Controlling Families’ Risk Allocation in a Business Group

• Author: Joon Chae,Hyung Cheol Kang,Yu Kyung Lee,Eun Jung Lee
• Date: 01 February 2020
• DOI: http://doi.org/10.1111/ajfs.12284
• Published date: 01 February 2020

Controlling Families’ Risk Allocation in a

Business Group*

Joon Chae

Graduate School of Business, Seoul National University, Republic of Korea

Hyung Cheol Kang

College of Business Administration, University of Seoul, Republic of Korea

Eun Jung Lee**

Department of Business, Hanyang University, Republic of Korea

Yu Kyung Lee

Graduate School of Business, Seoul National University, Republic of Korea

Received 24 June 2017; Accepted 23 October 2019

Abstract

This paper explores family business groups and their motivation for risk taking in each affili-

ate. We study whether the controlling family determines the level of risk taken by an affiliate

in its business group based on the amount of family wealth that is invested in the affiliate.

We find that the affiliates in which the controlling family has more (less) investment take less

(more) risk. Our results indicate that the controlling family decides the riskiness of each affil-

iate based on the family’s interests at both the firm level and the business-group level.

Keywords Controlling shareholder; Family firms; Risk taking; Business group

JEL Classification: G32, G11, G01, L22

1. Introduction

What determines corporate risk taking is an important issue in corporate finance .

The literature on a single firm’s choice to take risk assumes that the firm’s level of

risk can be determined at the firm level. However, corporate decisions that are

*This work was supported by the Ministry of Education of the Republic of Korea and the

National Research Foundation of Korea (NRF-2017S1A5A2A03067521). Joon Chae appreci-

ates the financial support of the Institute of Banking and Finance and the Institute of Man-

agement Research, Seoul National University.

**Corresponding author: Department of Business, Hanyang University, 55 Hanyangdaehak-

ro, Sangnok, Ansan, Gyeonggi-do, 15588, Republic of Korea. Tel: +82-31-400-5645, email:

ejunglee@hanyang.ac.kr.

Asia-Pacific Journal of Financial Studies (2020) 49, 67–98 doi:10.1111/ajfs.12284

©2020 Korean Securities Association 67

typically made at the firm level could deviate from those at the family-business-

group level (Morck and Yeung, 2003). We investigate how a controlling family that

pursues its interests at the family-group level decides on the level of risk each affili-

ate takes in the business group.

Many firms have a controlling shareholder who is often a founder’s family mem-

ber (La Porta et al., 1999). In several countries, a single individual or family controls a

large number of firms (Almeida and Wolfenzon, 2006); in Korea, these business

groups are known as chaebols. The controlling shareholder and general manager of a

chaebol is a family member known as a chongsu

1

in Korea. The chongsu has effective

control of all affiliates in the business group (Bae et al., 2012), and the risk taking of

each affiliate in the group is likely to be a consequence of the controlling shareholder’s

interest. In this paper, we examine whether the chongsu determines each affiliate’s level

of risk as they consider their overall interest in the business group.

Generally speaking, an investor allocates risk to maximize their portfolio’s per-

formance. In the pension- and asset-management industry, this process is referred

to as “risk budgeting.” For example, if a pension can tolerate up to 100 units of risk

in the entire fund, it can allocate 70 units of risk to blue chip stocks and 30 units

of risk to small cap stocks. An investor takes each asset’s risk level as given and

increases or decreases their portfolio’s risk level by changing the amount of money

invested in each asset. A chaebol’s controlling shareholder should also increase or

decrease their business group’s risk level, but they cannot easily increase or dec rease

the amount of money because they invest in each affiliate to maintain control over

these firms. In fact, the percentage of controlling shareholders’ shareholdings in

such affiliates does not change much over time (Lee et al., 2018). Still, they can

exercise unchecked power in all steps of the corporate decision-making process

(Bae et al., 2012; Hwang et al., 2013), especially when making decisions about cor-

porate risk taking (Su et al., 2017). In the asset management industry, riskiness is

an exogenous variable, and the amount of money invested in each asset is a

response variable; but in a business group, the amount of money invested in each

affiliate is an exogenous variable, and riskiness is a response variable. As a result,

the controlling shareholder tries to set a level of risk that can be tolerated at the

group level and then tries to decide each affiliate’s level of risk. Therefore, the

chongsu tends to allocate risk for each affiliate in their business group,

2

so a single

firm’s risk taking is not determined at the firm level but by the chongsu in pursuing

their own interests in allocating risk across affiliates.

