Using SAS analytical power to predict bankruptcies

Teikoku Databank realizes up to 80 percent accuracy rate

Teikoku Databank (TDB), with the help of SAS, has introduced the first corporate bankruptcy prediction model based on non-financial qualitative data in Japan. The new index for corporate evaluation is now used to offer various services to respond to the needs of clients engaging in new business transactions and partnerships.

A new benchmark in Japan
In today's fast-moving economic environment, businesses are exposed to various types of risk, including bankruptcy. When a company engages in new business transactions with a third-party company, using company-value indicators may help evaluate bankruptcy risk and manage credit exposure. These consolidated indicators, however, are generally based on financial data and to a certain degree are exposed to the subjectivity of the examiner. Many times, they can only offer a generic or limited evaluation of a company's safety. For this reason, TDB has deployed a new prediction model based on extensive qualitative corporate data and evaluation criteria, creating a new company evaluation benchmark in Japan.

This prediction model is the first model in Japan to run on qualitative data, such as management, company history, information on solvency, funding-reserve strength and mortgage-setting conditions. TDB's prediction model uses data from its approximately 430,000 client companies and ultimately selects 10 from the approximate 150 available variables to establish the bankruptcy risk. In addition, the company developed another prediction model which combines both qualitative data and quantitative data to predict bankruptcies. SAS efficiently manages and analyzes this huge amount of data, and with these two new models, it is possible to determine with an accuracy rate of between 70-80 percent whether a company will go bankrupt within a year.

Flexibility, speed and cost savings
The bankruptcy prediction model started back in 2000 as a corporate credit-risk research project. In March 2001, TDB already was providing some research data to financial institutions to back up their quantitative information based on credit ratings. As a result, examiners were able to fine-tune company evaluations and even differentiate the risk between companies that ranked equally. The first bankruptcy prediction service was released in July 2001. Today the service has evolved into four types: full service offers a prediction value for all the companies maintained by TDB; select service provides values for up to 500 companies; option service provides values for one company, and c-monitoring service offers prediction value and 10-grade ranking value via the COSMOSNET Internet service.

"Thanks to SAS, we can think and develop new services on the run," says Kohei Kimura, Assistant Manager of the Project Planning Section, Corporate Planning Department. "Because the SAS system is very flexible, the speed of service development makes it cheaper to use than any of our previous systems. The original model developed by SAS is still in use and has maintained its accuracy level throughout our monthly validation process. Additional models were developed as a backup, but a new release is still not needed at this point in time," he continues.

Aggressive business tool
SAS, which has become an essential business tool in TDB's Corporate Planning Department, has penetrated other departments in the company as SAS' excellent performance in data processing became well recognized. SAS came to be used in a variety of new areas.

For example, in the Service Management Department, SAS is used to control and maintain the quality of several product databases. The Business Promotion Division uses SAS in the management of daily business performance and on-the-job support. In the System Department, discussions have begun about using SAS to effectively and systematically manage and share data and programs, which are proliferating within the company.

In addition, SAS is recognized not only for its ability to manage system performance, but its detailed and thorough consulting and support are also highly valued.

"SAS has helped reconstruct the accounts database we had been using since around the late 1960s to deal with in-depth changes in the accounting system and bank reorganizations," says Masanobu Ojima, Manager of the Project Planning Section, Corporate Planning Department. "Several years ago we attempted to regenerate it, but the impacts in terms of costs and man-hours were enormous, so that it became impossible for us to achieve. Practically ready to give up, we engaged SAS, and with the help of its consultants we finally succeeded in reconstructing the database. This has been a very significant achievement in our company. The revolution brought by SAS to our company is not only the result of its analytical software, but is also the fruit of such quality services."

Thinking forward with SAS
A long-term issue for TDB is the building up of its human resources infrastructure. TDB needs to nurture second-generation personnel who can freely use SAS in order to expand the company's potential. The road to efficient use of corporate data depends on whether future strategic decision-makers understand SAS and make the maximum use of its potential. Rigorous use of the data to create even higher value-added services will lead to the development of TDB's future business. For example, one concept is to develop a forward-looking index service to measure the value and growth of companies.

"As long as there are credit-related inquiries, credit risk will be an issue for us," Ojima says. "But in the future we also plan to create many other types of predictive models. Something new like this cannot be done without trial and error, but we count on the contribution of SAS, both as analytical software and as human support, to make it possible. We continue to expect proposals from SAS that don't stop with the presentation of analytical capabilities, but also include the concrete know-how, consulting and best-practice methods for managing and using data."

The results illustrated in this article are specific to the particular situations, business models, data input, and computing environments described herein. Each SAS customer’s experience is unique based on business and technical variables and all statements must be considered non-typical. Actual savings, results, and performance characteristics will vary depending on individual customer configurations and conditions. SAS does not guarantee or represent that every customer will achieve similar results. The only warranties for SAS products and services are those that are set forth in the express warranty statements in the written agreement for such products and services. Nothing herein should be construed as constituting an additional warranty. Customers have shared their successes with SAS as part of an agreed-upon contractual exchange or project success summarization following a successful implementation of SAS software. Brand and product names are trademarks of their respective companies.

Copyright © SAS Institute Inc. All Rights Reserved.