Reden und Präsentationen

Ladies and Gentlemen,

I am honored to have been invited once more to speak at an important SUERF event. Let me use the opportunity of meeting in Viennato say a few words about how I see the relationship between the Oesterreichische Nationalbank (OeNB) and SUERF.

SUERF and the OeNB have had a longstanding (since the 1980s) productive and cordial relationship. In 2000, our cooperation reached a new quality. In April of that year we welcomed the SUERF General Secretariat to the premises of the OeNB and we supplied one of our most experienced staff members, Mr. Hochreiter, to serve as SUERF secretary general. 

We decided and embraced this package because of our trust in the research and network competence of SUERF. SUERF’s uniqueness that equitably brings together the academic profession, central and commercial bankers, has been especially dear to us. The maintenance of this balance remains a cornerstone for our continued support. Of course, housing the secretariat and covering most of the running costs have been a substantial investment for us. I am pleased to conclude that the return on investment for the OeNB has been way above average and above our expectations.

The Bank and I personally very much look forward to the further development of our collaboration. As today’s venue testifies, our collaboration also benefits the Austrian financial community.
 

Ladies and Gentlemen,

As you are well aware, one of the key areas of research in the OeNB centers on Central and Eastern European countries in general and on the New EU Member States in particular. We believe that we do have a strategic research advantage in this region as you can see from a broad range of state-of-the arts research output and from our conferences. If you want to know more, please visit our website www.oenb.at. Let me just mention one upcoming event, which might be of specific interest to you. This year’s “Conference on European Economic Integration”, which will take place in Viennaon November 14 and 15, will be jointly organized by the ECB, the Center for Financial Studies (CFS), and the OeNB and will deal with financial integration and in Central, Eastern and South-Eastern Europe”.

Our interest in the CEE area of course is also driven by the intense engagement of the Austrian financial sector. Keeping an eye on the risks implied for the Austrian financial sector both in the shorter and in the longer run certainly is a good investment for the central bank.

In a first step, I will re-call the institutional framework that has been playing an important role in the financial sector development in the New EU Member States in recent years. Accordingly, I will provide a brief outlook for the further integration process in the future.
Next, I will highlight key features of the financial sector in the New EU Member States and in the EU-15 or the Euro area (EU-12), presenting some of the findings of our own research.
Then, I will show the results of an OeNB study on the current state and development of banking efficiency in the NMS.
Finally, I will make some concluding remarks.


Monetary integration process of the NMS

Since their accession on 1 May 2004, all the New EU member states have already participated in the EMU (Economic and Monetary Union), but not to the full extent. Rather they have the status as „Member States with a derogation“. This means – inter alia – that their exchange rate policy has to be treated as a matter of common interest. 

The next stage of monetary integration after accession is participation in the ERM II.In legal terms, (new) EU member states outside the euro area have the right to enter ERM II participation at any time upon request, but subject to a multilateral agreement on the level of the central rate and the bandwidth. In December 2003, the Governing Council of the ECB and, thus, the Eurosystem took the policy position that it would be necessary that major policy adjustments are undertaken prior to participation in the mechanism in order to ensure a smooth participation in ERM II.

The third stage of the monetary integration of the New EU member states will be the adoption of the euroand the full participation in the EMU after the fulfilment of the EC Treaty’s convergence criteria. The decision on the fulfilment of the necessary conditions for joining the euro area will be taken:

• by the Council (ECOFIN) by a qualified majority
• on a recommendation from the Commission
• after consulting European Parliament and discussion in European Council
• based on convergence reports by Commission and ECB             

Let us take a brief look at the current official strategies for full monetary integration. In the meantime, six out of ten New EU member states have already started to participate, while the four larger economies, the Czech Republic, Hungary, Polandand Slovakia, the so-called NMS-4, are still outside the mechanism. Given their target dates for euro adoption and the legally required minimum length of ERM II participation of two years before the convergence examination, these countries can be expected to enter ERM II within the next few years.

The authorities of all the NMS-4 plan to adopt the euro or, at least, to be fully able to adopt the common currency within the next five years. Their plans hinge in particular on the progress in fiscal consolidation to which the authorities have committed. What is the financial market’s assessment of the credibility of these plans? One way of answering this question, we can look at the 5y swap spread against Euro in 5 years, which should not comprise any premium for exchange rate risk if the market participants considered these plans as fully credible. Thus, the spread should be relatively close to the spread of foreign currency-denominated sovereign bonds (Eurobonds) of these countries, which currently range between 15 and 30 basis points.

