime needs to credibly be able to achieve these objectives if financial stability is to be protected and market discipline and incentives are to operate effectively.

Any resolution involves the distribution of losses but these losses are generally much smaller under orderly resolution than under disorderly liquidation.

Resolution tools

Resolution tools to preserve the viability of a firm’s systemically important functions basically fall into three types:

- sale of the entire firm (or at least of all its viable activities) as an ongoing business to a new owner;

- separation and eventual sale of functions that are systemically important or have franchise value as a separate operation while the residual parts of the firm are wound down (or alternatively carving out and transferring the bad assets to a separate asset management vehicle);

- recapitalisation of the firm by restructuring its liabilities.

Resolving a firm in a sustainable way is likely to take time, particularly given the complexities of the businesses of SIFIs.

An interim solution, such as a ‘bridge bank’ (or a ‘bridge company’ more generally for non-banks) , may therefore be needed to maintain systemically important operations, including the funding for them, while a more permanent resolution is being sought.

Meanwhile, the bad assets in the financial institution’s balance sheet will need to be run down, while avoiding a destructive fire sale.

Any mechanism for addressing a firm’s assets and the associated allocation of losses while it is resolved will need to:

- allow authorities to take control of the firm within resolution, replacing management and directors if necessary;

- facilitate the continuity of essential financial functions by allowing for their transfer of the underlying financial contracts that support them to a sound third party or a bridge company;

- give the resolution authority all powers necessary to operate and resolve the firm, including powers to terminate contracts, continue or assign contracts, purchase or sell assets, and take other actions necessary to restructure or wind down the firm’s operations; and

- respect the hierarchy of claims that would apply in a liquidation, and ensure that no creditors are worse off than they would be in liquidation, so as to preserve creditors’ legal rights.

Legal capacity to enable cross-border coordination of resolution

Cross-border resolution is impeded by major differences in national resolution regimes, absence of mutual recognition to give effect to resolution measures across borders, and lack of planning for handling stress and resolution.

The complexity and integrated nature of many firms’ group structures and operations, with multiple legal entities spanning national borders and business lines, make rapid and orderly resolutions of these institutions under current regimes virtually impossible.

Legislative changes are likely to be needed in many jurisdictions to ensure that resolution authorities have resolution powers with regard to all financial institutions operating in their jurisdictions, including the local branch operations of foreign institutions.

Cross-border cooperation and effective pre-planning of resolution will be difficult if not impossible if the authority over failed institutions, including foreign bank branches, resides with the courts.

As part of its statutory objectives, the resolution authority should duly consider the potential impact of its resolution actions on financial stability in other jurisdictions.

It should have the legal capacity to cooperate and coordinate effectively with foreign resolution authorities, to exchange information in normal times and in crisis, and to draw up and implement RRPs and cooperation agreements on an institution-specific basis.

An international standard for effective resolution regimes

The Key Attributes of Effective Resolution Regimes are intended to address these failings.

They set out the features that all resolution regimes should have in order adequately to mitigate the disruption from the failure of a financial institution and reduce moral hazard.

These include the features necessary for cross-border cooperation and the requirements to improve authorities’ and firms’ readiness for resolution.

In the FSB’s view, in order to raise the effectiveness of resolution around the world in a consistent way, the Key Attributes should form an international standard.

This will entail that jurisdictions’ implementation of the Key Attributes standard will be subject to assessments under the IMF/WB Financial Sector Assessment Program.

Any further sector-specific operational guidance by individual standard-setting bodies should be consistent with this overall framework.

Scope of application

The objectives set out in the Preamble of the Key Attributes, as well as many of the attributes themselves, apply to financial institutions of all sizes that could be systemically significant or critical in particular circumstances; any ailing financial institution that can cause contagion and disruptive effects on financial markets therefore should be subject to the type of resolution regime set out here.

Yet, the crisis response needs to be tailored to the specific nature of the firm’s activities and to sectoral differences.

It is important that resolution regimes provide a wide range of tools and the flexibility to apply them on a case-by-case basis to achieve an effective resolution.

Not all resolution powers set out in the Key Attributes are suitable for all sectors and all circumstances.

For example, to the extent that insurers conduct activities which are bank-like, the application of banking sector resolution tools to such activities rather than to the insurer as a whole or to its core traditional insurance business may be appropriate.

The FSB will be working with the CPSS, IAIS, and IOSCO to develop sectorspecific guidance for the application of its framework to non-bank SIFIs, including insurance companies, financial infrastructures and other financial institutions.

Questions for public consultation

1. Comment is invited on whether Annex 1: Key Attributes of Effective Resolution

Regimes appropriately covers the attributes that all jurisdictions’ resolution regimes

and the tools available under those regimes should have.

2. Is the overarching framework provided by Annex 1: Key Attributes of Effective

Resolution specific enough, yet flexible enough to cover the differing circumstances of different types of jurisdictions and financial institutions?

Bail-in powers

The paper on Bail-in within Resolution sets out the essential elements of statutory powers within a special resolution procedure and possible contractual provisions to achieve a creditor-financed recapitalisation of systemically vital functions of an ailing financial institution.

Such powers enable the resolution authority to write-down or convert into equity unsecured and uninsured claims, with a view to maintaining continuity of systemically vital functions, by either recapitalising the entity providing these functions, or, alternatively, capitalising a newly established entity or bridge institution to which these vital functions have been transferred following closure of the residual firm.

Resolution authorities should have bail-in powers within resolution to implement at least one of the above mechanisms.

The existence of statutory bail-in within resolution tools does not prevent firms from issuing instruments that write-off or convert contractually, nor do they prevent national authorities from requiring them.

It may create incentives for firm to issue such contractual instruments which might reinforce the capacity of firms to recover from distress without going into resolution.

Where, at the point of entry into resolution, an institution has contractual instruments with write-off or conversion features outstanding, a contractual instrument that had not been previously written-off or converted will be written-off or converted according to the contractual terms and conditions of the instrument upon entry into resolution but before the application of bail-in within resolution or other powers by the resolution authority.

A contractual instrument that, prior to entry into resolution, has already been written-off or converted upon activation of a contractual trigger would be subject to a subsequent application of bail-in powers upon entry into resolution.

The objective of bail-in is to reduce the loss of value and the economic disruption associated with insolvency proceedings for financial institutions, yet ensure that the costs of resolution are borne by the financial institutions’ shareholders and unsecured creditors.

The FSB proposes that authorities put in place statutory bail-in powers within their resolution regimes as a complement to other resolution tools.

Bail-in powers could be activated alone but most likely would be used in combination with other resolution tools.

The capacity to bail-in creditors would enhance resolution options and foster market discipline by countering the expectation that public funds will be used to support failing financial institutions.

Resolution authorities should have the statutory power, but not the obligation, to apply a bail-in within resolution.

The legislation giving the resolution authority statutory bail-in powers should provide clarity and certainty as regards the authority triggering entry into resolution; the process and the threshold conditions under which bail-in, and other resolution tools, could be used; the relevant consequences for capital providers and creditors; and the scope of liabilities covered by the bail-in powers.

It is desirable that divergence is limited across countries.

In acting quickly and seeking to ensure sufficient resources for either restoration to viability, or orderly resolution, authorities may impose haircuts or write-downs that turn out to be greater than needed.

To address these situations, authorities therefore should have in place mechanisms for compensating the holders of bailed-in claims, or written-off equity when the amount of actual losses is finally determined, e.g., by the issuance of warrants.

As a general principle, bail-in within resolution should be initiated by the home authority with respect to debt issued by the parent firm in resolution (and/or subsidiaries in resolution in the jurisdiction of the parent).

Where subsidiaries issue bail-in instruments, host authorities should be able to exercise bail-in at the subsidiary level.

Recognizing that the exercise of bailin powers could result in a change of the ownership structure, host authorities should consult with the home authorities and, to the extent possible, in the CMG, and satisfy themselves that the subsidiary is not viable, that support from the group is not available and that no alternative group-wide solution would achieve a more favourable outcome from a domestic and crossborder

financial stability perspective.

Authorities should regularly monitor whether firms’ balance sheets contain a sufficient quantum of liabilities covered by bail-in powers within resolution to facilitate orderly resolution.

Questions for public consultation

3. Are the elements identified in Annex 2: Bail-in within Resolution: Elements for

inclusion in the Key Attributes sufficiently specific to ensure that a bail-in regime is

comprehensive, transparent and effective, while sufficiently general to be adaptable to

the specific needs and legal frameworks of different jurisdictions?

4. Is it desirable that the scope of liabilities covered by statutory bail-in powers is as

broad as possible, and that this scope is largely similarly defined across countries?

5. What classes of debt or liabilities should be within the scope of statutory bail-in

powers?

6. What classes of debt or liabilities should be outside the scope of statutory bail-in

powers?

7. Will it be necessary that authorities monitor whether firms’ balance sheet contain at all times a sufficient amount of liabilities covered by bail-in powers and that, if that is not the case, they consider requiring minimum level of bail-in debt ?

