EUROPEAN WORKING GROUP on OPERATIONAL RESEARCH in BANKING
Eurobanking 2004, Budapest May 16-19
| Eurobanking EUROPEAN WORKING GROUP on OPERATIONAL RESEARCH in BANKING |
|||
Eurobanking 2004, Budapest May 16-19 |
|||
| Paper Summaries | Conference Report | Conference Picture Gallery | |
Conference Picture Gallery - Mladen Stariha, photographer
|
A serious conference requires serious socialising
|
And good wine is very important
|
|
In a comfortable environment |
Between the main conference presentations
|
|
and the committee meetings |
which happen very early in the mornings |
|
Then there are the farewell sessions |
And some time to relax |
| Paper Summaries | Conference Report | Conference Picture Gallery |
Eurobanking 2004 Highlights
32nd Eurobanking Meeting, Budapest, Hungary, 16-19 May 2004 Hosted and organised by MKB
Introduction
Issues and trends
The papers that were presented touched on a variety of topics which were considered to be current practical issues by the speakers. Drawing on these presentations, the discussions that followed and the informal discussions throughout the Meeting, a number of issues and trends were noted which are briefly described - not in any particular order:
1. Credit risk has always been the most fundamental of the risks facing financial services organizations. Hence its core position in Basel I & II – the backbone of supervisory regulation - and the number of Eurobanking papers each year, viewing credit risk from all sorts of different perspectives. However, despite its ‘popularity’, it seems that there is still ample room for improvement in analysis and understanding, and how this is applied to doing business.
2. The 230-page final text of Basel II - the new capital adequacy framework - has just been released, incorporating the adjustments, among others, proposed in October 2003 and January 2004. As a presentation on the implications of recent developments has shown, such adjustments - made on the basis of further research, QIS 3 results, focused efforts to solve key technical problems and continued discussions with banking organizations - seem to have made Basel II perhaps somehow more palatable.
However, if smooth implementation of an effective and efficient supervisory regulation system is to be achieved, one cannot help but point to the ample room that exists for further improvement, including the achievement of a better balance between comprehensiveness-sophistication-fairness and complexity, and between control and market reality, as well as of an even closer coordination of all related authorities. The need to stress more market reality as regards the International Accounting Standards (IAS) was discussed by another presentation.
Within the framework of supervisory regulation as well as of the wider ‘regulation’ pattern (referred to in Eurobanking 2003), we had a number of other papers including one on their impact on IT systems, two papers on operational risk (compared to three last year), two papers on liquidity risk management, and three papers on payment systems.
Although financial institutions have long been preparing for Basel II, we have yet to see a paper comparing business strategy options against the background of changing regulatory, social, demographic and other key factors.
3. Supervisory authorities continue to seek, basically within the proposed Basel II/Pillar 2, effective and efficient ways to forestall and manage potential credit and other risk crises. At the same time, one notes an active interest in country risk shown by both government and private lending institutions.
4. Despite the increased emphasis placed on the need for corporate credit rating and the Risk Management expertise accumulated by top European banks, a paper has yet to appear presenting a financial institution’s selling Risk Management packages to corporate customers. Although such a strategy might help retain expert Risk Management professionals (who are very busy at least with Basel II), related strategic issues might involve a potential conflict of interest, as well as infringing on the turf of consulting firms offering such services, either directly or within the framework of Corporate Governance.
5. Several papers (e.g. on integrated Risk Management, scoring systems, customer profitability systems, IT response to regulation, and a wider framework for Risk Management) have again identified the pressing need for clear and consistent conceptualization as well as for coordination, consolidation and integration of systems, both internally within Risk Management and externally with other systems. Two papers emphasized the two-way link between Risk Management and Strategy, one of which clearly incorporated business risk in Risk Management.
