Good morning Ladies and Gentlemen,

A. I would like to thank the organisers for inviting me to participate in this important event, as I consider business information services a key and integral factor of risk management. In my brief intervention I will state my views on the actions of Greek banks, and the Bank of Greece (BoG) as the competent supervisory authority, concerning the management of credit risk.

I think that two issues are closely linked to this:

1. The first issue is the strong credit expansion, mainly to households in Greece, although Greece still has one of the lowest rates of household indebtedness in relation to GDP in the euro area (33% against an average of 55%).

While these developments are also evident in the EU and the euro area, in the case of Greece some specific structural factors exist, such as:

(i) the strong GDP growth in recent years,

(ii) the reduction of Bank of Greece reserve requirements on credit institutions in order to comply with the euro area Monetary Policy, a factor that increased credit institutions' liquidity,

(iii) the reduction of nominal and real interest rates, which has reduced the cost of borrowing to historically low levels in the euro area, and more significantly in our case the fact that interest rates were a multiple of interest rates of the core European countries and

(iv) the increase in competition, following the recent abolition of the remaining restrictions on credit, such as the ceiling applying to non-collateralised consumer loans, which took place not earlier than 2003.

The cumulative effect of these developments may have an impact on credit risk management requirements.

 At the same time, a discussion is held at the EU and euro area level (Banking Supervisory Committee - BSC and Committee of European Banking Supervisors - CEBS) about the role of competition on the possible loosening of credit standards, the possible correction of housing prices in some countries, the ability of households to service their liabilities and the adequacy of provisions under different macroeconomic and credit cycle scenarios.

2. In this context, the second related issue is the Basel II Αccord and the need to align supervisory measures with the risks taken by banks.

3. Having presented the changes taking place in the banking landscape and affecting credit risk, I would like to briefly review the regulatory context in which banks and supervisors work and then elaborate on the challenges that they face.

****

B. Αs already known, the overriding objective of the Basel II Accord is to establish a more risk-sensitive capital requirements regime, by aligning the regulatory capital requirements with the underlying risks run by banks and providing institutions with incentives to improve their risk management systems.

For today's presentation, suffice it to focus on the three approaches of Pillar I and the challenges posed thereby in the calculation of the minimum capital that each bank should hold against credit risk.

1. The Standardised Approach is quite similar to the Basel I method, but more risk-sensitive, since it introduces the differentiation of risk weights on the basis of the category of the obligor and his creditworthiness, as assessed by recognised rating agencies. It also recognises the use of credit risk mitigation, in the form of collaterals (e.g. various types of securities), which leads to the reduction of the undertaken credit risk.

2. The Foundation Internal-Ratings-Based Approach (F-IRB) relies on the internal ratings systems of banks for classifying customers at rating categories on the basis of the probability that they will not honour their obligations.

While in F-IRB the Probability of Default (PD) is estimated by the bank, the Loss-Given Default (LGD), the Exposure at Default (EAD) and the Maturity (M) of the exposure are set by the framework.

3. The Advanced Internal-Ratings-Based Approach (A-IRB) is actually an extension of the F-IRB, allowing the credit institution in question to calculate (besides the PD) the parameters of the Loss-Given Default, Exposure at Default and Maturity.

4. The use of the IRB Approaches is subject to supervisory approval under strict conditions to ensure that the systems used are of the appropriate quality.

The new framework is designed for broad application, as it satisfies different levels of requirements, by providing the three alternative approaches, which incorporate different levels of sophistication.

It is, therefore, not only an opportunity, as is often said, but also a necessity for credit institutions, irrespective of their size and complexity, to upgrade their risk management systems.

***

C. Moving to the challenges posed by the new framework, a number of issues have to be resolved before Pillar I can be efficiently applied. The three most important and partly interrelated issues in this respect are:

•  the choice of method,

•  the data availability and

• the establishment of risk management systems and a risk culture by banks

1. First of all, with regard to the choice of the approach that best suits each bank, the Bank of Greece had to make a decision on whether it would impose the adoption of a specific approach to the supervised credit institutions depending on predetermined criteria, like the approach applied in the USA, or only set the minimum requirements for upgrading Risk Management Systems, irrespective of which approach credit institutions would choose to adopt.