If so, how should the minimum amount be calibrated and what form should such a requirement take, e.g.,:

(i a certain percentage of risk-weighted assets in bail-inable liabilities, or

(ii)a limit on the degree of asset encumbrance (e.g., through use as collateral)?

8. What consequences for banks’ funding and credit supply to the economy would you

expect from the introduction of any such required minimum amount of bail-inable liabilities?

Cross-border cooperation

The recent crisis was made considerably worse by obstacles to the ability of home and host authorities to cooperate in the resolution of SIFIs.

Some of these obstacles are legal barriers, and a legally binding international treaty would be a comprehensive means of addressing this for the global good.

Although an internationally agreed model law exists that addresses crossborder cooperation in corporate insolvencies , there is no immediate prospect of an equivalently formal multilateral agreement addressing the set of issues raised in the resolution of financial institutions.

In its absence, bilateral or multilateral cooperation agreements are needed, setting out how those jurisdictions most affected will cooperate over the resolution of individual firms, both in the planning phase and during a crisis itself.

Lack of adequate tools for cross-border resolution

Resolution regimes and tools need to be able to cope with the international reach of SIFIs’ operations, and of their assets and liabilities, and to set out how resolution authorities in home and host jurisdictions interact with each other.

Resolution measures in a home jurisdiction, such as for instance the transfer of assets or liabilities to a bridge bank, will not have automatic effect in host jurisdictions unless there is an internationally binding arrangement to this effect.

Although there are some existing legal mechanisms under which foreign jurisdictions might give effect to transfers to a bridge bank or to a private sector purchaser, most of these involve court proceedings that may not be sufficiently predictable or timely to contribute to effective resolution.

They often rely on doctrines applicable to insolvencies generally, do not adequately take into account considerations of financial stability, and may not provide authorities with the necessary powers to implement the transfer unless the consent of the relevant counterparties is obtained.

The cross-border effectiveness of resolution measures would be improved if both home and host authorities had the requisite powers and regimes, applying not only to domesticallyincorporated banks but to domestic branches of foreign banks, and to assets, liabilities and contracts of foreign banks located within a jurisdiction.

These should empower authorities to cooperate in the application of a range of special resolution tools to local operations, including the power to transfer assets, liabilities and contracts of the bank to a foreign bridge bank or private sector purchaser without the consent of the counterparties.

Statutory mandates to foster cross-border cooperation

Cooperation and trust among resolution authorities should be built up.

The mandates of resolution authorities should be framed so that they have to duly consider the potential impact of their resolution actions on financial stability in other jurisdictions.

In applying resolution powers to individual components of a financial group, the resolution authority should have to take into account the overall impact on the group as a whole and the impact on financial stability in other jurisdictions concerned and undertake best efforts to avoid taking actions that could reasonably be expected to trigger instability elsewhere in the group or in the financial system.

There should be a strong encouragement of cross-border cooperation supported by robust abilities to cooperate.

However, the statutory framework for cross-border cooperation would not be so prescriptive as to deprive jurisdictions of the flexibility to act when necessary to achieve domestic stability in the absence of effective cross-border cooperation and information sharing, or in the event of inaction or inappropriate action by the home authority.

Provisions that hamper fair cross-border resolution need to be removed.

More specifically, jurisdictions should ensure that no legal, regulatory or policy impediments exist that hinder the appropriate exchange of information, including firm-specific information, between supervisory authorities, central banks, resolution authorities, finance ministries and the public authorities responsible for insurance guarantee schemes.

The sharing of all information relevant for recovery and resolution planning and in resolution should be possible in normal times and in crisis at a domestic and a cross-border level.

The modalities for the sharing of information relating to a G-SIFI should be set out in institutionspecific cooperation agreements.

They should, in particular, provide for holding at least annual meetings including top officials of the home and relevant host authorities to ensure that they are effectively involved in and informed of status of the work on recovery and resolution plans and that they provide the needed leadership to the process.

Where appropriate and necessary to respect the sensitive nature of information, information sharing may be restricted, but should be possible among the top officials of the relevant home and host authorities.

National laws and regulations should not discriminate against creditors based on nationality or location of their claim, or the jurisdiction where it is payable.

Where any such provisions exist, they should be transparent and properly disclosed.

Institution-specific cooperation agreements

Home and host authorities need to consult and cooperate on actions that may affect each other’s jurisdiction.

A policy framework for cross-border cooperation between resolution authorities is needed to enable advance planning and to avoid dealing with cross-border issues through the courts.

In the near term, it may be easiest and most flexible for authorities to reach cross-border cooperation agreements on resolution on an institution-specific basis.

The FSB report on Reducing the moral hazard posed by systemically important financial institutions endorsed by the G20 in November 2010 (SIFI Recommendations) called for institutionspecific cooperation agreements for all G-SIFIs.

The document on Essential elements of institution-specific cooperation agreements proposes essential elements that such agreements between home and host authorities should have, covering both crisis planning and actions during resolution, with an emphasis on the latter.

Cooperation agreements should build upon the principles the Financial Stability Forum set out in April 2009.

The agreements should cover institution-specific crisis management planning and cooperation amongst relevant authorities in the event of the institution’s resolution.

The agreements should contain provisions that authorities wished had been in place to facilitate cooperation with respect to failing firms during the most recent crisis.

They should provide an appropriate level of detail with regard to the cooperation procedures in place both in the “pre-crisis” (i.e. recovery and resolution planning) phase as well as “in crisis”.

To do so they will need to be firm-specific and also lay out how national legal regimes interact.

They should also set out the framework under which the home and key host authorities cooperate in the elaboration of RRPs and the conduct of resolvability assessments.

Firm-specific agreements are needed among all members of a firm’s Crisis Management Group (CMG), which should include the home and all key host jurisdictions.

Bi-national agreements between authorities of the home and a host jurisdiction, and firm-specific multinational agreements among authorities of the home and all key host jurisdictions may complement each other.

Questions for public consultation

9. How should a statutory duty to cooperate with home and host authorities be framed?

What criteria should be relevant to the duty to cooperate?

10. Does Annex 3: Institution-specific Cross-border Cooperation Agreements cover all the critical elements of institution-specific cross-border agreements and, if

implemented, will the proposed agreements be sufficiently reliable to ensure effective

cross-border cooperation? How can their effectiveness be enhanced?

11. Who (i.e., which authorities) will need to be parties to these agreements for them to be most effective?

Resolvability assessments

At present, few if any SIFIs could be effectively resolved in an orderly and speedy fashion, given the existing powers available to authorities, the lack of legal capacity for national authorities to cooperate, and the complex structures and activities of the firms.

To identify the changes needed to regimes, legal powers and individual firms, resolvability assessments need to be made at the level of each individual SIFI.

Annex 4: Resolvability Assessments proposes in detail the process and elements of such an assessment.

Authorities need to make a candid assessment of the current resolvability of SIFIs, and the obstacles that exist, in order to determine what changes authorities and firms need to make.

A separate assessment is needed for each individual SIFI, as firms differ greatly in their corporate structure and mix of activities, and each presents its own technical complexities to be addressed in the event of crisis.

Annex 4 also defines a framework for assessing the feasibility of existing resolution tools and regimes and the credibility of resolution strategies in the light of the systemic impact of their application to a SIFI.

These assessments will help focus authorities and firms on the implications of the current status quo, and the steps that need to be taken, by firms, by national regimes and globally, to reduce the current systemic threat from the lack of adequate resolution procedures.

Questions for public consultation

12. Does Annex 4: Resolvability Assessments appropriately cover the determinants of a firm’s resolvability? Are there any additional factors to be considered in determining the resolvability of a firm?

13. Does Annex 4 identify the appropriate process to be followed by home and host

authorities?

Recovery and resolution plans

During the recent crisis, efforts to cope with the failure of Lehman Brothers were greatly complicated by a lack of preparation.

Basic information was missing about organizational structures and relationships between subsidiaries.

This made it difficult to act quickly, to anticipate the effects of different actions in different jurisdictions, and to resolve conflicts between subsidiaries and jurisdictions. Much economic value was lost as a result.

When a firm falls into distress, the authorities and the firm need detailed contingency plans to implement rapid, well-planned measures to ensure that the firm can continue to perform critical functions, or wind them down if necessary, without spillovers that damage the wider system.

An adequate, credible RRP should be required for any firm which is assessed by its home authority to have a potential impact on financial stability, in the event of liquidation of that firm.

The SIFI Recommendations call for RRPs to be put in place for all G-SIFIs.

Authorities and SIFIs are currently working together to create RRPs for each firm.

RRPs should set out in advance the measures, in the event of a crisis, that a firm could take to recover as a going concern or else that the authorities could take to resolve it in an orderly way.

RRPs and resolvability assessment complement each other: RRPs should use as a base the conclusions of the resolvability assessments discussed above; indeed, an important benefit of the process of developing a plan is to identify actions that firms need to take to make themselves resolvable.