This chart shows that the financial market participants expect virtually no exchange rate risk in 5 years for the Czech Republicand Slovakia, but price-in still a considerable exchange rate risk premium for Hungaryand Poland. However, in case of Poland, the official Euro adoption target dates have attained increasing credibility in the course of the last 12 months.

Smooth ERM II participation and – to a further extent – the fulfilment of the exchange rate criterion of the Treaty require that the volatility of the national currency against the euro remains within certain limits. This chart shows the exchange rate development against the euro in the NMS-4 in recent years. Among these four currencies there is only one currency, namely the Polish zloty, that had a volatility in recent years that would not have allowed for the fulfillment of the criterion (under the assumption of an appropriate central rate). It fits to this observation that only the Polish zloty has been following a free-float regime in recent years.

While the further monetary integration of the New EU member states constitutes one major element in the institutional framework for the future financial sector development, let us now turn to the current starting point, the status-quo of the financial sector in these new EU countries compared to the EU-15 or the Euro area (EU-12), respectively.


Key features of the financial sector

In this and the following slide, I would like to highlight current features of the financial structure in the New EU member states. It is well known that, in terms of size, all major segments of the financial sector are considerably lower in the NMS than in the EU-12, not only in absolute euro terms, but also in relation to GDP. One may expect the ratios of financial assets to GDP to rise, as a result of real catching-up and further deepening of financial intermediation. In terms of ownership, a central specific feature of the banking sector in the NMS is the pre-dominant role of foreign-controlled banks.

A look at the relative size of the different segments within each region’s financial sector confirms some similarities between the financing structure in the NMS and the euro area, like the dominant position of the banking sector (including as an investor in the domestic debt securities market) and the larger size of the debt securities market than the equity market.

However, I would like to draw your attention to three important differences: 

• First, the share of domestic loans to the corporate sector is higher in the NMS. Conversely, the share of loans to private households is lower. However, it is rising fast.
• Second, the relative size of the corporate debt securities market is much smaller in the NMS. However, given the upswing of the corporate debt securities market in the EU-12 since the start of the common currency area, one can expect that this difference will diminish as result of progress in monetary integration.
• Third, the relative size of the equity market is bigger in the NMS. However, the large equity market capitalization is mainly the result of the listing of privatized companies and of the increase in the equity index and not of the volume of capital-raising equity issues. Indeed, the SUERF study No 16, which was contributed by the OeNB (Reininger/Schardax/Summer: Financial System Transition in Central Europe), found that corporate debt securities constituted a more important source of corporate financing than equity issues in the Czech Republic, Hungary and Poland.             

The comparison of the banking sector performance in the EU-15 and in the New EU member states provides evidence that 

• Bank profitability is higher in the NMS than in the EU-15 (measured by ROE and ROA),
• And: Net interest income in the NMS is twice the level in the EU-15
• But: the cost-to-income ratio as well as asset quality ratios – despite their improvements in recent years – are less good in the NMS than in the EU-15.             

It should be noted that the figures for the EU-15 are calculated on a consolidated basis and, thus, already include EU-15 banks’ subsidiaries in the new EU member states, which mostly have relatively higher profitability.

Against this background, an OeNB study by Walko/Reininger (2004), which was published in the OeNB Financial Stability Report No 8 in December 2004, analyzed the sources of net interest income in the NMS-4 (CZ, HU, PL, SK) and entailed a stock-taking exercise on interest rate margins in these countries and in the euro area.

This study confirmed that a very important factor for higher net interest income in the NMS-4 (than in the EU-12) was the positive overall margin spreads against the EU-12 in the NMS-4. These spreads ranged from 1.6% (in CZ) to 5.7% (in HU).

Thereby, both lending and deposit margins are higher (in absolute terms) than in the euro area. With the exception of Hungary, the deposit margin spread is of particular importance, as it is larger than the lending margin spread (in absolute terms) in CZ and PL and equal to the lending margin spread in SK.

Moreover, disaggregation by economic sectors shows that both lending and deposit margins and margin spreads and, hence, also overall margins and margin spreads are particularly high (in absolute terms) in the business with private households. It is very probable that this is only partly due to higher costs in this market segment. Rather, it strongly suggests that the business with households has not yet been as competitive as the business with the non-financial corporate sector.