Thus, one may anticipate more papers on Enterprise-wide Risk Management (ERM), still considered to be the ‘Holy Grail’ of Risk Management; integration of Risk Management with different systems within the organization; the Strategy/Risk Management link and treating business risk within Risk Management; and benchmark wider frameworks that help in adopting ‘the-big-picture’ view, thinking on a higher abstract level, locating weak spots, suggesting directions and priorities as well as providing a common focus for Risk Management professionals and non-expert business executives/users.
6. Bearing in mind the long string of studies that confirmed a low record of success of M&As (and the relatively better record observed in the banking sector), few recent studies pointing to possibly better results, and the relative and perhaps temporary subsiding of the recent wave of consolidation, it looks as if more M&A decisions are now being more carefully taken and planned to ensure success. This may be the result of the experience gained and the probability that in a number of countries – apart from some ‘adjusting’ decisions - any significant M&As would have to be of a cross-border nature and/or ‘mega’ size. One may also note the expected consolidation of ‘smaller’ banks that will probably be triggered by Basel II. Issues discussed during the Meeting related to the various types of ‘alliance’/consolidation between banks and insurance companies, the post-privatization situation in Central European countries like Hungary, and the wide range of quantitative methods techniques and tools that are currently available for evaluating M&A opportunities.
One might expect papers on such techniques/tools, a comparison of national/domestic Vs international consolidation strategies, and behavioral aspects of M&A decision making. Of particular interest might be a paper on HR Management exercised at the M&A implementation stage, and specifically the impact of HR strategies and practices relating to certain areas (e.g. selecting and retaining key staff, relocating/exporting jobs, changing job/work patterns) on staff loyalty/morale/motivation and eventually long-term financial performance.
7. Payment systems – the bedrock of business for most financial institutions - continue to be a crucial topic in financial services, with issues such as the successful implementation of self-regulation efforts (as opposed to supervisory regulation attempts) to integrate and continuously innovate European payment systems, security problems and approaching retail payment pricing in a novel way.
8. In our rapidly moving world of technology and increasing regulatory requirements and consulting packages, the full costs and benefits of any IT investment continue to be examined thoroughly by financial institutions. The same applies to any proposed model or system that our business executives are expected to use – whether this is required by regulatory authorities or not.
9. The perennial problem of building and ‘selling’ simple, yet robust, models/decision tools was noted once again. Simple-looking and user-friendly models seem to have been easily sold to non-expert executives/users, while the highly technical treatment of other models would put such users off. Whereas the complexity of life cannot always be simplified, and robustness is the first criterion of a good model, such considerations must always be kept in the background and be presented to non-expert executives/users in an appropriate and most convincing manner, if the model/tool is to be sold and used knowingly i.e. not by default or out of necessity. Furthermore, it was noted again that one must always consider the people/management aspects of our systems, and – above all – never forget that models and systems are tools that help the decision maker to make up his/her mind, using judgment in the final analysis.
10. The communication gap between Stream-A (traditional Eurobanking areas) and Stream-B (Risk Management) participants - which was first observed during a Eurobanking 1999 plenary discussion and was further discussed in Eurobanking 2001 – seems to be gradually but steadily receding. This is shown by the increased inter-penetration of papers between the two streams and the presentation of more papers that ‘bridge’ the gap and thus could be included in either stream. Both papers on customer performance used risk-adjusted performance measures, while the two papers on capital allocation/management clearly ‘bridge’ the two streams by merging risk-based capital concepts with traditional management accounting and financial management approaches. The same applies to the single papers presented on hedge accounting, IT’s response to Basel II and standards, operational risk, integrated Risk Management, and the wider framework for Risk Management, not to mention the papers on rating and scoring. Furthermore, there were four papers viewing pricing from different perspectives. Thus, one may well discern a blurring of the dividing line, or convergence of the two streams, leading to more effective and efficient systems and tools as well as to a much needed risk consciousness or culture within our organizations.
| Paper Summaries | Conference Report | Conference Picture Gallery |
Antoine Frachot (1) of CA/CL presented the formal overall approach to operational risk as developed and used in CL, which was prompted internally in the late 1990s - not as a response to Basel II drafts, which have subsequently strengthened the function further. Loss data collection began in 1999 and by 2001 such data were used to calculate capital charges. The 3-step process used includes risk identification/assessment, prevention/mitigation, and monitoring/control. Both quantitative and qualitative approaches and respective problems were described and discussed, work in progress was referred to, and four useful references were provided in slide # 25.