RRPs of G-SIFIs will be reviewed, subject to adequate confidentiality agreements, within the institution’s CMG at least annually.

To ensure the involvement of the key decision makers and keep them informed, the adequacy of RRPs of G-SIFIs should also be the subject of a formal review, at least on annual basis, by top officials of home and relevant host supervisory and resolution authorities, where appropriate, with the firm’s CEO.

Questions for public consultation

14. Does Annex 5: Recovery and Resolution Plans cover all critical elements of a

recovery and resolution plan? What additional elements should be included? Are

there elements that should not be included?

15. Does Annex 5 appropriately cover the conditions under which RRPs should be

prepared at subsidiary level?

Improving resolvability

Complex organisational structures and business models, with economic functions and business lines spanning multiple legal entities with a web of intra-group exposures, make resolution more difficult.

The FSB has focused in detail on some particular areas arising from the complexities of SIFIs’ operations that can create practical obstacles to resolution:

- the need for information systems that can provide rapid, comprehensive data on the position of each of the firm’s legal entities when a crisis hits;

- the reliance on service providers, which may help firms capitalise on economies of scale and increase efficiency in normal times, but may pose obstacles to effective resolution and threaten the continued performance of systemically important operations;

- intra-group transactions, which may be a source of strength for a firm in normal times but can impede actions to deal separately with individual business units of a group during a crisis; and

- challenges in recovery and resolution of the essential services that a firm’s global payment operations provide to customers.

Proposals to help address these issues, including for the powers of supervisory and resolution authorities to require firms to reduce unnecessary organisational complexities and intra-group exposures, are included in Annex 6: Measures to improve resolvability.

Questions for public consultation

16. Are there other major potential business obstacles to effective resolution that need to be addressed that are not covered in Annex 6?

17. Are the proposed steps to address the obstacles to effective resolution appropriate?

What other alternative actions could be taken?

18. What are the alternatives to existing guarantee / internal risk-transfer structures?

19. How should the proposals set out in Annex 6in these areas best be incorporated

within the overall policy framework? What would be required to put those in place?

Timelines for implementation of G-SIFI related recommendations

The gap between the arrangements necessary for effective resolution of firms and the arrangements that are currently in place is wide.

The proposals in this paper need much detailed follow-up work, including on planning for individual firms and on bilateral and multilateral cooperation between authorities.

Authorities need to be accountable to each other and to the public for taking the steps needed to achieve the objectives of the proposals.

The SIFI Recommendations call for RRPs and firm-specific cooperation agreements to be put in place for all G-SIFIs.

In this respect, the time line and key milestones that authorities should work towards in their immediate tasks of developing RRPs and conducting resolvability assessments for G-SIFIs; and enhancing cross-border cooperation among home and key host authorities of G-SIFIs are set out below.

Cross-border Cooperation Agreements

- Before the end of 2011, home authorities of G-SIFIs should have begun engaging with key host authorities as regards institution-specific cooperation agreements.

- By June 2012, the modalities for information sharing within the CMGs and the first drafts of the cooperation agreements should be completed.

- By December 2012, home authorities of G-SIFIs should have entered into cooperation agreements with the key host authorities.

Recovery and Resolution Plans

- By December 2011, the first drafts of the recovery plans should be completed.

- By June 2012, the first drafts of the resolution plans should be completed.

- By December 2012, both the recovery plans and the resolution plans should be completed.

Resolvability Assessments

- By June 2012, home authorities of G-SIFIs should have entered into discussions with firms and members of their respective CMGs as regards the preliminary assessment of the firms’ resolvability.

- By December 2012, the first resolvability assessments should be completed.

CMG Outreach

- By June 2012, CMGs should have identified the jurisdictions where the respective firms have a systemically important presence, but are not represented in the CMGs.

- By December 2012, the modalities for cooperation and information sharing between the home authorities and the host authorities of jurisdictions not represented in the CMG where the G-SIFIs have a systemically important presence, should be established.

Questions for public consultation

20. Comment is invited on the proposed milestones for G-SIFIs.

II. Discussion notes

The FSB is exploring the policy issues surrounding two specific additional measures to strengthen effective resolution.

In particular, the FSB is assessing the pros and cons of greater convergence in creditors’ hierarchy, depositor preference and depositor protection in resolution and of the possible introduction of a brief stay on contractual early termination rights upon entry into resolution in order to facilitate the implementation of resolution measures.

1. Discussion note on creditor hierarchy, depositor preference and depositor protection in resolution (Annex 7)

An important feature of effective resolution regimes is to “make it possible for shareholders and unsecured and uninsured creditors to absorb losses in their order of seniority.”

This should be achieved in an equitable manner and in a manner that keeps the risk of contagion to a minimum and obviates the need for bail-outs.

Clarity and predictability as regards the order of seniority or statutory ranking of claims in insolvency is a necessary prerequisite for effective resolution.

It determines the allocation of losses.

It shapes the incentives of market participants and pricing of risk.

It affects the ease with which certain resolution measures can be applied.

Differences in ranking can complicate cross-border resolutions.

Effective protection of local depositors and assurance of financial stability will be crucial considerations in the determination by the host authorities of whether to cooperate.

The FSB is seeking public comment on the issue of whether or not existing differences in statutory credit ranking represent an impediment to effective cross-border resolution and greater convergence in particular in, the treatment of deposit claims, could be pursued further at the international level.

Questions for public consultation

21. Does the existence of differences in statutory creditor rankings impede effective crossborder resolutions?

If so, which differences, in particular, impede effective crossborder resolutions?

22. Is a greater convergence of the statutory ranking of creditors across jurisdictions

desirable and feasible?

Should convergence be in the direction of depositor preference or should it be in the direction of an elimination of preferences?

Is a harmonised definition of deposits and insured deposits desirable and feasible?

23. Is there a risk of arbitrage in giving a preference to all depositors or should a possible preference be restricted to certain categories of depositors, e.g., retail deposits?

What should be the treatment of (a) deposits from large corporates; (b) deposits from other financial firms, including banks, assets managers and hedge banks, insurers and pension funds; (c) the (subrogated) claims of the deposit guarantee schemes (especially in jurisdictions where these schemes are financed by the banking

industry)?

24. What are the costs and benefits that emerge from the depositor preference? Do the benefits outweigh the costs? Or are risks and costs greater?

25. What other measures could be contemplated to mitigate the impediments to effective cross-border resolution if such impediments arise from differences in ranking across jurisdictions?

How could the transparency and predictability of the treatment of creditor claims in a cross-border context be improved?

2. Discussion note on conditions for a temporary stay on early

termination rights (Annex 8)

Under standard market documentation for financial contracts, contractual acceleration, termination and other close-out rights (collectively, “early termination rights”) in financial contracts may be triggered when the resolution authorities initiate resolution proceedings or take certain related resolution actions with respect to a financial institution.

In the case of a SIFI, the termination of large volumes of financial contracts upon entry into resolution could result in a disorderly rush for the exits and frustrate the implementation of resolution measures, such as the transfer of critical operations to a bridge bank, or the implementation of bail-in within resolution, which are aimed at achieving continuity of critical financial functions and of the financial contracts that support them.

The FSB is seeking public comments on the introduction of a possible statutory provision that would allow for a brief suspension of early termination rights pending the use of resolution tools, as well as the length and scope of such a stay, possible exemptions and its cross-border application.

Questions for public consultation

26. Please give your views on the suggested stay on early termination rights.

What could be the potential adverse outcomes on the failing firm and its counterparties of such a short stay?

What measures could be implemented to mitigate these adverse outcomes?

How is this affected by the length of the stay?

27. What specific event would be an appropriate starting point for the period of

suspension?

Should the stay apply automatically upon entry into resolution?

Or should resolution authorities have the discretionary right to impose a stay?

28. What specific provisions in financial contracts should the suspension apply to?

Are there any early terminations rights that the suspension should not apply to?

29. What should be an appropriate period of time during which the authorities could

delay the immediate operation of contractual early termination rights?

30. What should be the scope of the temporary stay?

Should it apply to all counterparties or should certain counterparties, e.g., Central Counterparties (CCPs) and FMIs, be exempted?

31. Do you agree with the proposed conditions for a stay on early termination rights?

What additional safeguards or assurances would be necessary, if any?

32. With respect to the cross-border issues for the stay and transfer, what are the most appropriate mechanisms for ensuring cross-border effectiveness?

33. In relation to the contractual approach to cross-border issues, are there additional or alternative considerations other than those described above that should be covered by the contractual provision in order to ensure its effectiveness?

34. Where there is no physical presence of a financial institution in question in a jurisdiction but there are contracts that are subject to the law of that jurisdiction as the governing law, what kind of mechanism could be considered to give effect to the stay?

If you want to see the Annex, download this document and look for the annexes on page 63 – 127:

Basel 3 News October 2011

Best Regards,

George Lekatis
President of the Basel iii Compliance Professionals Association (BiiiCPA)

Basel III: Are the basel iii rules in or out?

oktober 27, 2011 in Columns-Artikelen by George Lekatis

We had another interesting month. To hope that some of the Basel iii rules will be more flexible is one thing. The belief that the Basel iii rules are anti-American is another. Jamie Dimon, the Chief Executive Officer of J.P. Morgan finds the Basel III rules “anti-American.”

