Why is the Inter-firm Credit Market in Korea Special? An Agency View of Trade Credit Use by Chaebols

Abstract

We study inter-firm credit (also known as trade credit (TC) or the delayed payment a supplier allows its downstream customer on a product sale) with an emphasis on its unique features in Korea. It is the prominence of chaebol-affiliated firms in the Korean economy that makes Korea unique because chaebol firms are linked together by common ownership and thus distort the assumption common to all theories of TC that the input supplier is independent from its customer. We find several aspects of TC use by chaebols. First, viewed as input purchasers, these firms take more TC (the chaebol effect). The amount of TC taken is positively related to purchases from other chaebol affiliates (the affiliate effect), and with the degree of the chairman's ownership-control disparity (the ownership multiplier effect). Also TC as proportion of short-term debt increases significantly after profit growth (the profit effect). Furthermore, chaebol members give less TC although this is strongly correlated with transactions with affiliates. We suggest a unifying interpretation of these patterns based on agency cost: TC, by substituting for external financing and reducing monitoring by outsiders, presents controlling minority owners the opportunity to misuse their greater control for their own benefit.

Key Words:

• Inter-firm credit
• trade credit
• agency cost
• chaebol
• ownership-control disparity

Jel Classification:

• D20
• G30
• G32
• G34

Acknowledgement

This work was supported by Konkuk University in 2012.

Notes

1. Credit provision may also flow in the opposite direction as, for example, when the downstream firm pays prior to input delivery. In such a case, the payment is recorded as “prepaid expense” by the downstream firm and as “advance revenue” by the upstream firm. Advanced payment is less common than delayed payment. In Korean data, delayed payment makes up more than 13% of sales and cost, whereas advance payment accounts for less than 6% of sales and cost. We therefore leave advanced payment for future studies.

2. Some international studies include Korean firms in the analysis, e.g. Love, Preve, and Sarria-Allende (2007).

3. The agency problem in Korean corporations, especially in chaebols, is still an active research agenda. See, for example, Kim et al. (2013) and Sung and Park (2014).

4. An empirical test of the externality theory is made more difficult since the theory requires precise comparison of margins of any two firms. As Jeong and Kim (2012) have verified in the Korean manufacturing firms, margin varies significantly over time, making estimates from short time-series imprecise.

5. Media often characterizes delayed payment as an example of “abuse of market power” by large customer firms. According to a 28 July 2012 article of Korea Times (title: Minister voices concern on corporate bashing), Knowledge Economy Minister Hong Suk-Woo “pointed out that most of the criticism of chaebol stems from their wrong business practices, such as delaying payments or demanding excessive discounts from smaller suppliers.” According to a 16 May 2013 article of Korea Herald (title: FTC expands antitrust probe into companies), the Fair Trade Commission “launched an inquiry into Cheil Worldwide Inc., an advertising unit of Samsung Group, on suspicions that it forced price cuts or delayed payment in deals with its subcontractors.”

6. The sample period has been determined by the data availability. Chaebol-related variables are not available prior to the beginning of our sample period. Our sample period includes two major events – credit card crisis of 2003 and global financial crisis of 2008. To assess the sensitivity of our results to these two events, we have repeated our main analysis excluding 2003 and 2008 and have obtained essentially identical results.

7. We identify the observations with extreme values based on the right-hand side variables only. An observation is declared extreme (and is dropped) if value x of variable X is greater than the mean plus three standard deviations after log transformation (i.e. ). We make the log transformation as the distribution is smoother after the transformation. For those variables that are always positive (e.g. inventory ratio), we only look for extremely large positive values, as just described. For those variables that can be negative (e.g. ROA), we check for extremely small negative values as well, in a similar manner.

8. The numbers for NCBG are approximate because the data do not inform us which firm belongs to which (non-chaebol) group. We assign a firm as independent in a particular year if it reports zero transactions with affiliates in that year. Thus, the same firm can be classified as “independent” in one year and as “a member of a NCBG” in another year, so the total number of NCBG firms in the sample cannot be exactly determined. For chaebol members, our identification is constant across years.

9. The controlling shareholder typically holds the “chairman” title of the largest member firm. For example, Lee Kun Hee is the controlling shareholder of Samsung Group, and his official title is the chairman of the board of directors of Samsung Electronics, the largest company of Samsung Group.

10. Unfortunately, the chaebol member listings are unavailable for later years of our sample. We therefore apply the 2008 assignments to the years 2009–2011.

11. In principle, we could have constructed four variables from the Annual Audit Report but spot checks indicated that some firms reported “sales to affiliate” as “income for affiliates.” In the interest of conservatism, we elected to construct more consistent variables.

12. We have tried an alternative measure of “profit shock”: instead of the difference in ROA, we have defined two dummy variables – the positive shock and the negative shock. The positive shock dummy is 1 if the change in ROA is greater than 10% (which is approximately the standard deviation of the change in ROA), and 0 otherwise. The negative shock dummy is 1 if the change in ROA is less than −10%. We obtain essentially the same results from the alternative profit shock measure.

13. The preference for internal funds is influenced by a number of other factors including financial market conditions as Zulkhibri (2013) emphasizes.

14. Exporters may be different from non-exporters in many other ways, as shown by Aw and Song (2013), and this difference may influence their short-term financing behaviours as well.