Business Intelligence (BI) supports decision-making capabilities of banks and helps them to attain and maintain a competitive advantage in today’s turbulent markets. Its usage makes the conditions, procedures and mechanisms for creating business knowledge. BI supports reaction to external pressures and enables effective risk management and implementation of regulatory compliances such as Basel II accord. The purpose of this paper is to outline BI techniques and their role in analysis of key business factors in banking industry
Keywords: Business Intelligence (BI), Data Warehouse (DW), On-Line Analytical Processing (OLAP), Data Mining (DM), Key Business Factors (KBF), Banking, Financial Industry
Financial institutions’ long history of computing has created a collection of data that is measured in pentabytes. Large banks produce hundreds of millions of transactions daily which are stored inside complex IT systems. Efficient analysis of these data is crucial for a success in the financial market. Detection and suppression of fraud, risk management, customer management, banking product management, and loss prevention are some of the primary concerns of institutions providing financial services. BI technology can collect and transform millions of records for comprehensive analysis. Financial institutions exploit BI technology to analyse and understand the behaviour of their clients, to better satisfy their clients’ needs in an endless chase for a competitive advantage in the market.
The paper gives a review of BI techniques and their role in banks’ customer and risk management applications as well as implementation of regulatory compliances. This paper is structured as follows. Section 2 outlines the business intelligence and its main features. Section 3 describes the environment of banking operations. The role of BI technology in analysing key business factors in banking with a brief review of typical applications is presented in section 4. At the end, the main conclusions are drawn.
2. Business Intelligence
Business intelligence is the ability of an organisation to understand and use information to its gainful operation (Osterfelt, 2000). Enterprise BI is a way that brings synergies to business processes and new efficiencies across business areas (Liautaud, Hammond, 2000). BI offers to enterprises „one version of truth“, providing consistent and harmonised data to every department in an organisation (Bochner, Vaughan 2004). How can one achieve data consistency (also known as “one version of truth”) across different applications in a complex organisation? There are three important goals that need to be accomplished in order to achieve data consistency (Arents, 2003):
Ø Timeliness: the data within system should be synchronised with all other applications;
Ø Accuracy: the data should encompasses every data from any other application;
Ø Acceptance: the users, convinced of timeliness and accuracy of data, should be able to actively use the system as support for decision making.
In today's companies, BI plays an important role in support of the decision making process to augment competitiveness, making an efficient link between business strategies and IT. Business intelligence technology has been continuously expanded and improved and more and more complex business questions can be answered using these technologies. The most widely used business intelligence enabling technologies, described in more detail below, are: data warehousing (DW), on-line analytical processing (OLAP), and data mining (DM).
3. Data Warehousing
A data warehouse (DW) is an integrated collection of historic detailed and summarised data that is supplied by the spider web environment from internal and external data sources. It is organized by business areas (subject oriented) and is user-friendly, especially for a manager and business analyst. The original label that pre-dates the data warehouse is still the best description of what we are designing: a decision support system (Kimball, Ross, 2002).
The most delicate part of the data warehousing is the extraction of data from various data sources with variable data quality. It has to be decided which internal and external data will be fed into the warehouse, and how the inconsistencies among data sources will be resolved. Large amounts of operational data is accessed by end users and stored in different systems and the same data are represented differently in different systems (Turban et al, 1999).
Benefits of the data warehouses are most obvious in companies with several computer platforms and versions and with many different data sources. The most important benefits organisations seek from their DW efforts are: better business intelligence (39%), reduced time to locate, access and analyse information (21%), consolidation of disparate information sources (20%), strategic advantage over competitors (11%), faster time-to-market (5%), and replacement of older decision support systems (3%) (Hall, 1999).
The best known knowledge discovery techniques are online analytical processing (OLAP) and data mining (DM) techniques (Turban et al., 1999). OLAP provide users with the means to explore and analyse large amounts of data, involving complex computations, their relationships, and visually present results in different perspectives. OLAP tools are a combination of analytical processing procedures and graphical user interface. The key features of an OLAP application are: multidimensional views of data, calculation intensive capabilities and time intelligence (Forsman, 1997).
