逾期信用卡持卡人背景因素之探討-以國內某銀行為例

Abstract

隨著民國八十一年開放十五家民營銀行核准設立，宣告台灣金融市場進入春秋戰國時代，銀行開始調整業務重心，從傳統的企業金融業務轉為以消費金融商品為主力的業務，舉凡信用卡、現金卡、小額信貸等，信用卡是一種小額免擔保的信用貸款，具有高利差、高報酬與高效率的優點，是銀行消費金融商品獲利主要的來源之一。

然在無擔保的背景因子下，呆帳風險的管控對發卡行來說是一門重 要的課題，信用卡交易對發卡行僅有微薄的交易手續費收入，故會使用循環信用且具有償債能能與償債意願的人會是銀行珍視的好客人。有短期現金需求而會使用循環信用卻不會倒帳的客人有什麼共通性﹖銀行如何透過有效的信用評分模式，減少呆帳損失﹖

本論文主要研究範圍在實證持卡人之職業等級在婚姻、性別與預借現金行為的交互作用下對呆帳比的影響。以有效健全發卡行在信用卡評分系統之參數化模式。本研究樣本資料取自國內某金融機構95~99年信用卡倒帳之持卡人共計13,394筆樣本資料，利用多元迴歸方式，探討出職業等級高低顯著影響著呆帳比率(呆帳金額/信用額度)，然在婚姻、性別與預借現金行為的交互作用下，職業等級影響呆帳比更加顯著。

綜上所述，信用卡持卡人因為職業等級的不同，會產生不同的呆帳比率，如果發卡銀行在設計評分系統參數化時，如何針對不同職業等級做好風險管控的能力，增加評分系統的鑑別能力，方可有效減少銀行呆帳損失，強健銀行的經營。

Taiwanese banks entered into a new era of fierce competition as government approved 15 new private banks in 1992. They started to change their traditional corporate business strategies and focused on consumer financial products including credit cards, cash cards and personal credit loans. Credit cards business is a credit loan of small amount with no pledge of collateral. With such attributes as high spread, high return and cost efficiency for banking business, Credit cards loans become one of main sources of profits for banks. As credit loans without collateral are prone to default, Banks becomes more concerned with how to manage and control default risk. Because credit card transactions generate fee incomes for banks, revolving credit card borrowers with good credit history become banks’ good costumers. What are the common properties among short-term revolving credit borrowers with no risk of default? How would banks reduce default loss by using effective credit scoring model? This study mainly analyzes the effect of job title on default risk by controlling the interactions among marriage, gender and cash advance behaviors. Banks could use the findings to improve effectively parameterization of credit scoring models for better credit risk management. The credit card data contains 15,000 sample points of credit card default from 2006 to 2010 in a Taiwanese bank. The study, using multiple regression method, finds that the lower the job title, the higher the default risk after controlling such factors as marriage, gender and cash advance. The results suggest that card issuing banks could fine tune job title parameters in credit scoring models, which could enhance identification capacity and effectively reduce credit card default loss.