A Study on Efficiency of Domestic Banks’ Lending to Small and Medium Enterprises (SMEs) in Taiwan- Applied by The Data Envelopment Analysis Methodology

[[abstract]]2001年第一次金改、2004年二次金改及2005年為將實施新巴塞爾協議做風險控管準備，本論文旨在對此三波變革對銀行中小企業授信業務的影響，再者對本國銀行中小企業授信經營效率與此三波變革之相關議題作探討。 本論文採取中介法的觀點，以利息支出、非利息支出為投入變數；利息收入、非利息收入、中小企業放款金額轉換值為產出變數，使用資料包絡分析法、生產力指數與視窗分析法評估36家本國銀行2001至2005年間對中小企業授信經營效率、生產力變化及效率的穩定性。 橫斷面分析之實證結果顯示，有32％屬效率銀行，非效率銀行中有63％是導因於無純技術效率，在無規模效率的銀行中，有35％是大型銀行，另有55％則為小型銀行。大型銀行家數或因擴充或合併而成長，但全體銀行具效率的銀行卻減少。 縱剖面之實證結果則顯示，大型銀行與中型銀行5年技術效率均高於全體銀行，小型銀行均低於全體銀行。全體銀行平均生產技術變動值小於整體技術效率變動值。中型銀行的視窗分析平均效率值最優，而全體銀行對中小企業授信波動幅度顯得不穩定。 此外以迴歸分析來探討其他因素對銀行效率之影響。資產報酬率、存放比、經營規模均呈正相關，送保規模呈負相關。移送信用保證逾期比率，呈正相關且顯著影響，同為衡量資產品質的逾放比為負相關，惟影響並不顯著，此矛盾現象或是銀行對中小企業授信效率，與銀行使用信用保證效率不一致之佐證。[[abstract]]The first financial institution restructuring started in 2001, the second financial institution restructuring happened in 2004, and lending practice redesigning starting from 2005 for fulfillment of the requirements set by BASEL II, three waves of changes happened in the financial sector. The combined impacts of the three changes on banks’ lending practice, especially on their attitude toward lending to the SME sector is one of the purposes of this study. The other purpose is to examine operation efficiency of each domestic bank and examine relative difference among them due to the three changes. In this study the author adopts the viewpoint of Intermediation Approach, selects interest expense, non-interest expense as input variables; interest income, non-interest income, the transformed value of SME-loan amount as output variables. It examines operating performance, the changes of productivity, efficiency stability of 36 domestic banks by means of DEA, Malmquist-index and Window analysis from 2001 to 2005 . Empirical results of cross-section analysis conclude that 32% are the efficiency banks. In the non-efficiency banks, 63% of them caused by the non-pure technical efficiency. In non-scale efficiency banks, 35% are large-scale banks, while 55% is small bank. The large-scale banks increase in number because of expands or the merges, but the efficiency banks of all banks actually reduce in number. Empirical results of vertical-section analysis conclude that 32% are the efficiency banks. The medium-sized and the large-scale banks’ 5 year technology efficiencies are higher than that of all banks while that of the small bank is lower. The average variation of production technology is smaller than that of the overall technical efficiency. In the Window analysis, the average overall efficiency values of medium-sized banks are the best. However the range of fluctuation by all banks is unstable. In addition, this study uses regression analysis to investigate the influence of other factors on efficiency of banks. The return on assets, the ratio of deposits to loans and the business scale present positive correlation demonstrate while it shows a negative correlation in terms of the volume of the loans covered by credit guarantee. The default rate of loans covered by credit guarantee shows a positive correlation and has notable influence. The nonperforming loan ratio which is also used to appraise the quality of assets has a negative correlation but has not much influence. Perhaps these contradictory phenomena are the evidence of the inconsistence between the efficiency of lending to SMEs by the banks and the efficiency of utilizing credit guarantee