銀行對中小企業融資之私有訊息探討

臺灣之中小企業占全體企業家數比率逾97%、就業人數占全體就業人口比率亦近8成，是穩定經濟及促進發展的重要推手，並在提升國家競爭力及社會安定之地位上扮演舉足輕重之角色；惟在財務結構及融資能力方面不及具規模之大型企業，且資訊揭露程度較不透明，加上經營體質脆弱，因此違約風險機率相對偏高。隨著數位、金融科技、人工智慧及大數據的蓬勃發展，可以將企業之經營及財務等數據確實反應，降低受到資訊不對稱影響融資條件不佳之情形；銀行業者則可透過市場公開訊息及非公開之私有訊息，藉由分析技術，轉化為對中小企業融資風險的評估資訊。本研究蒐集國內某家銀行2019年、2020年去識別化共計5,252家中小企業之融資相關資料，利用二階段估計方法求得私有訊息數值，並就市場變數予以標準化後，據以衡量其影響私有訊息之程度。實證結果所算出之估計係數呈極為顯著，且存在私有訊息影響程度亦呈極為顯著，證實私有訊息確實存在。另外，根據進一步實證結果之正、負向私有訊息數值比較，銀行對中小企業融資時，受到負向私有訊息影響之程度，遠大於受到正向私有訊息影響之程度。

關鍵字

中小企業融資；授信政策；二階段估計方法；私有訊息

並列摘要

Small and medium enterprise (SME) accounts for more than 97% enterprises in Taiwan. Also, the employees of SMEs account for about 80% of domestic employed population. That is to say, SMEs play an important role in stabilizing economy as well as contribute to competitiveness and development of a nation. However, the capital structure and borrowing capacity of SMEs are usually poorer than large scale enterprises (LSEs). Besides, SMEs also have problems with information disclosure and operating structure. Therefore, the default probability of SMEs is higher than that of LSEs. With the prosper of digital development, financial technology, artificial intelligence and big data, it becomes possible that properly disclosing true business performance and financial data helps to reduce the negative influence of information asymmetry on financing conditions. Bankers can now take both public and private information into consideration to evaluate the condition of the enterprises. This study collected de-identified data of 5,252 SMEs from 2019 to 2020. We first numericalize the private information through hybrid regression model, and standardize market variables in order to evaluate the degree of influence on private information. The result suggests that there actually exists private information of SMEs and the market variables do have significant influence on private information. Moreover, we also found that the impact of negative information is greater than positive information on SMEs loans.