Silvano Silvestri (2) of UCI presented a risk integration approach, which allows efficient capital allocation, affords Risk Management the opportunity to play a strategic role and leads to measurement of risk categories not yet quantified. Integration at UCI moves vertically – across all risk types within each Business Unit (BU) and the whole Group; and horizontally – across all BUs to assess a bank-wide risk measure. Risk integration is achieved in three stages: (1) Risk analysis to identify the risks assumed by each BU, using a three-step process; (2) Development of measurement models for new risk typologies, giving priority to business risk, which is generated by the bank’s servicing activity; and (3) Development of an integrated risk model in two phases: Top-down (4-step development of a model based on VaR for each already available risk type); and Bottom-up (development of a more precise model on simultaneous variations of a set of risk drivers affecting fair values of financial instruments, using correlation matrix, copula functions, and Monte-Carlo simulation).
Ferenc Muller (3A) of MKB overviewed the M&A developments in Hungarian financial services, referred to the recent privatization/consolidation developments, the current ownership structure of the 38 Hungarian banks (70% foreign banks, 23% domestic private sector, and 7% Hungarian State), their strong financial structure and higher performance comparatively to European bank averages, the fact that Hungarian banks enjoy the confidence of their owners, the need to break a specific strategy vicious circle, and the strategy options that are currently open to them.
Jens Larsen (3B) of Nordea referred to the range of quantitative methods/techniques that are currently available for the evaluation of M&A opportunities. During the discussion, these methods/techniques were grouped into the evaluation of (1) a target’s financial and operational performance on a stand-alone basis, (2) the quality of a target's balance sheet/the target's net asset value, (3) the strategic implications, synergies and risks of combining the target's and the acquirer’s operations, and (4) public market comparables of similar transactions. For each of the four groups of techniques, specific tools and their practical applicability were discussed.
Papers on such techniques and tools were invited for the 33rd Eurobanking Meeting to be held in Copenhagen in 2005.
Raimo Voutilainen (Aktia Savings Bank) and Pekka Korhonen (HSE) (4 & 5) presented an empirical study of choosing the most attractive alliance-consolidation structure model between banks and insurance companies. Using the analytic hierarchic process (AHP), an expert panel of top managers of Finnish banks and insurance companies ranked six alliance structure models along nine criteria. The panel preferred the (strong and tight) financial conglomerate model and the bancassurance/assurfinance model, which are not always feasible. A risk-averse manager prefers a cross-selling agreement with no overlapping service channels, while the risk-taker prefers the financial conglomerate model. In short, for the risk-taking executive, stronger alliance-consolidation structures are preferred.
In the discussion that followed, as to which part of the insurance business has been retained in other countries, it was noted that some financial organizations have retained the life-insurance business, others have retained the non-life (general) insurance, while others have opted to retain both life- and non-life-insurance. Also noted was that the channels used by banks and insurance firms are basically different and difficult – but not impossible - to integrate, which does not help cross-selling.
Joergen Willum (6) presented a paper on customer performance management, and specifically, the system of measuring the economic profit (EP) of the SME (small- and medium-sized entity) customers of Nordea – the nordic bank that was born from a four-sided cross-border merger. The system, which aims at shifting management’s attention from operational profit (OP) to EP (slide # 15), is being implemented in four phases: ‘Communication’, ‘Parallel reporting/EP understood’, ‘Transition to EP’, and ‘EP embedded in business decisions’ (slide # 25).
