UCLA

In this thesis, I empirically analyze the transactions between firms and their related parties. I construct a novel granular dataset containing South Korean firms' related-party transactions, including intra-group sales and purchases as well as loans and debts. On the topic that in many contexts has remained a black box due to the lack of data, this thesis aims to uncover novel stylized facts and analyze the determinants and implications of these transactions.

In Chapter 1, I focus on sales and purchases between related parties and study how measurement issues have critically influenced the literature's view on such trades. Despite the importance of intra-firm trades in theories of the firm, an empirical literature using proxy measures has documented surprisingly little such trade. I revisit this conclusion using the economy-wide firm-level data from Korea, where related-party trades are directly observable. I show that the true prevalence and volume of the trades are much greater than previous measures indicate. Past proxy measures that rely heavily on input-output tables appear to dramatically underestimate the trades, capturing only 17.6% of related parties that trade and 32.6% of their sales volume.

In Chapter 2, I propose novel methods to infer supply chains within business groups. Research into topics such as vertical integration has strove to discern firms within business groups that constitute a supply chain aside from those held for unrelated reasons. Due to the lack of detailed data, extant literature has had to rely heavily on proxy measures of such information, despite mounting evidence against their accuracy. Using random forest algorithms with detailed intra-group trade data from South Korea, I propose alternative methods to infer trade within ownership networks. Compared to traditional proxies that rely on input-output table coefficients, this measure shows substantially improved performances while requiring only a small amount of additional information that is easily accessible to researchers.

Chapter 3 focuses on another key aspect of related-party transactions, the flow of capital through loans. Utilizing intra-group transactions data of South Korean firms, this paper provides novel evidence that business groups' internal capital markets are utilized for the group, instead of the private benefits of controlling shareholders. While higher cash-flow rights of controlling group owners do not predict lending relationships to the firms, empirical evidence consistently points to loans flowing to firms with higher marginal benefit of capital. Proximities within ownership and supply chains are highly correlated with lending relationships, suggesting that major factors behind intra-group capital flows include concerns of performance spillover between firms, as well as extending and supporting internal supply chains. Finally, this paper documents novel stylized facts on a specific type of loan, trade credit, that highlight the need to analyze them in a different context from non-trade credit loans.