In an interview with Financial Times, Dimon said:

“I’m very close to thinking the U.S. shouldn’t be in Basel anymore. I would not have agreed to rules that are blatantly anti-American”.

Mr Dimon has now some new friends and some new enemies. The Street columnist Glenn Hall, for example, said: “Jamie Dimon appears to be suffering from a form of delusional paranoia that makes him believe everyone is out to get his bank.”

Jean-Claude Trichet, the head of the European Central Bank, had another point of view:

“I see resistance of some in the financial sector against Basel III. For me, it is crystal clear: what has been decided is decided”

So, what should Mr Dimon say?

I believe he should say that European banks can use different accounting and modeling principles, and they may have a competitive advantage, so the American regulators must ensure that the standards in the States are similar to the European ones.

But there is no point to attach to the Basel iii framework, that already has been endorsed not only by the States, but also by the G20 leaders.

At the other part of the world, the Australian Prudential Regulation Authority announced a proposal that would require Australian banks to adopt the minimum capital requirement ratio two years ahead of the Basel iii timetable.

Also, they would implement the capital conservation buffer three years ahead of the Basel III’s schedule.

Of course the Australian Bankers’ Association (ABA) does not like the idea of the accelerated timeline (no, they have not called it Anti-Australian yet).

In China, systemically important banks will be subject to a minimum capital adequacy ratio (CAR) of 11.5 percent; other banking institutions will be required to adhere to a minimum CAR of 10.5 percent.

We also had the next hit for the investment banking sector. UBS trader Kweku Adoboli is accused of fraud and two charges of false accounting.

The Financial Services Authority and its Swiss counterpart have asked for a “comprehensive, independent investigation” into the events that led to the trading losses at the bank’s London operations.

Chancellor George Osborne found the opportunity to support the unique and very strict new measures in UK, AND speak about the need for a shake up of the banking sector in Britain.

Oh, not another hit.

He said: “I draw two lessons. One is we need a better system of regulation, and that is why the British Government is proposing to give to the Bank of England much greater powers of supervision, so it can look across the piece at issues of stability and proper conduct in our banking system, and indeed we will have a financial conduct authority specifically looking at market abuse. ”

“But also I take the lesson that John Vickers drew to our attention a few days ago which was: how do you protect retail banking from those kinds of activities in investment banking that we saw at UBS this week? ”

“If you ever wanted a better example of why the kinds of ideas that John Vickers was putting forward were right for Britain, look at what happened at UBS just a few days later.”

Oh…

The FSA and the Swiss Financial Market Supervisory Authority said a third party would investigate the details of the alleged unauthorised trading activity and why the activities remained undetected.

IMF Working Paper, August 2011

Possible Unintended Consequences of Basel III and Solvency II

Prepared by Ahmed Al-Darwish, Michael Hafeman, Gregorio Impavido, Malcolm Kemp, and Padraic O’Malley

Important parts

Efforts to strengthen the quality of capital for banks and insurers through Basel III and Solvency II are well advanced.

On the one hand, the Basel Committee on Banking Supervision (BCBS), the organization responsible for developing international standards for banking supervision, adopted the Basel II framework in 2004 and, in response to the financial crisis, has taken steps to strengthen it in an incremental fashion to form what is now known the Basel III framework (BCBS 2009, 2011a, 2011b, and 2011c).

On the other hand, the European Commission (EC) is leading the Solvency II project, in close cooperation with the European Insurance and Occupational Pensions Authority (EIOPA), to develop harmonized standards for insurance supervision within the European Union.

A directive was adopted in 2009, and work—which included a series of quantitative impact studies studies—has been underway to develop supporting rules.

The regional scope of application of the two accords varies.

Basel is an “accord”/agreement with no legal force but potentially global applicability, whereas Solvency II is a legal instrument that will be binding in 30 European Economic Area (EEA) Countries (27 European Union (EU) states plus Iceland, Liechtenstein, and Norway).

However, Solvency II has also implications beyond Europe through, for example, its influence on the international standards being developed by the International Association of Insurance Supervisors (IAIS), and because external insurance groups will be more easily able to operate in the EU if their home supervisory regimes are considered equivalent.

Although these standards have much in common, differences do exist.

Both take a risk-based approach to minimum capital requirements and supervision and promote the integrated use of models by institutions in managing risks and assessing solvency.

However, their objectives overlap only partially.

In particular, Basel III attempts to increase the overall quantum of capital and its quality as a means of protecting against bank failures, including improved quantification of risks that were poorly catered for under Basel II.

However, Solvency II attempts to strengthen the quality of capital and tailor the quantity of capital required more closely to the risks of each insurer, without necessarily increasing the quantity within the sector as a whole.

Finally, the two accords have been tailored to the business characteristics of the respective sectors, often as a result of bilateral negotiations, and shaped by the views of those involved in their development in a piecemeal manner.

Accordingly, they have generated long and complex documents which define the same concepts in different ways and often deal differently with the same or similar issues.

While different, business activities of banks and insurers often overlap.

The business differences affect the nature of their assets and liabilities, the manner in which these entities are funded, and the overall structure of their balance sheets.

Accounting rules, which affect the valuation of assets and liabilities—and the resulting level of capital—also differ in some respects by sector.

At the same time, banks and insurers compete on both investment and protection products.

In summary, while the banking and insurance businesses are fundamentally different, substantial overlap exists in key areas.

This section discusses similarities and differences in business models, accounting standards, ability to modify the risk profile of assets and liabilities, and product competition between the two sectors.

A. Differences in Business Model

Banks’ business is asset-driven and often supported by leveraged balance sheets.

Banks provide a large set of fee-based services, such as payment and settlement processing, and financial services.

Commercial/retail banks provide mainly loans to borrowers, typically funded by pooling the savings of depositors; i.e., the business is based on maturity transformation and the loan books represent the largest component of their assets while deposits represent the largest components of their liabilities.

However, if deposits are insufficient to support the growth of the loan book, banks leverage their balance sheet by seeking additional funding by issuing commercial paper or bonds, borrowing from other banks or the central bank, or engaging in securities repurchase transactions (repos).

Investment banks provide a large set of financial services to clients and engage in securities trading activity for both their customers’ accounts and their own account.

The funding needed to support trading activities and their associated risks typically comes from shareholders or wholesale sources.

Finally, universal banks combine both commercial and investment banking activities.

Accordingly, their assets, liabilities, and funding sources share characteristics with both commercial and investment banks.

Different activities subject banks to different risks.

The loan book subjects a bank to credit risk, should borrowers default on the obligation to repay their loans, but such credit risk is capped by the value of the loans.

The loan book (in combination with short term funding) also subjects a bank to liquidity risk.

Indeed, the duration of deposit liabilities is typically shorter than the duration of loan assets.

This maturity transformation usually serves to enhance the spread, but it can subject a bank to significant liquidity risk which, in turn, contributes to the need to protect depositors with safety nets.

In addition, since there is no direct link between the liabilities and the assets, if there is a run on the bank, where customers withdraw their deposits and other banks refuse to provide interbank loans, the loan assets might not be readily realizable.

Investment, and universal and commercial banks are also exposed to market, counterparty, and operational risk through their trading books.

Finally, fee-based services subject banks to operational risk and, in some cases, to counterparty credit risk.

Banks’ profits arise from the spread between the income earned on assets and the cost of liabilities.

Fee-based services generate fee income.

The loan book produces interest income and the trading book produces investment income.

On the liability side, costs duration.

Insurance activities generate mainly underwriting, credit, and market risk.

The policyholder protection activity exposes both life and nonlife companies to underwriting risk, market risk, and perhaps (depending on the guarantees provided by the policy) credit risk.

Life products (especially when combined with savings components) can also create liquidity risk, although many include restrictions or penalties intended to mitigate this risk.

For some classes of nonlife insurance business, such as credit insurance and financial surety, the underwriting risk is similar to the credit risk borne by a bank.

Finally, insurers typically seek to invest the premiums in assets that are suitable to the nature of the corresponding insurance liabilities.

Mostly, there is a close and direct link between the value of the insurance liability and the value of the related assets.

However, when adequate assets are not available, the duration of an insurer’s liabilities is often longer than the duration of its assets.

Any interest rate duration mismatch creates market risk for the insurer in the form of interest rate risk, but some investment types (i.e., asset-backed securities) can also create credit spread duration risk.

Insurers are less leveraged and therefore, less exposed to liquidity risk than banks.

Insurers are largely funded by policyholders with technical provisions representing the largest component of insurers’ balance sheets and they are less reliant on debt and other external funding than are banks.

However, some life insurance products generate losses (under current methods of calculating liability values) during the first years of the policy and insurers sometimes purchase reinsurance or use own capital to help meet this new business strain.