Ulf Stolzke (7) of FSB presented a paper on the introduction of an integrated document management and workflow system to a bank’s back office, and specifically, an early adopter’s designing, selection and implementation of an integrated document management system that has been installed in FSB (Fonds Service Bank), which administers its founding partners’ fund accounts. Noting that investment is being recovered through reduced process times, the major lesson learnt was ‘Take your time – early on’. One should obtain management attention and support; involve employees thoroughly; stick to one’s timetable and not be too keen to gain quick gains. Early adopters can profit on first mover advantage – but this involves risks.
Eric Salomon (8) presented CL’s system of detecting card fraud through real-time scoring. Following card fraud definitions, statistics and trends, CL statistics and existing tools for fraud detection were reviewed. Focusing on the credit card international authorization process, the CL system of card fraud detection using a scoring approach (slide # 17) – fusing a transaction score with a client score into a card fraud detection score - was described. Results, limitations and key success factors (slide # 23) were discussed. Looking ahead, the EMV standard and smart cards were reviewed.
Jouke Wieringa (9) of ING presented the two-year old and continuously developing customer profitability systems at his Bank for retail customers, mid-corporates and (soon) medium-large corporates. Users are 5000 salesforce (using the click-call-face concept), almost all management levels (for portfolio views), risk departments (for loan pricing), controllers (for analysis) and marketing/product management. Current system usage comes to about 150 actual users a day. Stressing the need for customer-oriented performance tools, the systems emphasize RAROC and EVA (the former being preferred when capital constraints are comparatively low) (slide # 9), expected loss and economic capital for loans (slides # 10-13), have a flexible drill-down/aggregate-up capability, and allocate costs (slide # 27), provide a range of analysis possibilities (slide # 28) and are continuously improved (slide # 31). The presentation included a list of query results (slide # 16) and query screen examples.
Per-Goeran Persson (10) of UBS presented an in-depth discussion of the role of risk in capital management by focusing on two key questions: (1) How much capital does a bank need? Answer: This is a mainly a market-driven adaptive process. (2) Should we allocate regulatory capital, risk capital or a combination? Answer: We should attribute regulatory capital for capital management and performance measurement purposes; and allocate risk limits in line with our overall risk appetite, which is based on an earnings volatility. The conclusions of the discussion relating to the problems involved in treating risk capital as the solution in capital management were summarized in slides 17-18.
Esa Vilhonen (11) of OKO Bank discussed the problem around IAS/IFRS 39 Hedge Accounting, and specifically the one arising from IASB’s view that the change in value (MtM) of every derivative instrument should be booked in either P&L or equity - currently the usage of an instrument defines the accounting rules of a specific deal. The in-depth discussion concludes that the proposed approach to value a specific type of instrument to fair value regardless of their original aim of use should not be used. Instead, the valuation principle of the hedged instruments should be used to value the hedging instrument. The fair value effect could then be added to the disclosures of the annual report.
Petra Muenchau (12) presented WestLB’s model for the internal evaluation of loans, using the funding oriented yield method (calculation of the NPV with zerobond factors resulting from the market interest curve used). Following a discussion of internal evaluation, the implementation of the project was described, including a comparison with other methods. Used, so far, only for EUR loans, the funding oriented yield method allows faster adjustment of internal evaluation to market changes, better interpretation of margin, and optimal adjustment of the method to WestLB requirements.
Elias Halamandaris (NBG) and John Sparkes (NIMBAS) (13) focused on the need for a wider framework of Risk Management – wider than the narrow RM process or an IT-based RM system chart – that can offer all the benefits of ‘the big picture’ and a common focus for Risk Management practitioners and their clients. Following a literature review of such frameworks and the RM survey results, a wider framework was proposed that groups all key factors involved in RM into three interrelated elements - Strategy, Systems and Soft Factors (slides # 13-16) - which were then briefly described.
