市場結構對銀行兼營電子支付業務獲利能力之影響

Abstract

自《電子支付機構管理條例》實施後，臺灣電子支付市場規模持續擴大，惟用戶數與交易金額主要仍集中於少數業者，形成「大者恆大」的競爭態勢。有鑑於此，本研究旨在探討電子支付市場結構是否會顯著影響銀行兼營該業務的獲利能力，並檢驗「相對市場力量假說」(Relative Market Power hypothesis，以下簡稱RMP假說) 與「結構─行為─績效假說」(Structure-Conduct-Performance hypothesis，以下簡稱SCP假說)於數位金融環境下的適用性。本研究以2020年至2023年間，臺灣19家取得電子支付執照之商業銀行為樣本，除了使用普通最小平方法作為對照模型，亦根據Hausman檢定結果，採用隨機效果模型進行縱橫資料分析。文中以「手續費淨收益占整體淨收益之比率」作為銀行經營電子支付業務獲利能力之代理變數，並分別將「個別銀行市占率」與「整體市場集中度」視為結構面的解釋變數，同時納入銀行內部特性與外部環境因素之控制變數。實證結果顯示，銀行滯後一期之代理收付交易金額市占率與手續費淨收益占比呈現不顯著正相關，僅部分支持RMP假說；而整體電子支付市場滯後一期之赫氏指數對銀行手續費淨收益占比呈現顯著正相關，支持SCP假說。研究結果可作為金融機構發展電子支付業務策略及主管機關制定監理政策之參考依據。After the implementation of the Act Governing Electronic Payment Institutions, Taiwan's electronic payment market has expanded rapidly, with substantial growth in both the number of users and transaction volumes. However, market activity remains concentrated among a few dominant providers, resulting in a "winner-takes-all" competitive landscape. This study aims to examine whether the market structure of electronic payments significantly influences the profitability of banks engaged in such services, and to assess the applicability of the Relative Market Power (RMP) hypothesis and the Structure-Conduct-Performance (SCP) hypothesis in this context.This study uses panel data from 19 Taiwanese commercial banks that obtained electronic payment licenses between 2020 and 2023. In addition to using the Ordinary Least Squares (OLS) method as a benchmark, the Random Effects Model is adopted for panel regression analysis based on the results of the Hausman test. The ratio of net fee income to total net income is used as a proxy for banks' profitability in electronic payment operations. The key explanatory variables include each bank's market share and the overall market concentration, measured by the Herfindahl-Hirschman Index, while control variables account for both bank-specific characteristics and external environmental factors. The empirical results show that banks' lagged market share of transaction value is positively but insignificantly associated with their net fee income ratio, only providing limited support for the RMP hypothesis. In contrast, the lagged Herfindahl-Hirschman Index of the overall electronic payment market is significantly and positively associated with banks' net fee income ratio, thereby supporting the SCP hypothesis. These findings provide practical implications for financial institutions developing strategies in the electronic payment sector, as well as for regulators formulating policies in response to changes in market structure.