Insurers are less exposed to the risk of mass lapses of policies than are banks to runs on deposits because of the locked-in nature of much of their funding.

The ability of insurers to anticipate future profits under Solvency II could increase the exposure to mass lapses, because it means that some might not have sufficient readily-available capital to repay all policyholders in the event of a run.

However, the mass lapse stress component in the solvency capital requirement (SCR) helps to mitigate this risk.

Insurance profits arise from the difference between income earned and the costs of providing insurance benefits.

The relative importance of the different income and expense items varies by business line, but, in general, the largest source of income is typically represented by policyholders’ premiums followed by the investment income that accumulated premiums generate prior to the payment of insurance benefits.

The largest component of expenses is typically constituted by claims payments to which underwriting, reinsurance, investment, and administrative expenses need to be added.

Insurance liabilities have more loss-absorption capacity than bank liabilities.

Insurance liabilities, particularly those in respect of life insurance policies with profit participation, provide some risk-absorption capacity.

If experience is adverse and there is profit participation in place, some of the adverse impact would reduce policyholder liabilities, thereby mitigating the impact on capital.

Although the same partial pass-through can also arise in banking (for example, with a derivative contract, if defined in an appropriate manner), such contract terms are much less common than in insurance.

These differences in business models translate into differences in both levels and quality of capital.

The largest 50 European banks have on average 6 percent total capital over total assets, of which only 67 percent is represented by high-quality Core Tier 1 capital.

Large insurers world-wide have on average 8 percent of total capital over total assets, of which 84 percent is represented by high-quality equity capital.

Finally, large global reinsurers have on average more capital (with total capital amounting to 15 percent of total assets) but are also more leveraged (with high quality equity capital amounting to 73 percent of total capital).

B. Key Differences in Accounting Standards

Banks and insurers use different accounting standards.

In many jurisdictions, the financial statements of banks are prepared in accordance with International Financial Reporting Standards (IFRS) or similar accounting rules.

The rules are a mixture of fair-value accounting (for trading activities and available-for-sale banking assets) and historical cost accounting (for other banking assets).

The accounting standard was not designed specifically for banking regulatory purposes, so various adjustments and deductions are made in moving from the financial statements to the determination of capital adequacy.

Solvency II uses instead a “total-balance-sheet approach” based on the fair values of both assets and liabilities.

The Solvency II balance sheet must be prepared independently of any general-purpose financial reporting requirements.

Although some assets and liabilities will be valued in the same manner as under IFRS, others will be valued differently.

Given the different underlying accounting bases, it is unsurprising that the prudential adjustments under Solvency II differ from those under Basel III

Bank assets are typically valued at amortized book value or fair value.

Bank loans, under IFRS, fall under financial instruments and are accounted for by using IAS 39.

Under IAS 39, financial assets are categorized as either

(i) loans and receivables;

(ii) financial assets at fair value through profit or loss;

(iii) available for sale; or

(iv) held to maturity.

Other financial assets are initially measured at fair value plus, if not valued at fair value through profit or loss, transaction costs that are directly attributable to the acquisition.

Financial assets at fair value and assets available for sale are subsequently carried at fair value, while loans and receivables and held-to-maturity assets are carried at amortized cost.

All derivatives, whether held for trading or for hedging, are carried at fair value.

IFRS 9, which will apply from January 2013, proposes that all financial assets will either be measured at amortized cost or fair value.

In addition, loan provisions are only “retrospective.”

Banks are not permitted to make provisions for expected future loan losses, no matter how likely they are to arise.

IAS 36 “Impairment of Assets” requires a case-by-case assessment of the facts and circumstances surrounding the recoverability and timing of future cash flows relating to a credit exposure.

Should there be an expectation that all contractual cash flows would not be recovered (or recovered without full future interest payments), an exposure would be classified as impaired.

The extent of impairment would be measured on a present-value basis using the effective interest rate of the exposure as the discount rate.

For groups of loans that share homogenous characteristics (such as mortgage and credit card receivables), impairment can be assessed on a collective basis.

The aim of an individual or collective assessment is to capture the incurred loss for a specified portfolio.

General provisions are permissible only to the extent that they relate to a specified risk that can be measured reliably and for incurred losses.

For investments, a similar analysis is conducted, the key difference being that the fair value of the investment is also considered as an input.

However, the assets of insurers are valued at fair value under Solvency II and provisions are prospective.

IFRS is used as a proxy for fair value, except for some specific instances when IFRS is not equivalent to fair value.

Provisions for expected losses on the assets of insurers would thus be inherently reflected in the valuation.

Bank liabilities are also valued at amortized book value or fair value.

Bank deposits are typically valued at book value or amortized book value.

A retrospective methodology is used, reflecting the amount that would be paid upon exit.

Other financial liabilities of banks are initially measured at fair value plus, if not valued at fair value through profit or loss, transaction costs that are directly attributable to the acquisition.

After initial recognition, financial liabilities can be carried either at amortized cost using the effective interest method or at fair value through profit or loss.

A new accounting standard, IFRS 9, will apply from January 2013 and will require companies choosing to measure a liability at fair value to present the portion of the change in its fair value due to changes in the entity’s own credit risk in the other comprehensive income section of the income statement, rather than within the profit and loss.

However, insurance liabilities under Solvency II are valued at the amount for which they could be transferred or settled between knowledgeable willing parties in an arm’s-length transaction.

The values can be calculated using one of two methods: best estimate plus risk margin; or as a whole.

The first methodology involves projecting the probability-weighted future cash flows and required capital associated with the insurer’s obligations and discounting the results back to the valuation date.

The risk margin is included to ensure that the value is equivalent to the amount that another insurer would likely require in order to take over the insurance obligations.

The cost-of-capital rate most recently tested was 6 percent.

The best estimate liability may be negative.

The second methodology uses the market value of a portfolio of replicating assets, which implicitly results in a prospective valuation.

This method is appropriate if the future cash flows associated with the insurance obligations can be replicated reliably, using financial instruments for which a reliable market value is observable, such as the obligations in respect of unit-linked products.

Another difference relates to the effect of an entity’s own credit standing in the valuation of its liabilities.

Under IAS 39, a bank’s financial liabilities can be adjusted for changes in its own credit standing.

This allows a bank, the position of which is weakening, to book profits based on its theoretical ability to buy back debt at a reduced cost.

Insurers will not have this option under Solvency II, which does not permit any adjustment to take account of the own credit standing of the insurer.

However, this accounting difference does not provide an advantage to banks in the calculation of regulatory capital, because Basel III requires  that common equity Tier 1 capital (CET) be adjusted by cumulative gains and losses due to changes in own credit risk on fair-valued financial liabilities.

These differences in accounting standards for banks and insurers suggest the following:

1. Banks have greater flexibility than insurers regarding the valuation of financial liabilities.

Banks can value liabilities using either amortized cost or at fair value through the profit and loss—although a bank cannot freely switch valuation bases, while insurers will be required to value them at fair value under Solvency II.

The reported values of financial liabilities that are valued at amortized cost are more stable than those valued at fair value, which will move in response to changes in interest rates.

2. The retrospective valuation of bank balance sheet items generates greater stability in the reported values.

Banks have the ability to classify financial assets as loans or held to maturity, which enables them to value the assets using amortized cost.

The assets of insurers change in value as interest rates move or credit spreads change, whereas bank assets that are valued using amortized cost will not.

Even the new approach proposed by the International Accounting Standards Board (IASB) would still differ from a fair-value approach because, for example, movements in credit spreads will not automatically result in any change to the valuation of the asset.

It is not until a loss event occurs that the bank would move the asset from the good book to the bad book and recognize the likely cost of default.

3. The relative prudence of the different valuation methods in provisioning of banks and insurers depends upon the profitability of the business written.

If a contract is expected to have positive future profitability, then the retrospective approach is more prudent, because the liability will be valued at book value and will not recognize the future profit that the company expects to earn on the contract.

Under Solvency II, the insurer will be able to recognize this expected future profit upon writing the contract.

However, if a contract is expected to produce future losses, then the prospective approach is more prudent, because those losses would be recognized upfront rather than spread over the remaining life of the contract as with the retrospective approach.

Another factor that affects the level of prudence is that insurers are required to include a margin in the liabilities in respect of the cost of capital, whereas banks do not include such a margin in their retrospective valuation of liabilities or in its loan provisioning.

Finally, important changes are underway at the IASB, which will affect the accounting of banks and insurers.

In addition to the aforementioned IFRS 9, they include changes regarding the definition (IAS 32) and classification of financial instruments (IAS 39), and the recently-issued Exposure Draft on Insurance Contracts (IFRS 4).

In addition, the IASB is currently consulting on a requirement for balance sheet loans to bear an expected loss provision throughout their life.

In November 2009, the Board issued an exposure draft proposing the adoption of an expected loss model for recognizing credit losses.

The IASB has also agreed to incorporate an incurred loss model on top of the expected loss model, often referred to as a good book (expected loss model)/bad book (incurred loss model) approach.