A multidimensional view of
data that is usually used in OLAP applications provides quick and flexible access
to data and information. Typical applications performed on multidimensional
data views are: roll-up (data is summarized with increasing
generalization), drill-down (increasing
levels of detail are revealed),
slice and dice (performing projection operations on the dimensions), and pivoting (cross
tabulation is performed) (Jarke et al, 2000). Complex analyses are possible, such as
time series (sequence of events) and model charting, forecasting, modelling,
statistical, and “what-if” analysis.
Presentation of information via user interface in the form of text, picture and graphics determines the way to execute queries and display of query results. It is important that the interface enables pleasant work within a graphical environment, with simple and fast running of queries, and a visually appropriate display of query results.
The OLAP technology potentially provides several benefits to an organisation: increases the productivity of business managers, analysts, and whole organisation by the inherit flexibility and timely access to strategic information; leverages IT developers to deliver solutions to business users faster; and it provides the ability to model real business problems and to respond more quickly to market demands (Forsman, 1997).
3.2. Data Mining
In contrast to OLAP being retrospective in nature (Turban et al, 1999), data mining provides prospective knowledge discovery. Data mining is a process of discovering meaningful new correlations, patterns, and trends by sifting through large amounts of data stored in repositories, using recognition technologies as well as statistical and mathematical techniques.
Data mining technology discovers hidden trends and patterns in large volumes of data. A significant distinction between data mining and other analytical tools is in the approach they use in exploring the relationships among the data. The analytical tools usually support a verification approach, in which the user hypotheses about data interrelationships are verified or refuted. This approach relies on the intuition of the analyst to pose the question and his or her ability to refine the analysis based on the results of potentially complex queries against a database.
Data mining uses discovery-based approaches in which pattern matching, clustering, neural networks, genetic, and other algorithms are used to determine the significant relationships and correlation among data (Kennedy et al, 1998). Data mining algorithms can look at numerous multidimensional data relationships concurrently, highlighting those that are dominant or exceptional. Data mining enables users to discover knowledge and provides them with greater depth and understanding of data than ad hoc querying and using of OLAP applications.
4. Banking Environment
Banks operate in one of the most dynamic environments: new markets are being opened, new products are being launched, new competitors enter markets that were previously reserved only for banks, new regulatory requirements are being imposed, and new customer needs are being identified. Rapid external changes and high pressures affect banking operations with immediate impact on development of banking IT systems. Continuous innovation and launch of new products with ever shortened life cycle has led to development of many non- or loosely connected applications making banks IT a heterogeneous collection of systems and data. With mergers and acquisitions occurring in the banking sector all over the globe, heterogeneousness of systems and data is augmented. The need for unified resource of information for decision-making led to an integrated collection of data known as the data warehouse. More and more companies worldwide use and/or develop the data warehouses. Usage of techniques and tools for extracting useful knowledge from the available data is necessary for a bank that has to respond to the business pressures.
Banks operate in a complex environment, as depicted in Figure 1. Rigorous competition and market requirements dictate bank’s operations which are further restricted and regulated by several national and international authorities who demand constant and prompt reporting and auditing to assure supervisors and stock holders of stability. Currently, all international banks and national banks of most countries are implementing, or make plans and preparations to implement the ‘Basel II accord’ (BIS 2005). The implementation of Basel II accord relies heavily on the bank’s IT infrastructure, particularly on BI and DW systems. To manage newly introduced component of operational risk, requires new data and databases. Detailed logs of banking business processes become a new source of data that will be subject to BI analysis in order to optimize bank’s processes and minimize operational risk.
Considering the way banks operate and use their data, the banks’ needs for information can be divided into two basic categories: Customer Management and Risk Management data and applications.
