本文將關係人交易區分為正常關係人交易與異常關係人交易,並檢視其與產品市場競爭之關聯性。過去文獻發現在中國,產品市場競爭程度越高之公司,會進行較多正常關係人交易以降低交易成本。而關係人交易在台灣亦是企業間普遍的交易形式,因此本文欲探討台灣上市櫃公司是否亦有此現象。另一方面,許多文獻指出產品市場競爭可以替代部分的內部治理機制,產品市場競爭應可抑制公司進行異常關係人交易。然而亦有許多文獻認為產品市場競爭會導致公司有動機進行盈餘管理,而異常關係人交易通常是管理當局進行盈餘管理的手段,因此本文也欲探討產品市場競爭程度與異常關係人交易之關聯性。 本研究以2010年至2019年台灣上市櫃公司為研究樣本,實證結果顯示,產品市場競爭程度越高,企業越傾向進行正常關係人交易。區分電子業與非電子業後,發現電子業在競爭程度較高的環境下較不傾向進行正常關係人交易,而非電子業在競爭程度較高的環境下將會進行正常關係人交易以降低交易成本。另一方面,實證結果顯示較高之產業競爭程度會導致較多的異常關係人交易,可能原因為產品市場競爭將導致企業有進行盈餘管理的動機。另外本文也發現電子業在產業競爭程度較高的環境下會有較少之異常關係人交易,而非電子業則是與電子業相反的結果。 基於本研究之發現,在產品市場競爭程度較高的產業,正常關係人交易有助企業降低交易成本以維持競爭力,法規不需加以禁止。而產品市場競爭雖然被視為外部的監督機制,但亦會使企業產生盈餘管理的動機,而實證結果顯示產品市場競爭導致企業有盈餘管理的傾向較為明顯,因此有較高之異常關係人交易。
This paper provides an analysis about the relationship between product market competition and related party transactions. Base on the methodology of previous studies, this paper classifies related party transactions as normal or abnormal. Previous studies have shown that normal related party transactions can reduce transaction costs (Efficient transactions hypothesis). However, abnormal related party transactions are one of the earning management tools or wealth exploitation (Conflict of interest hypothesis). Prior research examining Chinese firms finds a significant and positive relationship between product market competition and normal related party transactions, consistent with efficient transactions hypotheses. Using a sample of Taiwan’s listed firms from 2010 to 2019, this paper also finds a significant positive relationship between product market competition and normal related party transactions. In addition, this paper finds a significant and negative relationship between product market competition and normal related party transactions in electronic industry, while non-electronic companies show a positive association. As to the test of the association between product market competition and abnormal related party transactions. This paper shows that there is a positive relationship between product market competition and abnormal related party transactions, consistent with earning management prediction. This paper obtains a similar result when separately testing electronic and non-electronic industry subsamples.