臺灣金控與非金控多角化經營績效比較

Abstract

本研究探討了 2016–2023 年期間，台灣金融控股公司（FHCs）與獨立商業銀行（non-FHCs）採取的收益多元化策略，對傳統獲利能力及風險調整後報酬的影響。本文以台灣經濟新報資料庫中 14 家 FHCs 與 20 家 non-FHCs 的經審計年報為基礎，構建了收入結構指標（利息收入比重、非利息收入比重及赫芬達爾–赫希曼集中度指數）、核心績效指標（資產報酬率 ROA、權益報酬率 ROE）、風險調整後指標（RAROA、RAROE）與淨利差，同時控制成本效率、成長性、資本結構與流動性等變數。運用描述性統計、群組比較（t 檢定與 Wilcoxon 等級和檢定）及固定效果面板迴歸，我們發現營運費用率是影響全體銀行 ROA 與 ROE 的最強單一預測因子，凸顯了嚴謹成本管控對價值創造的關鍵性。 多元化策略所帶來的效益有限：對 FHCs 而言，較高的非利息收入比重可提升傳統獲利，但若未搭配完善的風險治理，透過收費業務擴張或快速資產成長拉高波動後，反而可能侵蝕風險調整後的績效。non-FHCs 則可藉由收入分散與選擇性收費增長提升 ROA，卻對費用與資本緩衝品質高度敏感。無論是規模擴張抑或加速資產成長，都不必然保證優異報酬；在兩類銀行中，過度追求規模或成長往往超出風險管理能耐，進而壓低 RAROE。適度的貸存比能增進風險調整後 ROA，而過高的資本適足率則可能帶來機會成本。 最後對 Tobin’s Q 進行的截面 OLS 分析顯示，子公司數量與市場評價間並無顯著關聯，凸顯擴大樣本或採用替代技術的必要性。整體而言，本研究證據表明，台灣銀行應透過嚴格的費用控制，並結合謹慎校準的多元化與成長策略，以及審慎的資本與風險監管，方能兼顧獲利與穩健。

This study examines how income‐diversification strategies affect both traditional profitability and risk‐adjusted returns at Taiwan’s financial holding companies (FHCs) versus stand‐alone commercial banks (non‐FHCs) over the 2016–2023 period. Drawing on audited annual financials for 14 FHCs and 20 non‐FHCs obtained from the Taiwan Economic Journal database, we construct measures of revenue mix (interest‐income share, non‐interest‐income share, and a Herfindahl–Hirschman concentration index), core performance indicators (return on assets and return on equity), risk‐adjusted metrics (RAROA, RAROE), and net interest margin, while controlling for cost efficiency, growth, capital structure, and liquidity. Employing descriptive diagnostics, group comparisons (t-tests and Wilcoxon rank‐sum tests), and fixed‐effects panel regressions, we find that operating expense ratio is the single most powerful predictor of both ROA and ROE across all banks, underscoring that disciplined cost management is a prerequisite for value creation. Diversification delivers modest gains: higher non‐interest‐income shares raise traditional returns for FHCs but, when scaled by volatility, fee‐based expansion or rapid asset growth can erode risk‐adjusted performance if not matched by robust risk governance. Non‐FHCs benefit from revenue dispersion and selective fee growth to boost ROA but remain highly sensitive to expense and capital‐buffer quality. Neither “getting bigger” nor accelerating asset growth guarantees superior returns in both groups, size and rapid expansion often outpace risk‐management capacity and depress RAROE. A balanced loan-to-deposit ratio enhances risk-adjusted ROA, while excessive capital adequacy may incur opportunity costs. Finally, cross‐sectional OLS on Tobin’s Q shows no significant link between subsidiary count and market valuation, highlighting the need for broader samples or alternative techniques. Overall, the evidence suggests that Taiwanese banks will optimize both profitability and stability by pairing rigorous expense control with carefully calibrated diversification and growth, supported by vigilant capital and risk oversight.