4.2. Customer Management
The customer is a focus of
all business activities. This is not specific to banking, almost any company is
struggling to understand who the customer is, what the customer wants, when,
how, and why the customer wants it. It has become essential for companies to
find new ways to attract new customers, to maximize the value of each existing
customer, and to retain the most profitable ones (Liautaud,
Ø Control in every aspect of relation with clients;
Ø Means to recognise and retain the most profitable customers;
Ø New ways to attract new customers (from competition);
Ø Efficacy of its processes and profitability of products;
Ø Understanding of new markets and need for new products.
Moreover, the financial industry is, and will be more, oriented towards the selling of new products than toward traditional services such as offering loans and holding deposits. That makes a modern bank’s employee more a salesman than a traditional banker. Armed with timely and accurate information, a modern banker knows all about his or her customer, and all the bank’s services that would be appealing for that particular customer, as well as profitable and risk-acceptable for the bank.
Having a strategy to leverage modern information technologies to gain an operational efficiency, enhance customer service, raise productivity and profitability, the banking industry is becoming less focused on its core business of holding deposits and giving loans, and more on managing information (Girish, 2001).
4.3. Risk Management
Due to the nature of its business, risk management is inherent to financial industry. In banking there is an ever present risk of payment default, fraud, theft, identity theft, and operational risk connected with internal procedures and processes.
The implementation of the new Basel II regulatory compliance (BIS 2005) means interconnection of IT systems with processes to gain higher transparency and reliability of bank’s operations. Evaluation and prediction of market changes and minimisation of capital reserves are also in focus of Basel II. Compliance with Basel II accord generated a need for central repository of purified, sorted, and aggregated data about financial transactions and associated risks. A data Warehouse is a central and crucial component of a software solution for Basel II enabling a flexible and modular infrastructure for risk management. Relevant surveys (Furlonger, McKibben, 2005) (Bender, Ding, 2005) made a comparison of modern software risk solutions and identified the following common Basel II components:
Ø Risk data warehouse;
Ø Risk analytics/risk engine;
Ø Loss severity and probability estimations;
Ø Data management and integration;
Ø Optimization and management of collaterals;
Ø Asset liability management;
Ø Applications for reporting and regulative compliance;
Ø Credit portfolio management.
The same surveys revealed the following significant components found in some Basel II solutions:
Ø Global limits management – used for consolidation of various exposures to risk for real time control;
Ø General ledger reporting and IAS compliance;
Ø Risk databases - consolidated third party data about operational risks;
Ø Profitability management – financial modeling and analysis of possibilities of capital allocation using business rules, expected losses and profitability of clients for pricing of products.
5. Bi Technology And Key Business Factors In Banking
5.1. General Description
Key business factors (KBFs) are measures or indicators that are significantly related to the business success of a particular company or industry. Their uses contribute to the overall improvement of results and as such have to be closely and continuously monitored. Defining KBFs requires specific knowledge and understanding of a bank’s core business and services, behaviour and habits of its clients and the ways its services are being used. Implementation of KBF monitoring requires BI tools.
BI can provide business value by helping enterprises identify risk early and identifying material changes in business condition that require attention (Friedman, Hostmann, 2004). BI technology can collect and show data about each client, account, and service, and provide an aggregate panoramic view of business performance. BI enables efficient analysis of key business factors such as:
Ø Assets and liabilities analysis
Ø Risk analysis
Ø Revenue analysis
Ø Client profiling
Ø Account analysis
Ø Campaign analysis
Ø Sales analysis
Ø Customer loyalty analysis
Ø Customer care analysis
Ø Credit scoring
In addition to financial KBF such as profitability and efficacy which can be derived from financial data in general ledger, other factors can roughly be categorised as either customer or risk related.
5.2. Customer Related Key Business Factors
Client information can be scattered across multiple accounts and bank’s divisions and that challenges client analysis. BI technology through data warehousing techniques accumulates these scattered data and with appropriate tools (OLAP and DM) identifies clients with multiple accounts and their assets in order to develop an appropriate approach to the individual client.