Fabio Mercurio (14) of IMI presented a lognormal-mixture smile-consistent (or uncertain-volatility) option pricing model - a rather general uncertain volatility (and rate) model - with special application to the FX options market. Starting from the premise that the Black-Scholes model cannot consistently price all options in a market, and aiming at consistently pricing all (plain-vanilla) options in a given market, an alternative model for the asset price S is resorted to. The proposed lognormal-mixture uncertain-volatility (LMUV) model has the tractability required, prices analytically a number of exotic derivatives, accommodates general implied volatility surfaces, and allows for Vega bucketing.
Tom Conlon (15) of the Irish Payment Services Organization (IPSO) and the European Payments Council (EPC) reported on the progress made toward integration of European payment systems through EPC. EPC was formed by banks in 2002 to implement the Single Euro Payments Area (SEPA) as a voluntary industry strategy i.e. through self-regulation - its objectives and mission are described in slides # 6-7. EPC is being registered in Belgium as a not-for-profit organization, its structure is changing (slide # 9) and relations with SEPA stakeholders are emphasized. Recent EPC progress relates to customer and business requirements, infrastructure (Pan-European Automated Clearing House - PEACH), end-to-end straight-through processing (STP), cards and cash, while the great challenge ahead is how to achieve effective implementation.
Chiel Bakkeren (16) of ABN-AMRO’s Quantitative Consultancy unit (part of the quants group within Group Risk Management) presented a paper on channeling sales, and specifically a model that seeks to answer the question ‘Which distribution channel/s should be used for which products in the most economical way?’ A simplified problem was given before the development of the model was described. Users consider the model as a tool that makes their planning work a lot easier, while it constitutes a firm base for negotiation among the parties involved. Preliminary results point to a potential 15% cost reduction. Screen examples and Key Success Factors (slide # 18) are included.
Wilko Bolt (17) of DNB presented a paper on the retail payment market in the Netherlands. Following a brief description of the Dutch retail payment systems in terms of market structure and performance, usage of payment instruments and tariff structures, it was concluded that this market has, over the last decade, shown a dynamic potential in terms of product and process innovation. However, some parts of the Dutch payment services industry have recently attracted controversy and antitrust scrutiny, relating to pricing transparency, infrastructural arrangements, organization of the debit card network and accessibility of retail payment systems. It is proposed that such market imperfections – particularly with regard to optimal social pricing of retail payment transactions - be viewed through newly developed theories such as the two-sided market theory, which delineates who has the greater interest in a transaction being made. Such perspectives may lead to new insights and relevant policy conclusions.
Anders Wulff-Andersen (18) of UBS presented a paper on the counterparty credit risk challenge - ‘one counterparty, many products’ – or the aggregation problem of different products to counterparty level and how the credit risk of different products compare. Looking at exposure is insufficient to aggregate and compare credit risk. Following a discussion of the importance of a stand-alone single name credit risk measure from a credit risk control perspective, the desired properties of such a measure (fairness, coherency, consistency) were presented and a set of risk coherency axioms was applied in the context of a term structure of credit exposure and time of default. Having discussed the fairness of a single name credit risk measure with regard to the treatment of different products, the fair market value of credit protection was proposed and applied as a coherent stand-alone single name credit risk measure. Suggestions were made as to how the concepts presented can be implemented.
George Christodoulakis (19) of BoG presented a Dynamic Panel Data (DPD) model with fixed effects for forecasting non-performing credit and loss provisions for corporate, mortgage and consumer loans in the Greek commercial banking sector (22 financial institutions). Having reviewed the various approaches available to a forward-looking supervisory authority, a non-publicly available data set for the period of 1994-2003 was analyzed, maximum likelihood estimates for the DPD model parameters were provided, and the out-of-sample behavior of the model and data using both symmetric and asymmetric forecast error loss functions were presented extending a rationality test under non-normality using a specific class of density functions. Empirical findings suggest that both endogenous and exogenous effects contribute to the forecasting power of the model, which was found to outperform a number of alternative specifications.