The Bank of Greece (BoG), both in the dialogue with banks and in its 6 consultation papers, stressed, on the one hand, that each credit institution would have the discretion to adopt the approach it prefers and, on the other hand, that no deviation from the best practices of risk management would be condoned.

However, the BoG is obliged to encourage banks to move to the IRB approach more quickly, as it feels that this approach will lead to a further improvement in their risk management, thus strengthening their competitive position and especially their ability to successfully address risk management challenges.

2. The second challenge I would like to address is the availability and suitability of the data used for the calculation of the parameters of PD and LGD.

It is certainly not an easy task for banks and authorities to establish a reliable relationship between capital requirements or Non-Performing Loans (NPLs) and the economic cycle in Greece, because of the relatively recent deregulation of consumer lending as well as the fact that Greece has not experienced a full economic cycle since the onset of the financial liberalisation in the mid-90s.

This task becomes further complicated when calculating downturn LGDs, as required by the new framework.

An issue closely connected with the collection of information is the further enhancement of the credit bureau. Having recognised the importance of this issue, amongst many new measures, new categories of loans (mortgages) will be reported in the very near future. In addition, some progress has been made in finding the appropriate balance between the protection of customer rights and the need to enhance the ability of banks to assess customers' creditworthiness.

The above highlight the difficulties Greek banks face in collecting information for their rating systems. However, to be fair, I have to say that other countries too face similar problems, underestimating, to some degree, expected losses. It is also true that data availability touches upon the availability of ratings for corporations.

In the context of the new capital adequacy framework, credit institutions may use external credit assessments to determine the risk weight of their exposures, provided by external credit assessment institutions (ECAIs) that are recognised as eligible for that purpose by the competent supervisory authorities.

This recognition is only granted if the competent authorities judge an ECAI to meet the criteria of objectivity, independence, ongoing review of the its methodology as well as transparency and disclosure, credibility and market acceptance of the ECAI's individual credit assessments.

Given the low number of rated firms in the EU and especially in Greece, the BoG would welcome the operation of national agencies fulfilling the criteria of the Capital Requirements Directive (CRD) and is committed, within the CEBS (the Lamfallusy 3 Level Committee of supervisors) convergence plan, to ensure common standards that will allow free access to new agencies.

***

3. To wrap up the challenges that banks face with regard to the Basel II implementation, I will elaborate on the third issue of the establishment of the credit risk management system and practices.

In recent years, most of Greek banks have taken major steps towards upgrading their credit measurement and management systems, as they have realised that the benefits of Basel II offset the implementation cost, through the full implementation of more advanced approaches is a difficult and lengthier process than anticipated. Several banks have already introduced credit scoring models for the approval of applications for consumer loans, while some banks have developed probability-of-default models.

In particular the issues that banks have been addressing are:

• The design of their rating systems and, more specifically, technical characteristics such as expanding the risk rating grades in order to avoid granularity. The majority of Greek banks (accounting for 76.6% of the total commercial bank assets) have already adopted 10 grades in their rating systems, while the remaining ones are moving in this direction, making all the necessary adjustments.

• The establishment of the criteria and methodology of classification according to the category of each portfolio, the use of internal ratings and the incorporation of financial parameters. Banks representing 53% of the total commercial banking assets (at 31 Dec. 2004) use external software to rate large corporations.

• Banks also use scorecards to either approve or reject an application for loan but there may be some room for improvement in this area, too.

• Besides the adoption of the necessary methodologies, another significant issue is the establishment of the required IT facilities.

This poses one of the most important practical challenges, as IT infrastructure does not only cover the collection, storage and accessibility of the data needed for PD and LGD estimation, but also reporting requirements.