One of a bank’s profit drains are delinquent accounts. The term applies to credit product lines whose clients consume credit and fail to make payments. The occurrence of delinquent accounts has to be promptly detected and anticipated. BI techniques enable grouping of accounts by geographical, demographical and psychological variables. The accounts that don’t fit usual standards can be identified and monitored.
The 80/20 rule is being applied particularly in the banking industry where the top 20% of the customers generate 80% of the revenue. BI techniques enable identification of the best revenue generating clients across countries, regions, cities and even affiliates and help to develop business strategies and approaches for different client segments. Depending on the client’s size, revenue, and needs, new products and services appealing to the customer and revenue generating for the bank, are being promoted.
Understanding of how its products and services are being used is crucial for a bank’s efficiency in customer relations. Applications for profiling, segmenting, and credit scoring of clients are founded on BI technologies. Customer segmentation is an important area of CRM where BI tools can be particularly useful for producing finer grained segmentation that result in better focused marketing campaigns. Customer value management is another area where BI tools can be used to predict which customer segments are likely to become more profitable in the future.
Marketing campaigns are both revenue and cost generating. To make a profitable campaign it is important to predict the campaign’s scope and size in terms of cost, media type being used, duration, etc. BI tools can easily measure the client’s response to a campaign, perform cost and benefit analysis and measure the overall results. These results and findings become a part of a bank’s knowledge base repository for future use and campaign enhancement.
BI applications for sales help banks analyse new accounts from different perspectives such as product category, client profile, and geography. The sales perspective gives a better view on marketing campaign results, helps to understand the results and market trends, and leads to better results. In selling to existing customers a probability of a positive customer response is calculated. Thus, an outcome of a marketing campaign which targets carefully selected customers can be very appealing and have very high response rate with low costs. By offering the right products to the right clients, overall customer relationship and customer loyalty are being improved. Profitability is also increased since the costs of selling to new potential customers are several times higher than selling to the existing ones.
The capability to attract and attain a customer is key to long term profitability. BI application for loyalty analysis monitors the duration of the customer relationship, the span of services used, and measures demographical, geographical and psychological impacts. Customer care for any service providing organisation is of great importance, especially in the banking industry, where users interact with banks through different channels and in different ways to accomplish their daily needs. Levels of customer satisfaction can be measured by analyzing the contact history with the client. Data warehouse and data and knowledge mining tools give an integrated view of clients through their contact history, interactions with the bank and their contribution to the bank’s revenue (Berry, Linoff, 1997).
Availability of enhanced data helps banks to shape new opportunities for revenue, strengthen relationships with existing customers, attract new customers, adapt to growth and development, leveraging existent technologies and skills. BI is a powerful tool at banks disposal to understand and satisfy their customers and achieve competitive advantage.
5.3. Risk Related Key Business Factors
Risk management is essential for banking and to the financial industry in general. Traditionally, a bank’s risk managers were highly skilled and experienced employees who, besides credit scoring and risk assessment, had an important task of training of younger personnel. Today’s bank workforce consists of predominantly young, less experienced personnel, whilst staff with a high level of expertise is either unavailable or too expensive. Therefore, information, knowledge, and information technology become the main resource in support of banking operations.
For example, credit scoring applications which are founded on sophisticated algorithms and parameters, many of which are gained through data mining, have largely replaced the need for human analysis. Another example of successful implementation of BI techniques is an early detection of credit card theft. Based on the fact that the volume of transactions following card theft increases rapidly, an on-line comparison of previous and expected credit card holder behaviour with the actual transactions, can trigger an early warning and suspicion of card theft. Such systems can save the bank considerable sums otherwise spent to compensate for losses.
The Basel II accord requires the gathering of risk data at all organisational levels. The data has to be relevant to every aspect of bank’s business ranging from customer to operational data. The ultimate objective of collecting and warehousing risk data is to increase a bank’s competitiveness in the market. These data can be categorised as data about losses and data on the effects of risk management.
