Design And Implementation Of Bank’s Assets And Liability Management System Based On Data Warehouse

In recent years，the supervisory authority at the higher level, for example, theChina Banking Regulatory Commission and people’s bank of China, put forwardthe new regulatory standards on the system risk and individual risk which mayhave to face when the bank capital cover bank sufficiently. At the same time, as thefinancial market at home and abroad is getting competitive, higher requirements tothe bank business development and expanding construction are requested.LONGJIANG BANK is a city commercial bank soon after establishment, and itshould build up effective risk management mechanism to avoid interest rate risk, tomeasure the demand for capital in future, to identify customers who generateconsiderable profits, and to analysis the difference of different clients.This paper based on the realistic project-the strategic asset-liability items ofLONGJIANG BANK, deal with the design and realization of transfer pricing in thefirst phase of this project. With using synthesis business intelligence technology,such as data warehouse technology, on-line analysis and data mining, the systemachieve query function of each dimension report of transfer pricing.By modeling analysis of historical data, obtained personal housing mortgageloans LONGJIANG BANK prepayment modeling. Model as a result of subsequentdata mining operations into the system operational parameters. Using On-LineAnalytical Processing technology, this article divided dimensions, defineddimension of agencies, departments, subject, currency, customer, customermanager, industry and so on in details. Data warehouse and balance system toanalyze various data interface description, given the overall system designframework. Detailed system design and implementation, ETL processes for E(Extract), T (Transformation), L (Load) all aspects of the actual situation, using atargeted approach: First, a unified data format standardized definitions usingDataStage JOB data sources provided by the data warehouse extract; systembackend use stored procedures, each dimension table or fact table corresponds to astored procedure, to achieve data processing work in a stored procedure, front-endstatements Cognos show.Through the test and the actual movement confirmation, the business systemdata can be imported successfully, and business data, manipulation data, result datacould be stored separately. The ETL scheduling tasks can be initiated automaticallyand periodically in accordance with the setting time and task in the system, and thedata can also be batch processing. At the same time, the system supports that the query task can call the run results, which meet the demand of the balance system.The paper is a good basis for the product pricing, profitability analysis, costsharing and other mining analysis work in the project follow-up.