Ali Higgins (20) of ECGD presented a paper on modeling transfer risk. ECGD supports British exporters by issuing guarantees against non-payment and has a highly concentrated risk portfolio with long horizons of risk. Its primary credit risk model Credit Explorer (CEx) is an extension of CreditMetrics methodology. Transfer risk (i.e. the additional risk - in a country-default situation - which is due to the possibility that foreign exchange may be unobtainable) is modeled by CEx using the ceiling rule method, which is fairly simple. Its use, however, is valid only if the ratings used in the model are local currency ratings. Such modeling of transfer risk increases ECGD’s VaR by 5-10%, an effect that is magnified if concentration grows. And yet, the methodology has some major advantages over alternative modeling techniques – as well as some clear weaknesses. Four points were suggested for focused discussion.
Michael Maerz (21) of UBS presented a paper on the capital allocation framework for a credit portfolio, comparing market and model based approaches. Three views of capital were considered: regulatory (minimum capital that a bank must have), economic (capital the bank should have), and (recently) market-based (subordination level that the market requires). Capital allocation is essential if performance measures are to be reliable. In effect, running a successful business requires full alignment of capital, risk and return (slide # 5). The treatment of regulatory capital under the two Accords was first compared and the impact of using different models for a credit portfolio was then discussed. Given that different models normally yield different results, the question was posed as to whether there can be a unique view of capital. Market-based capital was then discussed, which can be estimated either by engaging in a hypothetical securitization transaction applying the rating agency’s point of view; or by defining a collateralised debt obligation (CDO) / collateralised loan obligation (CLO) / residential mortgage backed security (RMBS) peer group. In conclusion, the best approach might be to build a model with the appropriate sophistication to come up with a reasonable loss distribution for the credit portfolio. The model should then be calibrated to internal parameters. Finally the model should be benchmarked with CDO/CLO/RMBS market subordination levels for ‘peer transactions’ thus performing a ‘reality check’. Four issues that must also be considered were discussed (slide # 19).
Pavel Finger (22) of CCB presented the joint PChC/CCB project ‘Rating of SMEs (small- and medium-sized entities)’ which allows an easy, not-so-expensive and standard assessment of SMEs. There are about 80,000 SMEs and until the project was launched, information about them was lacking. The pilot project was carried out between October 2003 and April 2004. CCB (with a staff of 8) has founded BCB together with five Czech banks. Currently, there are nine banks-users of BCB and another seven are negotiating to join it. There are many potential partners of the project – six banks, two corporates and some government institutions. CCB is owned by CRIF (Italy), Trans Union (USA) and Mozart Two (Ireland), will set up the credit bureau in the Slovak Republic and is working on the creation of a non-banking credit bureau. Today there are 63 contact points through the regional chambers of commerce throughout the Czech Republic. The rating tool (slides # 14-19) uses 18 financial ratios, 7 non-financial ratios and 7 rating levels, focusing on current creditworthiness and financial stability within the context of the future development of the respective industrial sector, as well as other factors. A ‘comment’ is also produced, which is a standardized report on the strong and weak aspects of the SME’s economic health.
Joerg Stensinski (23) of Group Risk Management Programme Lending (Scoring Policies & Development) at ABN-AMRO presented a paper on scoring in the consumer credit cycle. With almost 20 business units all over the world making use of more than 60 different types of risk scorecards and implementations, one of the unit’s objectives is to develop and maintain a global self-sufficient, uniform, high quality and cost aware credit scoring infra-structural network. Bearing in mind a variety of scoring restrictions and temptations, the potential scoring issues that need to be dealt with at ABN-AMRO and at many other banks are divers. Some of them may be: Scorecard development (the determination of the objective function, necessity and types of reject inference, selection of characteristics and attributes), general planning and setting of objectives (e.g. use of behavior scoring), testing of systems, adverse selection through the implementation of untested strategies, and the role of consultants and external vendors (e.g. credit bureaus) in the process of implementing and maintaining scoring environments. Also the reasons for the problems and issues are divers. The road map to succeeding in dealing with and resolving issues includes internal raining, strict scoring policies and guidelines (project plan, formal model assessment, implementation, validation, MIS and global reporting) and regular reviews (scoring reviews and consumer risk inspection) and bank internal consultancy. First results were very positive.