The BCBS supports the use of more forward-looking provisions.

The effects of these changes are unclear.

C. Differences in the Ability to Change Assets and Liabilities

Banks and insurers manage differently the risk profile of their assets and liabilities.

They typically do so to reduce overexposure to particular risks or to free up capacity to take on more business.

The ways in which this is accomplished can differ significantly by sector.

Banks commonly securitize loan assets and sell them into the capital markets.

Securitization can reduce exposure to a particular type of loan, type of borrower, or geographic area.

By removing assets from the balance sheet, it may also release regulatory capital, which can then be used to support additional loans.

However, a securitization must be of sufficient size to make the transaction economically viable.

Sometimes, as in collateralized debt obligations (CDOs), securitizations are repackaged into another securitization.

Insurers commonly use reinsurance and, less frequently, capital market instruments to transfer risk.

Reinsurance can be used to cover individual risks

(i)to share the underwriting risk on a particular portfolio;

(ii) to prevent the absolute scale of any loss exceeding a particular amount;

(iii) to stabilize experience;

(iv) to provide capital support;

(v) to release future profits; and

(vi) to protect against catastrophes.

It offers significant flexibility to insurers to alter the nature of the risk that they hold on their balance sheet and allows them to tailor the exact form of underwriting risk that they wish to take, and it can also be structured to transfer credit risk or market risk.

Many types of reinsurance enable an insurer to reduce the insurance liabilities held on its balance sheet.

However, because an insurer remains liable to its policyholders in respect of the benefits that have been reinsured, reinsurance essentially transforms underwriting risk into counterparty credit risk.

This counterparty risk can be protected against using methods such as depositing assets back with the insurer or providing collateral.

Much less frequently, insurers limit their exposure to extreme nonfinancial underwriting risks to others through the capital markets, using insurance-linked securities (ILSs).

Similarly to reinsurance, ILSs limit the risk exposure, but they do not take the liability off an insurer’s balance sheet.

For many insurance liabilities there is simply no secondary market, which impacts on an insurer’s ability to offload risk and also on the ability to value insurance liabilities.

Finally, both banks and insurers use derivatives to manage risks, but generally with different purposes.

Banks use derivatives mainly to hedge exposures arising from their trading activities, as well as credit default swaps (CDSs) to limit risks associated with their lending activities.

Life insurers use derivatives to hedge exposures to market risk as part of their asset-liability management (ALM) programs.

D. Different Treatment of Similar Products

Banks and insurers compete both on products that are predominantly investments in nature and those that predominantly provide protection.

Examples of investment products that are written by both banks and insurers include investment bonds, term deposits offered by banks, and term-certain annuities offered by insurers.

Examples of protection products include:

(i) Investment guarantees and options written by investment banks and variable annuities with investment guarantees written by insurers;

(ii) CDSs written by both banks and insurers; and

(iii) Trade finance offered by banks and surety bonds offered by nonlife insurers.

Different tax and capital treatments of such products can generate opportunities for product and capital arbitrages.

Although these products offer similar economic benefits to customers and expose financial institutions to similar economic risks, differences in the manner in which they are taxed—with respect to either the financial institution or the consumer—can affect the related assets or liabilities.

For example, in some jurisdictions savings products offered by life insurers provide for the deferral of tax on related investment income until the policy benefits are paid, while the income on bank deposits is taxed annually.

This may discourage withdrawals from the insurance policies, subjecting an insurer to less liquidity risk than a bank that offers a similar savings product (unless the tax regime changes in a way that creates an incentive for the insurance policyholder to withdraw).

In addition, differences in their capital treatment under the two accords can generate opportunity for arbitrage across sectors.

E. Implications

The aforementioned differences between banks and insurers reflect the

fundamentally different roles played by these intermediaries in the financial system.

Banks are a reflection of money serving as a means of payment in exchange.

Their comparative advantage is to screen projects and finance them over the short term.

This shortterm maturity transformation calls for the existence of safety nets to protect depositors against liquidity shocks.

Insurers, instead, (especially life insurers), are a reflection of money serving as a store of value, permitting deferred consumption and smoothing.

Their comparative advantage (strictly as investors) is to invest in long-term securities and benefit from their illiquidity premium.

The different types of intermediation of banks and insurers already suggests capital needs to protect entities against unexpected shocks should be different across sectors.

The aforementioned differences can have a range of implications.

For instance, the assets and liabilities of insurers are more often linked to one another (due to the liability driven business model of insurers and strict matching requirements of ALM programs) than are those of banks (where links exist mainly through a network of financial exposures).

This reduces their liquidity risk and need for external funding.

Also, due to liquidity risk, the need for funding, and the effects on loan defaults, banks have more short-term exposure to changing economic conditions than insurers.

For banks, the credit risk can be transferred through securitization, but the liquidity and funding risks are more difficult to deal with.

For insurers, reinsurance represents an additional tool—unavailable to banks—that can alter the nature of their balance sheet.

This could provide insurers with additional degrees of freedom to reduce their cost of capital.

Furthermore, there is a much greater degree of interconnectivity among banks, especially through the interbank market, than there is among insurers.

Therefore, there is generally a much greater risk that the failure of a single bank could have systemic implications (Geneva Association, 2010).

However, if an insurer is significantly involved in activities—such as providing protection against credit exposures—which are closely linked to the broader financial sector—then its failure could likewise have systemic implications.

Finally, insurers—especially life insurers—have more long-term exposure to changing economic and other conditions than banks because of the nature of their liabilities.

ALM can be used to help deal with the economic risks.

However, without effective ALM, the prospective nature of liability valuation can accentuate the effects of economic changes on the balance sheet.

More importantly, the aforementioned differences justify many but not all the differences in the respective capital accords.

In particular, Basel III reflects the general realization that capital has failed (because deemed ex post insufficient and of poor quality) during the recent crisis to protect banks and tax payers from unexpected losses.

This issue does not primarily guide Solvency II, as insurers have proved to be more resilient to the same shocks.

Notwithstanding this observation, several differences in the two accords (as we will see in the next sections) are not supported by differences in the nature of capital needed in the two sectors, potentially generating unintended consequences.

Basel III and Solvency II aim to ensure sufficient regulatory capital is held by banks and insurers.

Basel III aims to increase the quality and level of capital (and to reinforce banks’ liquidity risk management) in an incremental fashion to the existing regime.

Solvency II does not have the same objective: it aims to increase the protection of policyholders by creating incentives for good risk management and ensuring that the quantity/quality of capital required is calibrated to risks that insurers’ are exposed to.

Basel III and Solvency II have had completely different and independent development processes.

These two international accords have different histories:

Basel III has been developed over the last two years in reaction to the recent financial crisis, while Solvency II is the first attempt to develop a fully risk-based solvency standard for the European insurance industry, and has had a much longer gestation.

The differences in the nature of the respective standard setters, their objectives, and their largely independent development processes have affected the accords and their scope of application.

Finally, at the time of writing, the two accords are yet to be finalized.

Key Elements of Basel III

Basel III has been developed by the BCBS in reaction to the recent financial crisis.

The BCBS consists of senior representatives of bank supervisory authorities and central banks from and central banks from the major financial centers, including all G20 jurisdictions.

It is intended to apply to internationally-active banks, on a fully-consolidated basis.

Basel III evolved from the Basel II framework, and now includes liquidity requirements and gives increased attention to dealing with system-wide issues, such as the interaction of prudential requirements and economic conditions.

In order to implement Basel III in a particular jurisdiction, its requirements generally must be transposed into local legislation.

For example, in the EU this requires a directive as well as, possibly, regulations, which need in turn to be transposed into legislation in the member countries.

Many jurisdictions that are not members of the BCBS also seek to implement the Basel accords because of the international character of the banking business.

Furthermore, some jurisdictions apply the Basel accord, although sometimes with simplifications, even to banks that are not internationally active.

Although the basic framework of Basel II has been retained in Basel III, many of its elements have been strengthened, including the following:

- Early reforms to Basel II were agreed in July 2009, which will raise capital requirements for the trading book, complex securitization exposures, and securitizations in both the banking and the trading books.

The reforms immediately raised the standards of the Pillar 2 supervisory review process, and the capital requirements and related disclosures are to be implemented by the end of 2011.

- Basel III requires that Tier 1 capital must be of higher quality —focusing on common equity; harmonizes Tier 2 capital instruments; and eliminates Tier 3 capital instruments, which had been available to cover market risks.

All elements of capital will be required to be disclosed, along with a detailed reconciliation to the reported accounts.

A leverage ratio will also be applied to provide an extra layer of protection against model risk and measurement error.

- It includes measures to strengthen the capital requirements for counterparty credit exposures arising from derivatives, repo, and securities financing activities.

The changes were designed not only to increase capital requirements, but also to reduce procyclicality and to provide incentives to move derivative contracts to central counterparties.

- It introduces a framework designed to promote the conservation of capital and the build-up of buffers that can be drawn down in periods of stress.

The capital buffer range can be adjusted in response to signs that credit has grown to an excessive level.