Positive margin spreads are a major, but not the only factor that supports the higher net interest income ratio in the NMS. Another important factor lies in differences in the balance sheet structure between the EU-12 and the NMS on the liabilities side:

• first, deposits from domestic residents play a much greater role in the NMS-4 than in the EU-12; this fact magnifies the effect of significantly negative deposit margin spreads between NMS-4 and EU-12.
• second, the share of capital and reserves is higher in the NMS-4 than in the EU-12, amounting to cheaper financing of interest bearing assets in the NMS-4.             

However, there are differences in the balance sheet structure between the EU-12 and the NMS on the asset side, which tend to lower net interest income (ceteris paribus):

• first, claims on general government have higher weight in the NMS-4 than in the EU-12 and
• second, the structural liquidity surplus in the banking sectors of the NMS-4 leads to sizeable investment in central bank instruments.             

Thus, the banking sector in the NMS does not exploit available margins to the full extent possible, which is a sign of inefficiency.

Up to now, I have sketched out the status quo of the banking sector in the NMS-4. In order to get some indication for the upcoming future development of the banking sector there, it is worth to look at the experience in the EU-15:

Not only did overall interest rate margins decrease over the 1990-ies in the EU-15, but also did overall interest rate margins approach the EU-15 average in the peripheral countries (like Greeceor Portugal) of the EU-15. Thereby, lending margins fell significantly in the EU-15 between 1997 and 2000, continuing a longer trend.

Among the factors leading to this decline in the overall margins, I would like to highlight the following:

• increased competition
• falling interest rate level and flattening yield curve
• charging explicit fees for some services
• reduction in banks‘ operating costs
• securitization             

It is obvious that one may expect similar developments also for the new EU member states.

Against the background of the experience in the EU-15, what are the challenges for the banking sectors in the new EU member states? 

• The positive margin spreads against the euro area in the NMS – or, in other words, the relatively high profitability of the banking industry in the NMS – is an important incentive to expand credit supply. Together with the catching-up related structurally strong credit demand (in particular from private households), this contributes to the high actual credit growth that we can observe currently in these countries. Thereby, the comparatively higher lending margin spreads in the household business are consistent with the relatively higher actual growth rates of loans to private households and the on-going rise in the share of private household loans in total domestic loans, starting from levels below those in the euro area.
• However, the increasing competition will likely erode interest rate margins (and margin spreads) in general and margins in private household business as well as deposit margins in particular. For instance, emerging alternative investment opportunities will put pressure on deposit margins.
• In order to compensate at least partly for the general decline in the margins, it is important to tap non-interest revenues to an increasing extent.
• Moreover, the reduction of claims against the government and the central bank could set free resources to exploit available margins to a larger degree. However, this will imply also higher risk. Given the still higher share of non-performing loans in the NMS than in the EU-15, any higher risk business needs caution and strengthening of credit risk management.
• Finally, the expected decline in interest rate margins renders the issue of banks handling their operating costs efficiently very important – not only for the banks themselves, but also for financial stability reasons, as banks’ ability to generate an adequate return in a competitive environment is a cornerstone for the shock absorption capacity of a banking system.             

In the next several slides, I will give you some evidence on the current state of both cost and profit efficiency in the banking sector of the NMS.


Banking efficiency

As raising efficiency will be the key to success for the banking sectors in the new EU member states, the OeNB conducted also a study that took a close look at the efficiency of the banking sectors in the new EU member states. Parts of this paper by Rossi/Schwaiger/Winkler were published in the OeNB Financial Stability Report No. 8 in December 2004. In the following, I will go over some of the main findings of this study.

The analysis of banks’ efficiency levels continues to be important from both a macroeconomic and a microeconomic point of view. From the micro perspective, the issue of banking efficiency is crucial in the face of increasing competition in a rapidly growing market. From the macro perspective, the efficiency of the banking sector influences the costs of financial intermediation and the overall stability of the financial markets. For the new EU member states, improvements in bank efficiency can have a significant impact on the allocation of financial resources since this sector still remains the most important external source of financing private investment of firms and households.