Janez Barle and Anton Zunic (24) of the Slovenian Ministry of Finance presented a paper on deriving principles of loan portfolio diversification from Markowitz’ theory of portfolio diversification. Noting some key differences between securities’ and loan portfolios (Table I), a simplified probabilistic model of loan portfolio losses is used, which is similar to the one used in Basel II documents. Portfolio loss is explored from the viewpoints of homogeneity in exposure size, correlation and volatility of individual loans. The majority of derived rules are basically a confirmation of rules already used in practice, but some of them contain quite innovative concepts. Using Sharpe’s CAPM model as a starting point, some rules are also derived aiming at minimization of loan portfolio systemic risk. References are provided.
Wolfgang Schmidt (25) of HfB presented a paper on explicit hedging and pricing of basket credit derivatives. The paper investigated the pricing of basket credit derivatives and their hedging with single name credit default swaps (CDS). The market in credit default swaps quotes fair insurance premiums whose dynamics is the natural starting point of our model. Pricing basket credit derivatives requires a model for the dependencies between the default times. In case of a pure jump filtration, dependencies are characterised by default implied spread changes. In this setup, a simple system of integral equations was provided involving the notional amounts of the dynamic hedge positions, the price and the spread of a basket derivative. The presentation included some numerical examples of explicit hedging strategies and valuations of first-to-default baskets illustrating the approach.
Mordecai Avriel, Oren Parnas and Alon Raviv (26) of Bank Hapoalim presented a model for forex liquidity reserve. While the Bank is required to provide liquidity at all times - including difficult ones – cash reserve (liquidity) management is a dynamic process, which is influenced by mainly random factors that cannot be controlled by the Bank. A cash reserves management policy was developed which is based on a model drawn from inventory theory and has two connected ‘reservoirs’ – cash overnight reserves, and short deposits. The model, which was developed in the presentation, includes an optimization algorithm that has cost minimization as the objective function, and is constrained by an upper limit on the probability of liquidity default. The model was implemented in a user-friendly computer code written in Delphi that provides the optimal policy on a daily basis.
Jes Gerstrom (27) of Danske Bank presented a paper on the management of long-term liquidity risk. The Bank uses the Net Funding Need (NFN) analysis which is a series of gap reports designed to show the overall long-term funding profile of the Bank - its net liquidity position in a single time bucket is defined as NFN. The basis for NFN analysis is the BASE CASE model, while a BASE CASEMOD NFN is calculated where the assumptions are modified so that the resulting NFN curve can be used in the Bank’s funds transfer pricing (FTP) process, and forms the basis for long-term funding planning. The preparation of a base-case gap analysis was then analyzed. The NFN analysis is distributed in the form of an interactive application (the NFN visualizer – a flexible drill-down tool) which is built on top of the separate ALM data warehouse that contains the data for the NFN and other ALM statements. Examples of how NFN analysis is used were given and specifically in monitoring the Bank’s long-term liquidity mismatch and funding gearing; ensuring that the Bank is not building up an oversize funding imbalance in more distant periods; facilitating the Bank’s planning of medium-term and long-term funding activities; and serving as the foundation of the FTP process within the Group. Areas for improvement refer to including assumptions for new business, and scenario development.