The controlling shareholder can exercise control rights that affect corporate

decision-making to maximize their expected utility (Morck, 2000). Generally, a

1

Achongsu might not be the chairperson of the board or the CEO of any of the group’s main

affiliates, but they can control companies in the group through cross-shareholding.

2

The chongsu uses their authority to strategically allocate not only tangible resources but also

intangible resources like R&D and advertising intensities among affiliates in their business

group (Chang and Hong, 2000).

J. Chae et al.

68 ©2020 Korean Securities Association

risk-averse investor’s expected utility function is a decreasing function for the vari-

ance of their wealth (Faccio et al., 2011). If a chongsu is risk averse, they will

increase the expected utility by lowering the risk of affiliates in which they have

made significant investments. Therefore, when allocating the level of risk to each

affiliate in the business group, they strategically curtail the level of risk in affiliates

in which their money is heavily invested in order to pursue private benefit. Thus,

affiliates in a business group may take the risk for group-level reasons even when it

may not accord with their interests at the individual firm level.

We construct a variable, Controlling Shareholder’s Value (CSV), to investigate a

controlling shareholder’s risk-allocation behavior across affiliates. Following Kang

et al. (2017), we measure an affiliate’s CSV as the value of the controlling share-

holder’s shares in the affiliate, divided by the combined value of the controlling

shareholder’s shares in all of the group’s affiliates.

3

By using this variable, we can

more readily identify whether variations in the level of risk among the affiliates in a

group can be explained by the amount of the controlling family’s wealth that is

invested in each affiliate.

We focus on Korean business groups and use panel data on 1892 firm-years dur-

ing the period from 2001 to 2011. The Korean financial market is one of the best test-

ing platforms from which we can observe the controlling family’s behavior, as most

companies have concentrated family ownership structures and are part of busin ess

groups. To determine the relationship between corporate risk taking and the amount

of the controlling shareholder’s wealth invested in each affiliate, we follow Faccio et al.

(2011) in measuring firm riskiness (STD_ROA) as the volatility of a firm’s return on

assets (ROA) over a given five-year period. We find strong statistical evidence that

firms with higher CSV invest more conservatively than do those with lower CSV.On

average, an increase in CSV from the first to the third quartile of the distribution

results in a 4.79% decrease in the STD_ROA relative to its mean. This result suggests

that the choice regarding the amount of risk to be taken by each affiliate can be deter-

mined by the amount of the controlling shareholder’s wealth that is invested in the

affiliate. These results are also robust to including various controls suggested in the lit-

erature (Faccio et al., 2011; Boubakri et al., 2013; Kang et al., 2017).

Since our independent variable, CSV, does not change much over time, it is dif-

ficult to control for firm fixed effects. To address this concern, which is faced by

many studies in the ownership field, we employ a fixed effects vector decomposition

(FEVD) model, as in Pl€

umper and Troeger (2007). Another issue is that the results

3

Kang et al. (2017) argue that CSV contrasts with the commonly used ownership variable,

Cash Flow Right (CFR), measured as the percentage of an affiliate’s shares owned by the con-

trolling shareholder. For example, a controlling shareholder of a business group owns 100

and 10 shares of two affiliates, A and B, respectively, the outstanding shares in which are 200

and 100, and prices are $1 and $100. In this case, CFR

A

and CFR

B

are 50% and 10%, while

CSV

A

and CSV

B

are 9.1% (=$100/($100+$1000)) and 90.9% (=$1000/($100+$1000)), respec-

tively.

Controlling Families’ Risk Allocation

©2020 Korean Securities Association 69