Basel III also introduces internationally-harmonized liquidity standards.

The Liquidity Coverage Ratio (LCR) is designed to measure whether a bank has sufficient highquality liquid resources to survive an acute stress scenario lasting for one month, while the Net Stable Funding Ratio (NSFR) has a time horizon of one year and helps to determine whether there is a sustainable maturity structure of assets and liabilities.

A long transition period is foreseen for the implementation of Basel III.

Although Basel III is effective from January 1, 2013, some items are subject to phase-in arrangements; for example:

- The Leverage Ratio must be disclosed beginning in 2015 and becomes a Pillar 1

requirement in 2018;

- The Minimum Common Equity Capital Ratio reaches its ultimate level in 2015;

- Various deductions from common equity are phased in between 2014 and 2018;

- The Minimum Tier 1 Capital requirement reaches its maximum in 2015;

- The Capital Conservation Buffer is introduced in 2016 and reaches its ultimate level in 2019;

- Capital instruments that no longer qualify as noncore Tier 1 or Tier 2 capital are phased out over a 10-year horizon beginning in 2013; and

- Minimum standards will be introduced for the LCR in 2015 and the NSFR in 2018.

Additional reforms are also being considered.

Looking to the future, the BCBS is promoting stronger provisioning practices by advocating a change in the accounting standards toward an expected loss approach, updating supervisory guidance, and providing incentives in the regulatory capital framework.

It is also conducting fundamental reviews of the trading book and the securitization framework.

Key Elements of Solvency II

By contrast, Solvency II originated from the European Commission and not from the IAIS, the insurance equivalent of the BCBS.

Solvency II applies to almost all insurers operating in the European Community.

This includes small, local insurers, except for those with gross premium income below EUR 5 million.

Its requirements are also relevant to reinsurers or groups based outside the European Community but operating therein, which can be treated more like local insurers with respect to reinsurance, group solvency, and group supervision if their home supervisory regimes are assessed as equivalent by EIOPA.

Solvency II is broader in scope than Basel III.

It attempts to create a comprehensive, principles-based system of regulation and supervision.

In addition to establishing risk-based capital adequacy requirements, it consolidates 14 existing insurance directives and deals with such topics as the valuation of assets and liabilities, group supervision, and the winding up of insurers.

Several levels of legislation must be in place for Solvency II before it takes effect on January 1, 2013.

The Level 1 directive was approved by the Council of Ministers and the European Parliament in 2009.

Various consultations and quantitative impact studies have since taken place, including the recent QIS5.

A proposed revision to the Level 1 directive was published by the Council of the European Union on June 7, 2011 and is not expected to be finalized until early 2012.

Level 2 implementing measures are currently expected to be published in Quarter 1 2012.

EIOPA is expected to submit draft binding technical standards to the European Commission in Quarter 2 of 2012 and to publish Level 3 guidelines in Quarter 3 of 2012.

There is a wide range of issues on which decisions remain to be taken in arriving at the final requirements and guidelines, including the following:

- Assessment of the adequacy of the calibration of quantitative requirements;

- Determination of the risk-free interest rate curve used for discounting technical provisions and the use of transitional measures;

- Development of actuarial guidelines for the calculation of the best estimate;

- Assessment of the consistency between IFRS valuation and Solvency II valuation approaches;

- Assessment of the complexity of the Solvency Capital Requirement (SCR) standard formula and the need for simplifying the standard formula;

- Calculation of the group SCR (including group-specific risks);

- Based on the data gathered during QIS5, considering the grandfathering of existing capital items;

- Development of guidelines to allow insurers to apply for the use of an internal model;

- Transitional periods for a number of items;

- Development of guidelines on own-risk and solvency assessment (ORSA); and

- Development of guidelines on the supervisory review process.

Implications

The implications of the different situations described above are somewhat uncertain.

In general, since Solvency II does not have the same explicit objective to increase capital as Basel III, capital costs for the two sectors could rise asymmetrically, possibly promoting inappropriate risk migration from banks to insurers.

However, under both accords, there are still aspects of the requirements that are yet to be finalized and which could result in additional unintended consequences.

Some of these, such as the calibration of quantitative requirements, could affect the opportunities for cross-sectoral arbitrage.

Within the European Community, the overall legal and regulatory framework should produce relatively consistent implementation of both Basel III and Solvency II from one jurisdiction to another.

However, among other jurisdictions, only peer pressure will limit the variations in Basel III implementation.

The influence of Solvency II outside the European Community will be even less certain, depending on such factors as the results of the equivalence assessments, IAIS deliberations, and the outcome of its implementation within the European Community.

In particular, the confined geographical application of the Solvency II framework could generate opportunities for insurers to reduce the cost of capital through the use of reinsurance in non-equivalent jurisdictions with weaker solvency and supervisory standards, if the capital charges required against the credit risks introduced by such reinsurance contracts are not set appropriately.

Best Regards,

George Lekatis
President of the Basel iii Compliance Professionals Association (BiiiCPA)

Preliminary international banking statistics, second quarter 2011

oktober 20, 2011 in Publicaties, Treasury Algemeen by onze redactie

Preliminary international banking statistics – second quarter 2011

ING: Groeikansen bedrijven liggen in opkomende markten

september 26, 2011 in Nieuws by onze redactie

Dow Jones – ING Commercial Banking, onderdeel van ING Groep nv (INGA.AE), ziet voor Nederlandse bedrijven vooral groeikansen in regio’s buiten West-Europa en de VS, omdat in die twee traditionele markten op zijn best slechts matige groei wordt verwacht in de komende jaren.

“We hebben net een eerste onderzoek naar internationaal zakendoen onder onze clienten afgerond”, zegt directeur van ING Commercial Banking Annerie Vreugdenhil maandag tijdens een persconferentie. “Daaruit blijkt dat 87% van de ondervraagde bedrijven importeert en 79% exporteert, terwijl slechts 51% een samenwerkingsverband met een buitenlandse partner heeft en maar 34% een buitenlandse vestiging heeft. Daarbij is het buitenland in dit verband vooral op te vatten als Europa. Wij willen bedrijven de komende jaren meer gaan wijzen op de enorme kansen die het buitenland biedt buiten Europa, in de opkomende markten.”

Veel bedrijven koesteren de nodige terughoudendheid om een vestiging in bijvoorbeeld een Aziatisch land te openen en dat is volgens het onderzoek van ING nog niet eens zo onbegrijpelijk. Taal en cultuur zijn twee barrieres die moeten worden overwonnen, maar ook het gebrek aan kennis over wet- en regelgeving komt bij veel bedrijven steeds meer voorop te staan.

Vanderlande Industries, gespecialiseerd in transportsystemen en logistieke toepassingen, werkt al geruime tijd voor lokale opdrachtgevers in China. Deze ING-klant roemt de mogelijkheden die China biedt, alsmede de langdurige en intensieve voorbereidingen die Chinese klanten zich getroosten, voordat een project start.

Het is de ervaring van veel bedrijven en ook Vanderlande, dat in China veel waarde wordt gehecht aan een duurzame relatie. Dit dwingt Nederlands bedrijven ertoe veel geduld te hebben en bereid te zijn bij een start aanloopverliezen te nemen.

Het advies, ook van Vanderlande, luidt altijd met lokale partijen te werken als een bedrijf in China wordt opgezet. Dit is overigens geen verplichting die de Chinese wet stelt.

ING stelt dat bedrijven gebruik kunnen maken van de investeringen die de bank heeft gedaan in zijn internationale infrastructuur, met een spreiding over circa veertig landen. “Het netwerk ligt er en wij willen dit de komende jaren ten volle verder uitnutten”, aldus Vreugdenhil.

De investeringen in het internationale netwerk van de bank staat volgens Vreugdenhil geheel los van het omvangrijke desinvesteringsprogramma van het ING-concern, dat onder andere zijn gehele verzekeringstak gaat afstoten.

ING wijst op de groeimogelijkheden in Azie, maar ook in Zuid-Amerika, zoals Brazilie. “In Europa, maar ook de VS is de groei naar verwachting de komende jaren bescheiden. In Azie wekken jaarlijkse groeicijfers van 10% nog steeds geen verbazing en er is weinig reden aan te nemen dat deze percentages de komende jaren dalen”, zegt Vreugdenhil.

Maar voordat ze die aantrekkelijke groeimogelijkheden kunnen benutten, willen bedrijven vooral juridische steun en daarin voorziet ING niet. Hiaten in de juridische kennis zijn er bijvoorbeeld op het gebied van arbeidsrecht, milieu-eisen en financiële en fiscale regels.

Illustratief is dat zowel de ING-woordvoerder als zijn klant geen antwoord kunnen geven op de vraag, of een bedrijf met winstgevende dienstverlening in China de winst kan uitkeren aan aandeelhouders buiten China. Ofwel: krijg je je geld het land wel uit? “We hebben geen idee of dat kan of niet,” aldus Vreugdenhil.