This enables banks' risk management and internal audit functions and of course helps supervisors to fulfil their task of validating credit management systems as a whole, while it covers the information requirements set by Pillar III.

Regarding risk culture, what will undoubtedly play a very important role for the acceptance or not of a bank's system will be its systematic use. In other words, the minimum expectations of the BoG are that the systems used for capital requirements will also be used in approving credit lines or loans, setting limits, measuring and reporting credit risk, making provisions and, where appropriate, pricing.

D. From the supervisor's point of view, it is clear that banks' risk management has improved significantly over recent years, but the degree of improvement is uneven across banks, including cooperative banks.

We are well aware of this factor, as well as that fierce competition will possibly result in some volatility in credit standards, which will most likely worsen loan quality if the current high rates of economic growth decelerate.

The Bank of Greece has already responded by taking some proactive measures. More specifically:

• It has increased the coefficients of supervisory provisions in 2003 and again in 2005. The issue of supervisory provisions has been extensively discussed. We believe that, especially for banks that will stay in the standardised approach, the buffer that those provisions offer should remain. For banks that will move to the more advanced approaches, supervisory provisions should not differ from their expected losses as estimated by their systems, but some buffers may be necessary until the system has been tested in real life.

• In addition, the BoG encourages banks to maintain a buffer above the minimum capital adequacy requirements, in order to shield their net positions against the potential increase of credit risk. Moreover, it has imposed a required capital adequacy ratio higher than the minimum of 8% (up to 3 p.p. for some banks) when the risk profile of individual banks had made this measure necessary. In this respect, in principle and not in the details, we have already adopted the Pillar II approach of Basel II.

In the spirit of the new framework, which requires individualised supervision, the BoG:

• monitors and validates more closely the level of sophistication of banks' risk management systems with targeted on-site inspections and reviews banks' internal control systems and their internal policy on loan classification (for instance the cut-off points in granting new loans) and the parameters used (income of borrowers in relation to their monthly instalments).

• encourages banks to improve data collection and statistical methods, so that they move from the assessment of transactions to that of borrowers , and

• has initiated an on-going dialogue with the Risk Management Units of each bank, in particular for sectoral or large individual exposures, taking into consideration stress-testing exercises.

On the issue of mortgages, I should emphasise that strong competition may induce banks to loosen their credit standards by increasing the LTV ratios to a level that could be no more in line with the favourable, lower than the 8%, minimum capital requirements prescribed by Basel I (4%).

E. Concluding, I would like to convey a clear message: Despite the considerable improvement in the risk management of banks, complacency for what has already been achieved over recent years may be regarded as an additional risk factor.

The clear improvement of risk management may not be sufficient, given that at the same time the ever growing complexity of bank products and some credit or market risk factors oblige all of us to keep pace with the new environment and dynamic developments.

For example regarding the credit derivatives and hedge funds, greater effort is needed by those banks that are very active in this field to track the risk that each one faces. However, as Greek banks have not resorted to these instruments to the same extent as many foreign banks, the situation in Greece is not, for the time being, a matter of concern.

In addition, not only factors such as oil prices may affect the risk profile of banks, but also other factors that may emerge, such as:

- the fact that the risk profile seems more likely to deteriorate when the economic environment is becomes unfavourable (procyclicality effect).

- the narrowing spreads due to competition resulting from low interest rates and ample liquidity, which may induce banks, by hunting the yield, to misprice some risks.

- in the medium to long term, possible difficulties to increase and maintain the interest income, a fact which is also evident in other economies.

 However, I believe that, as in the past, Greek banks have demonstrated their ability to adapt to challenges and that they will also successfully address these issues on a proactive and forward looking manner, by building on the improvement already made and further enhancing their risk management to the benefit of financial stability and economic growth. Most of the banks with substantial portfolios and cross-border presence will be ready, in the time frame provided for in the Capital Requirements draft Directive, to adopt the IRB approaches, making the required additional effort in adapting their systems.

Let me thank you for your attention and wish you success in the goals of the conference.