In this section of the presentation, I will provide evidence on three issues concerning the efficiency of the banking sectors in the new EU member states: 

• evidence on the level of and differences in cost and profit efficiency in the NMS
• evidence on the evolution of cost and profit efficiency over time as well as 
• evidence on banks’ management behavior in NMS             

The analysis itself is based on the so-called Stochastic Frontier analysis. To give you an idea about the methodology used, consider the following: Generally speaking, a bank is called efficient if it is not possible either to produce more with the same input or to reduce these inputs leaving the output unchanged. However, observable real data of banks often move away from this optimum. This is due to both bad luck and inefficiency. The stochastic frontier analysis tries to disentangle deviations from efficient behavior into an error term (bad luck) and a term that reflects true inefficiencies. Cost efficiency will only look at the way banks handle their operating costs, whereas profit efficiency focuses on both the cost and the earnings side, also taking into account net creation of loan loss provisions. 

In the present analysis, the investigation is based on a stochastic frontier model, in which the cost and profit functions are specified according to the flexible Fourier function, which is particularly suited to fitting empirical data. Roughly speaking, one takes the best banks within the NMS as the benchmark for evaluating efficiency.

Concerning the data covered, annual account data of 278 banks in the eight new EU member states from Central and Eastern Europeand Romaniaare included, with the observation period ranging from 1995 to 2002. (the Czech Republic (CZ), Hungary (HU), Poland (PL), Slovakia (SK), Slovenia (SI), the three Baltic countries Estonia (EE), Latvia (LV) and Lithuania (LT) as well as Romania (RO); Bulgaria misses for lack of data.).

The estimates on the overall sample for the whole period from 1995 to 2002 split by country are presented in the table you can see here. A value of 1 would mean perfect efficiency, values below one indicate inefficiency. Let me give you the broad picture of the evidence emerging from this table:

• The banking sectors of the NMS and Romania present, as expected, significant levels of cost and profit inefficiency, indicating that on average banks operate relatively far below the cost and profit efficient frontiers. 
  
• Inefficiencies are above levels usually encountered in EU-15 countries, which implies some potential for improvement as well as future competition.
  
• Efficiency levels vary considerably across countries. Statistical tests on the significance of these mean differences show that the variations in efficiency levels you can see here are significant in almost all cases.
  
• Overall, the profit efficiency scores are well below cost efficiency levels: NMS banks seem to deal much better on the cost than on the profit side. It should however be noted, that given the potential reward of maintaining or expanding market shares in a rapidly growing market, banks have little incentive to maximize profits by means of full utilization of their discretionary pricing power.           

Looking at the evolution of efficiency for the overall sample of banks, we can find a positive and significant increase over time of cost and profit efficiency, equal  to 6 and 4 per cent respectively – as you can see on the table here. This trend is also consistent with other research done on the issue (see for instance the contribution of Prof. Weill from the Universityof Strasbourgat the Conference on European Economic Integration in Viennalast year in this respect). The improvement is, however, not equally spread across countries. So, all in all, the picture brightens: we have comparatively low efficiency levels, but we see a convergence process towards higher EU-15 efficiency levels.

One focus of the recent contributions to the academic literature was to explain variation in efficiency across banks in the new EU member states in terms of their ownership yielding often mixed or inconclusive results. Whereas some authors provide evidence that higher bank foreign ownership is associated with higher efficiency, others show that privatization by itself is not sufficient to increase bank efficiency.

With the following analysis, I want to go beyond the relatively simple question of bank ownership and try to give you some indication on banks’ management behavior hidden behind the phenomena we observed so far in terms of efficiency. In order to do so, a Granger Causality approach is used. In this respect, four hypotheses on management behavior are tested for the sample at hand: 

1. The bad management hypothesis implies that bad managers do not adequately control for operating expenses and, thus, also poorly manage loan portfolio. 
   
2. The skimping hypothesis implies that banks that economize on monitoring costs to achieve higher efficiency will see their problem loans rise, as well.
   
3. The bad luck hypothesis means that unexpected and external factors, which stem in particular from the changes in the macroeconomic environment of the banks, increase problem loans, which reduces cost efficiency as monitoring costs have to be stepped-up. At the same time, the net creation of loan loss provisions has to be increased, as well, which lowers profit efficiency even further.
   
4. The moral hazard hypothesis suggests that managers of weakly capitalized banks are less risk adverse and engage in risk taking behavior.            