Ook Vanderlande blijft het antwoord schuldig. Net als veel andere ondernemingen met activiteiten in China, betoogt het bedrijf dat besteding van de winst buiten China niet aan de orde is, omdat de strategie gericht is op groei in China en het verdiende geld dus weer lokaal wordt geïnvesteerd.

Door Andre Sterk; Dow Jones Nieuwsdienst; +31-20-5715201; andre.sterk@dowjones.com

(END) Dow Jones Newswires

The next steps of the Basel III implementation

augustus 11, 2011 in Columns-Artikelen by George Lekatis

Today we will study a really important paper, that leads us to the next steps of the Basel iii implemantation. The paper we will study is titled; ‘Next Steps in Financial Regulatory Reform’ with remarks by Daniel K. Tarullo, Member Board of Governors of the Federal Reserve System, at the George Washington University Center for Law, Economics, and Finance Conference – Washington, D.C. November 12, 2010

Basel III makes a number of important changes to address deficiencies in the pre-crisis capital rules:

First, reflecting both intuitive good sense and market realities during the crisis, Basel III creates a new minimum common equity capital requirement.

Moreover, the agreement provides a definition of common equity that will prevent firms or national regulators from including in the calculation of common equity certain assets that could dilute its loss-absorbing character.

Second, the minimum common equity ratio will be set at 4.5% of risk-weighted assets, with an additional requirement for a 2.5% “conservation buffer.”

The minimum ratio should be understood as defining the amount of common equity needed for a firm to be regarded as a viable financial intermediary.

The conservation buffer is a new feature of capital regulation, intended specifically to reflect the losses that a firm may suffer during periods of financial stress

Thus the concept behind the two-level requirement is that a banking organization should be able to withstand losses associated with systemic stress and still be a viable financial intermediary.

This concept is comparable to the approach we adopted during the Supervisory Capital Assessment Program (SCAP) in early 2009.

There, you may recall, we used a special stress test to provide a rough estimate of losses that the large banking organizations could face in an adverse scenario and asked that they hold capital sufficient to absorb those losses and still be above common equity levels that would maintain the firms as viable intermediaries.

There is no direct way to calculate how much equity is needed to assure markets that a banking organization is viable.

In our internal analysis at the Federal Reserve in preparation for the Basel Committee deliberations, we analyzed distributions of actual losses suffered by larger institutions over the last several decades, on the assumption that an institution that could withstand such losses at a high confidence level would be regarded as a viable going concern.

For the conservation buffer, we looked at actual pre-SCAP losses incurred by large banking firms during the recent stress period and SCAP estimates of additional losses associated with the recent stress period.

Both these determinations required considerable judgment, and thus we developed ranges, rather than point estimates, for the levels we thought reasonable.

In particular, government capital injections and debt guarantees in the fall of 2008 complicated the estimation of losses that might have been incurred in the absence of too-big-to-fail support.

The ratios agreed to in Basel were at the lower end of, though still within, the ranges we had calculated.

The practical effect of the two-level approach is that banks under stress may let their common equity ratio drop below the 7% level that is the sum of the minimum and buffer requirements.
However, restrictions on capital distributions will result, which will become progressively more stringent as the common equity ratio drops closer to the 4.5% minimum.

The buffer is thus designed to forestall banks from continuing to pay dividends even as they come under stress, a practice observed in some institutions during the financial crisis.

Realistically, both regulators and markets will expect firms generally to maintain their common equity ratios above 7%.

Third, Basel III makes extensive changes to the risk weights assigned to a financial institution’s traded assets and counterparty exposures.

The market-risk requirements of the pre-crisis capital regime were woefully inadequate.

In many instances, they simply did not reflect the actual risk assumed by an institution. They also created an invitation to arbitrage credit risks by turning them into traded assets with lower risk weights.

It is also noteworthy that the changes in risk weights incorporate some elements of a macroprudential perspective as, for example, in higher capital requirements on equity investments in other financial firms and credit exposures to large financial firms.

Fourth, Basel III provides for a minimum leverage ratio, roughly similar to requirements already applicable under national law in the United States and Canada.

While the terms of this leverage requirement have been agreed to, there will be a supervisory monitoring period and then a parallel run to assess its impact, particularly in countries with no history of such a requirement, and to provide for adjustments if warranted.

Finally, Basel III has a rather lengthy and complicated transition period, with the new requirements to be phased in between January 1, 2013, and January 1, 2019.

We favored a significant transition period, so as to allow firms flexibility in adjusting to the new regime through such means as running off higher risk-weighted assets, adjusting their business models gradually, or using retained earnings to add any new capital that might be required under the new rules.

To be honest, however, we did not think the transition period needed to stretch over eight years. In fact, it appears that most U.S. banking entities expect to meet the new requirements considerably sooner.

However, the lengthy transition period was an important inducement for some countries to agree to the new, much stronger standards.

With the agreements reached in July and September at meetings of the Governors and Heads of Supervision (GHOS) in Basel, the structure and basic elements of Basel III are clear.

Details are still being worked out by the Basel Committee. Then, of course, the U.S. banking agencies will need to implement Basel III through domestic capital regulations. Basel III is not a perfect agreement, of course.

There are things we would have done differently if we were writing a capital regulation on our own.

There will surely be some technical challenges in implementing it. It does not really address some pre-crisis problems in capital regulation such as pro-cyclicality.

But it is a major step forward for capital regulation.

It will raise minimum requirements substantially, ensure that regulatory capital is truly loss absorbing, and discourage some of the risky activities for which the pre-crisis regime required far too little capital.

Basel III was also a major step forward in international cooperation.

In all candor, as recently as this past spring I was concerned that we might be unable to agree in Basel on such key issues as a distinct common equity requirement.

However, under the strong leadership of Nout Wellink in the Basel Committee and Jean-Claude Trichet in the GHOS, we concluded a good agreement in timely fashion.

I believe that another factor in turning things around was that, unlike at some times in the past, the U.S. banking agencies spoke with a single, unified voice in Basel.

Obviously, the benefits of Basel III for financial stability will be realized only if they are implemented rigorously.

In this regard, it is important to draw a distinction between, on the one hand, implementation in the sense of enacting national regulations that incorporate the Basel standards and, on the other, implementation in the sense that firms are actually holding the amounts of capital called for by the internationally agreed rules.

The Basel Committee must be able to monitor effectively implementation of, and compliance with, these new capital standards.

A number of market analysts have noted that, even under current market risk capital rules, there is considerable apparent variation in the riskweightings apparently applied by different banks.

We are urging the Committee to explore mechanisms for ensuring that these strengthened capital standards lead to a consistency in application, as well as in the provisions of relevant domestic regulations.

Along these lines, we have heard complaints from a few other countries that Basel II is not yet operative for our large, internationally active banking organizations in the United States.

As we have explained, despite the substantial resources devoted by both banking organizations and supervisors to the tasks of developing and validating the Advanced Internal Ratings-Based Approach in those institutions, we continue to encounter significant difficulties.

The suggestion that U.S. banking organizations have thereby gained a competitive advantage is misplaced, however.

For one thing, we required significant capital increases as part of the SCAP and the Troubled Asset Relief Program repayment processes last year.

Also, we note that the required capital levels for some foreign banks adopting Basel II apparently declined from Basel I levels.

In-depth oversight by the Basel Committee of implementation and compliance would allow supervisors from all member countries better to understand issues such as these.

I suspect it would also result in supervisors learning from one another and thus improving the quality of large institution capital regulation globally.

Although, fortunately, Basel III does not present nearly the degree of technical challenge posed by the advanced approach of Basel II, there will still be a good bit of opaqueness in how some of its components are implemented and thus a continuing need for significant monitoring by the Basel Committee.

One piece of unfinished business on the international capital regulatory agenda arises from the agreement by the GHOS in September that systemically important financial institutions should have loss absorbing capacity beyond the Basel III requirements.

This international position parallels the Dodd-Frank requirement that the Federal Reserve apply capital requirements to large, interconnected financial institutions that are more stringent than those applied to other banks.

We think it serves U.S. interests to develop our plans for implementing our domestic statutory obligation in tandem with our participation in this international process, so as to maximize the chances of convergence of international standards and our own practice.

Work on this issue in the Basel Committee and the Financial Stability Board will continue well into next year. The final set of major Dodd-Frank regulations will not be completed until early 2012.

A year from now we will be in the midst of a regulatory process implementing Basel III, and there will likely be an active debate over the future of the government-sponsored enterprises.

So I see little risk that the third annual conference on financial regulatory reform at George Washington Law School will be entitled “The End of the Road.” Maybe, just maybe, I have given you a title for the fourth. Then again, maybe not.

Best Regards,

George Lekatis
President of the Basel iii Compliance Professionals Association (BiiiCPA)

The role of banks in the economy  improving the performance of the New Zealand banking system after the global financial crisis

augustus 11, 2011 in Cash Management, Publicaties, Risico management, Treasury Algemeen by onze redactie

The role of banks in the economy  improving the performance of the New Zealand banking system after the global financial crisis