Markku Pehkonen (28) of Sampo presented a paper on modeling financial liabilities in Finland, and specifically: the key features of demand deposits and with-profit products and their risks on earnings and value; techniques, results and challenges of modeling liabilities; and how modeling results can be used in product design, pricing and ALM. Stressing that modeling is just a starting point and that we need to go from measurement to management, the conclusions were grouped as follows: (a) Modeling is just a starting point (modeling helps one understand the earnings logic and the related risks of the products where customer behavior is an important factor; modeling results are never a comprehensive picture – they seldom give ready-made solutions to complex problems; models are tied to history – if there are changes in customer behavior, modeling cannot see them); (b) Use modeling results in ALM (define and decide the financial instrument equivalency of liability portfolio; estimate the profitability in the market risk neutral position; modify the liability features and decide on the risks one is willing to take); (c) Use modeling results in active product design and pricing (if some features of the products are problematic, change them if possible; in many cases, the pricing is not a key factor for a customer).
Ulrich Krueger (29) of DBB presented an analysis of recent developments in the Basel II framework, focusing on risk-weight functions for a calibration based on unexpected losses. Following a detailed discussion of the implications of these developments – with fully worked out examples - it was concluded that: (a) In Basel II, risk weights are model-based for the first time, implying increased risk sensitivity and a bottom-up approach. (b) QIS 3, floors and parallel calculation will ensure that the level of capital in the system is maintained, implying a top-down approach. (c) The main goal is to establish a link between micro- and macro-prudential perspectives of banking supervision. (d) Open problems include the treatment of diversification and the validation of risk parameters – probability of default (PD), loss given default (LGD) and exposure at default (EAD).
Marco van der Burgt (30) of ING presented a paper on the development of a rating model for domestic macro risk – the risk that debtors are unable to repay their debt due to domestic macro events such as currency crises, banking crises, political crises and business cycle crises. Events and financials were gathered on 56 countries for the period 1970-2002. The sample included emerging markets, developing and developed countries world-wide. Approximately 220 explanatory variables were created and grouped into eight categories: economic structure, economic size and development, economic performance, financial/banking sector, government, current account, liquidity, external debt service. On the basis of statistical performance, non-colinearity, integrity and availability, one variable was selected per category. Some data are transformed before usage in the logit model. The logit model gives 1-year event probabilities, which are translated into a 1-21 rating scale by using an expert opinion model. This expert opinion is based on the ability to generate income/cash flows during a long period of time, vulnerability to external shocks, willingness to fulfil debt-service obligations, exchange rate system, independent central bank, fragility of banking system, and amount of short or floating rate debt. External sovereign ratings of Moody’s, International Investor and S&P are also included in the expert model. Further developments include further development of the country rating model, calculation of transfer risk and calculation of probability curves. References are provided.
Izok Valencic (31) of Nova KBM presented a paper on the impact of Basel II and standards and the response of a bank’s IT organization. In an increasingly competitive and challenging environment, a wide range of pressures are exerted on banks and their IT organization while an even wider range of response items is available to IT. Noncompliance is costly and IT Departments are being hit from all angles, as almost every aspect of compliance impacts on the IT function. IT governance has therefore become an essential element of successful compliance, ensuring that IT-related risk is minimized. Bearing in mind the IT Risk Management Connectivity Combination Object approach and that the 3600 Risk Manager must give equal emphasis on standardization/integration and on creative collaboration, a Risk Management-Security-Knowledge Management approach is proposed which focuses on knowledge, surrounded by people, process, technology and information. A framework of IT Risk Management – together with maps for facing risks - is proposed (slides # 14-16). Compliance is costly because it demands holistic change within a bank’s organizational processes and technology infrastructure. Thus, a holistic approach and related practices are required. This implies a full understanding of the nature of threats/risks and their vulnerability/final impact (Risk = Vulnerability x Threat, while Expected Loss = Risk x the required Resource Value x the Unavailable Response); a framework of team structures to confront risks; and control guidelines such as COSO/COBIT (Control Objectives for Information and related Technology) (slides # 21-23). Bearing in mind the Basel II approach to operational risk and related standards (e.g. BS ISO 17799), three charts are proposed (slides # 28-30). Finally, a concluding checklist is provided (slides # 31-32).
| Paper Summaries | Conference Report | Conference Picture Gallery |
