On Foreign Shareholdings and Agency Costs: New Evidence from China

ABSTRACT

This article examines the impact of foreign shareholdings on agency costs of Chinese firms from 2006 to 2012. The empirical results indicate that: (1) direct foreign shareholdings, in contrast to indirect foreign shareholdings, improve asset utilization, suggesting low agency costs; (2) qualified foreign institutional investors play a significant role in firms because they are less subject to political pressure, which is consistent with lower agency costs, but this effect could be eroded by government control; and (3) foreign shareholdings reduce the cost of equity and improve firm performance. The results contribute to the privatization of state-owned enterprises and the domestic/foreign ownership structure of firms.

KEY WORDS:

• agency costs
• cost of equity
• firm performance
• foreign shareholdings

JEL CLASSIFICATION:

• G12
• G32
• G34

Supplementary Data

Supplemental data for this article can be accessed on the publisher’s website.

Notes

^1. Lei, Lin, and Wei (2013) considered the agency problem between managers and shareholders is the major concern for SOEs, whereas the agency problem between large shareholders and minority shareholders is major concern for private firms.

^2. Since 2005, when China launched the split-share structure reform, converting nontradable shares into tradable shares, there has been a significant impact on listed firms. Thus, this article investigates the relationship between foreign shareholdings and agency costs starting in 2006.

^3. This article not only uses one-year lagged and two-year lagged foreign shareholdings together as instrumental variables (see Tables S8 and S9, available online), but also instrument the foreign shareholdings of each firm n by the average foreign shareholdings of the other n-1 firms in the Chinese market(see Tables S10 and S11, available online). All the results indicate that the relationships are robust.

^4. The PSM method includes the same control variables as used in earlier regression models on agency costs.

^5. According to Becker and Ichino (2002), all treated units are matched with the closest control unit in the nearest-neighbor matching method and a weighted average of all controls with weights that are inversely proportional to the distance between the propensity scores of treated and controls in the kernel matching method. Each treated unit is matched only with the control units within a radius r from the propensity score of the treated unit in the radius matching method.

^6. See Table S5, available online.

^7. Specifically, TOBINB and TOBIND are ratios whose numerator is measured by the sum of the value of tradable stocks, the book value of nontradable stocks and the book value of debt, and the denominator is defined by the difference between total assets and both intangible assets and goodwill. However, the prices of tradable stock used in TOBINB are calculated as the equity divided by paid-in capital.

^8. RE_PEG is the cost of equity based on the PEG model, $r_e=\sqrt{\frac{EP{S}_2-EP{S}_1}{P_0}}$. RE_OJ is the cost of equity based on the OJ model, $r_e=A+\sqrt{A^2+\frac{EP{S}_1}{P_0}[g_2-(\gamma -1)]}$, where$A=\frac{1}{2}[(\gamma -1)+\frac{DP{S}_1}{P_0}]$; $g_2=\frac{EP{S}_2-EP{S}_1}{EP{S}_1}$; $(\gamma -1)$is the long-term growth rate; P₀ is the current price; EPS₁ is earnings per share in the next year; EPS₂ is earnings per share in year 2; DPS₁ is the dividend per share in the next year; DPS₁ = EPS₁ × k, k is the dividend payout ratio.

Additional information

Funding

We are grateful to the National Natural Science Foundation of China (71473103) and the Fundamental Research Funds for the Central Universities (Jinan Linghang Project, 15JNHLH001) for supporting our research.