Controlling Shareholders’ Preference in Business Groups: Evidence from Korea

ABSTRACT

This study examines the effect of controlling shareholders’ preference on the payout policy of Korean firms. Using a sample of 9495 firm-year observations, we find that firms with individual controlling shareholders (family-owned firms) have a lower payout ratio than those with non-individual controlling shareholders. Further, firms with higher family-individual controlling shareholder ownership by individual controlling shareholders are reluctant to pay cash dividends in family business groups. These results are consistent with the conservative payout hypothesis. An additional test indicates that family business groups’ group-level payout tendency influences all group-affiliated firms’ payout policies. The results suggest that controlling shareholders’ preference for cash dividends determines payout policy.

KEY WORDS:

• Business group
• controlling shareholder
• payout policy

JEL:

• G32
• G35

Supplemental material

Supplemental data for this article can be accessed here.

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Notes

1. In a pyramidal ownership structure, controlling shareholders hold equities, either directly or indirectly, through other group affiliates (Almeida et al. 2011). Meanwhile, the financial situation, ownership structure, and managerial structure of each business group are different. Thus, although controlling shareholders control the strategic corporate decisions of all of the group affiliates, the priorities of group affiliates for controlling shareholders are generally relative (Lee, Oh, and Park 2014).

2. In psychology, Mischel et al. (2010) conduct a “marshmallow test,” a typical experiment to examine hyperbolic discount, anomalies in inter-temporal choice, and delay of gratification. Laibson (1997) also shows significant bias toward the present time preference due to an increased discount rate.

3. In Korea, there were 43 family business groups comprising 1565 affiliated firms in 2012; family members held 172 director positions, and their relatives held 354 director positions in 2012. This shows that business group affiliates with a controlling shareholder often hire family members of the shareholder as top officials; thus, firms’ policies are affected by the interests of the controlling shareholder (Chang 2012; Minichilli, Corbetta, and MacMillan 2010).

4. We use 9495 firm-year observations that include 5385 firm-year observations with an individual controlling shareholder and 4110 firm-year observations with a non-individual controlling shareholder.

5. We refer these control variables from Hwang et al. (2013) and Brockman and Unlu (2009).

6. For more information about the model, see the Supplementary Material, available online.

7. DART is an electronic disclosure system in Korea. It is similar to the EDGAR system in the United States.

8. Several previous studies on the same topic use these databases to collect financial data. Black, Jang, and Kim (2006) and Kang and Kim (2006) use TS2000. In this study, we refer to both databases to collect clear data. FN Dataguide is a database similar to Thomson Reuters DataStream. Recent financial studies have often used this database to collect annual financial data and stock prices in Korea. For example, Joh and Jung (2012), Lee, Park, and Shin (2009), and Choe and Yang (2010) use this database for their studies in Korea.

9. Table S1 reports the annual distribution of the sample firms (available online).

10. Table S2 shows the payout policy for each type of representative controlling shareholder in more detail (available online). We classify representative controlling shareholders into six types: individuals, group affiliates, financial institutions, governments (and state-owned firms), foreign industrial firms, and NGOs (non-profit organizations).

11. We use 1% below the DPA dummy and 0.5% below the DPA dummy to re-examine the results.

Additional information

Funding

This work was supported by Global Research Network program through the Ministry of Education of the Republic of Korea and the National Research Foundation of Korea (NRF-2016S1A2A2912421).