The Basel II accord provides for the collection of data related to credit, market and operational losses, and other risks such as liquidity and interest rate change risks. Besides the data about losses, banks collect KRI's (Key risk indicators, and other data from statistical analysis, stress testing and simulations on loss data.
The time dimension is the most important element in risk analysis. The need for historical data is not only required by Basel II specifications, but it is also required for time series analysis of economical cycles.
The “risk engine” and the risk data warehouse are the central components of a Basel II solution. The risk data warehouse is filled in an organized and structured manner with purified data about risks in transactions and operations carried out. The warehouse is the source for credit, market and operative risk applications, as well as for financial reporting.
To manage risk successfully, the modern banker must be armed with “intelligent” information: timely, relevant, and precise information about the current customer and the current situation the banker is dealing with.
For banks to prosper in today’s complex business environments, specific information and knowledge about all operational details is required. The bank’s operations and processes have to be recorded and appropriately stored. This historical data has to be accessible for analysis and knowledge extraction.
The solution is to create data warehouses and extract knowledge from the data using BI technology. BI can leverage tactical and strategic decision making based on the vast amount of data that is gathered inside bank’s systems. The KBFs in banking are categorised as being customer and risk related. This customer related data and BI is exploited in all possible ways to augment a bank’s sales. The risk data warehouse and BI techniques are the foundation for risk management and regulatory compliances such as Basel II accord. Finally, we presented a review of typical BI techniques and their applications in the banking industry that points to a conclusion that the exploitation of BI in banking is on an upward trend and that more and more data warehousing and business intelligence applications are expected to be implemented in the future at all levels of banking operations.
Arents-Gregory M., 2003, One Version of the Truth, DM Direct, DMReview and SourceMedia, articleID=6359.
Bender, L., Ding C., 2006, Risk Management and Basel II: Comparing the Financial and Credit Risk Solution Vendors, Report Published by Celent LLC, www.celent.com.
Bochner P., J. Vaughan, 2004, BI today: One version of the truth, Application Development Trends, 2004, VOL 11; NUMB 9, pages 18-24, ISSN 1073-9564.
Forsman, S., 1997, OLAP Council White Paper, OLAP
Friedman T., Hostmann B., 2004, The cornerstone of Business Intelligence Excellence, Gartner Research, Decision Framework, Research Note, DF-21-9470, www.bi.umich.edu/startup/download/bi_cornerstones_gartner.pdf
Furlonger, D., McKibben,
D., 2005, Magic Quadrant for
Girish, G.V. 2001, Banking Business Unit
– Challenges and Achievements, Infosys Technologies Limited,
Annual Investor Meet,
Hall, C., 1999, Data
Warehousing for Business Intelligence,
Jarke, M., Lenzerini, M., Vassiliou, Y., Vassiliadis, P., 2000, Fundamentals of Data Warehouses, Springer Verlag.
Kennedy, R.L., Lee, Y., Van Roy B., Reed, C. D., Lippmann, R. P., 1998, Solving Data Mining Problems through Pattern Recognition, Prentice Hall, Upper Saddle River, NJ.
Kimball, R., Ross, M., 2002, The Data Warehouse Toolkit – The Complete Guide to Dimensional Modeling, Wiley.
Osterfelt, S., 2000, Business Intelligence: The Intelligent Customer, DM Review www.dmreview.com/editorial/dmreview/print_action.cfm?EdID=3107
Turban, E., McLean, E., Wetherbe, J., 1999, Information Technology for Management: Making Connections for Strategic Advantage, John Wiley & Sons.
Contact the Authors:
Dr. Goran Radonic, Croatian Institute of Technology, Planinska 1, 10000 Zagreb, Croatia; Phone +385-1-5494734;Email:.email@example.com
Curko, Faculty of Economics & Business University