有別於過往企業融資行為之相關文獻多以最適資本結構、融資順位理論與融資增額單期行為模式為研究主軸，本文乃試圖由資訊不對稱的角度結合企業多期存量負債融資縱橫資料（Panel Data），以期解析我國企業多期直接與間接負債融資模式。本文以民國80年至89年間皆具十年完整財務資料之75家上市公司為樣本，援引資訊不對稱性推論之發行成本、負債代理成本、逆向選擇、成長機會等構面，進而控制股權結構與償債能力構面，並運用隨機效果Panel Tobit模型共同驗證渠等對企業直接與間接舉債融資決策的多期性、不偏性及一致性影響。實證結果顯示，若公司規模越大、規模經濟越顯著、資訊揭露程度越高、負債代理與監督成本越低、董監事持股比率越高或獲利能力越強，皆使得企業傾向以直接負債融資，而使間接負債比率下滑；然成長機會越高、機構投資人持股比率越高或長期償債能力越高的公司，多傾向間接負債融資，反減少其直接負債金額，而企業之逆向選擇問題越趨緩和，其直接與間接負債皆有所提升。是以，資訊不對稱性明顯影響企業多期直接與間接負債融資模式。

In contrast to prior studies in corporate financing centering on the optimal capital structure, pecking order theory, and one-period additional debt financing pattern, this paper examines the heterogeneous firms' multi-period financing arrangements of direct and indirect debt mainly from the viewpoint of asymmetric information. Specifically, this study empirically analyzes the impacts of the flotation costs, agency costs, growth opportunities, and adverse selection variables based on information asymmetry, as well as corporate governance and prepayment ability variables on the multi-period financing pattern of direct and indirect debt of 75 publicly traded companies during the period of 1991-2000. The results, estimated by panel data and panel tobit model, reveal that the enterprises with a larger size, a higher extent of information disclosure, lower agency and monitoring costs, a higher ownership ratio of board of directors, or higher profitability are more likely to finance from direct (or public) debt, which directly lowers the proportion of indirect (or private) debt. In addition, the firms with abundant growth opportunities, a higher ownership ratio by institutional investors, or higher repayment ability are more inclined to finance from indirect debt and thus lower the level of direct debt. Finally, the companies with lower adverse selection problem have both the higher level of direct debt and the higher ratio of indirect debt. Therefore, the impacts of asymmetric information on the firms' multi-period financing pattern of direct and indirect debt are verified.