From a bank regulator’s perspective, each of these four hypotheses would of course entail a different set of regulatory implications. Whereas the bad luck hypothesis would highlight the need for regulators put an effort into insulating the banking system towards external shocks, the bad management hypothesis as well as the skimping hypothesis would indicate that supervisors should focus their attention towards bank-internal credit-risk management systems. The moral hazard hypothesis would alternatively suggest a close monitoring of banks with comparatively low capitalization levels.

The two equations of the model were estimated using the Arellano-Bond dynamic panel data model. Generally speaking, what emerges from the data is the relevance of loan loss provisions in explaining efficiency scores. In contrast to what we could have expected from the ownership literature, the evidence does not support the presence of the bad management hypothesis in the banks of the new EU member states. In the same way, neither evidence for the skimping hypothesis nor for the moral hazard behavior can be found for the banks in the sample. 

On the other hand, the results do provide some support for the bad luck hypothesis. Therefore, exogenous shocks on bad loans drive efficiency. 

From a regulator’s perspective, the evidence in favor of the bad luck hypothesis stresses the need to put an effort into increasing the shock absorption capacity of the banking sectors with respect to external shocks. Stress testing exercises spotting weaknesses in the banking sectors of the new EU member states therefore are especially important for both regulators and banks in these countries. Only these exercises can help spot and cure neuralgic points in the respective banking sectors. Furthermore, in terms of economic policy, this result implies that the prudent management of macroeconomic policy is of particular importance for the development and stability of the banking sector in the NMS. 

Let me put a special emphasis in this regard on the need for banks of the new EU member states to keep up in their development of risk management systems as they rapidly expand their loan portfolios. Otherwise, changes in external market conditions could prove harmful to banks that have merely focused on growth targets and neglected controlling the rising risk exposures going along. 

To conclude this section, I want to mention, as always, some caveats on these results. One reason why the evidence on bad management behavior, as cause for the lower efficiency of the banking sectors in the new EU member states, is scarce may be due to the sample size, which of course decreases with the number of lags used in the estimate, and the other data limitations in the database used for these countries. In fact, also data restrictions on problem loans should not go unmentioned, with several banks not reporting this figure. Therefore, further investigation with more data being available as time goes by, will certainly enable researchers to gain additional insights into the management behavior of banks in the NMS.


Concluding remarks

So, what can we conclude from the stylized facts and the findings of the empirical studies outlined above? 

• First, the financing structure in the NMS has already important similarities to that in the euro area, as it is pre-dominantly bank-based, with banks playing an important role also in domestic debt securities markets. The most important differences concern the still relatively lower level of private household loans and the less developed role of the corporate debt securities market.
  
• Second, the relatively better performance of the banking sector in the NMS (as measured by ROE and ROA) can be traced back to higher net interest income, which – in turn – has resulted primarily from higher margins. Margins are generally higher than in the euro area and this is true even more for margins in the private household business and for deposits.
  
• Third, for the time being, the high net interest income ratio has more than compensated for the fact that the banking sector in the NMS does not exploit available margins to the full extent possible, has higher cost-to-income ratios and has worse asset quality ratios. This is confirmed also by an empirical study that finds generally lower levels of cost and profit efficiency for the banking sector in the NMS than in the EU-15, certainly with quite significant differences across countries. Moreover, there are large gaps between cost and profit efficiency – banks in the NMS seem to be more efficient in controlling costs than in generating profits. 
  
• Fourth, however, the results also reveal a significant tendency of efficiency (both cost and profit efficiency) to increase over time in this country group. This is all the more important, as one can expect the margins to decline towards the euro area average, in particular as a result of intensified competition, given past EU-15 experience.
  
• Fifth, looking at the managerial behavior behind these phenomena, we find no evidence for bad management hypothesis explaining the relationship between efficiency and loan quality. However, we do find evidence for the bad luck hypothesis, which means that external factors have increased bad loans, which in turn has increased inefficiency.These results should focus the attention of the regulatory authorities even more on spotting and controlling exposures of banks to external shocks. In addition, this result highlights the importance of prudent macroeconomic policy for the development and stability of the banking sector in these countries.
  
• Finally, the successful completion of the on-going monetary integration process will further support the development of the financial sector in the new EU member states and will enhance European financial integration. This, in turn, will additionally improve the financing conditions for the economic agents in these countries. In particular, it will probably foster the issuance of corporate debt securities, which would result in further alignment of the financing structure in the new EU member states with that in the euro area.           

Thank you for your attention!