Agency cost of CEO perquisites in bank loan contracts

“The Toledo, Ohio, auto supplier, which emerged from Chapter 11 bankruptcy protection in January 2008, spent $2.3 million last year on chartered planes to fly its chairman and chief executive, John M. Devine, and its vice chairman and former CEO, Gary L. Convis, to and from their California homes, according to its latest proxy statement.” Wall Street Journal, “For some CEOs, the perks keep flowing”. April 4, 2009. “Executives keep health perks while work benefits are cut” ~ USA Today December 10, 2010.

The term “perquisite” is used in the text; the term “perk’’ is used for convenience in variable names.

We follow Andrews et al. (2009, 2017) in our classifications for which the detailed variable definitions are in the “Appendix”.

Kim and Shin (2017) investigate the ratcheting up of executive bonus targets after the compensation disclosure reform in 2006 and report that the ratcheting has asymmetric effects and fluctuates with differing degrees of equity compensation and investment opportunity. The authors also confirm that the performance and bonus targets are serially correlated. Ferri et al. (2018) apply the reform of the 2016 compensation disclosure rule as an experiment to examine whether the enhancement of compensation disclosure affects the information content of financial reports. The authors find a higher coefficient for earnings responses when firms have higher improvement in the quality of the compensation disclosure.

Adithipyangkul et al. (2011) also confirm that executive perquisites strengthen the incentive effect and productivity. Voßmerbaümer (2013) contends that work-related perquisites improve work efficiency and that the employer should pay for any tax expenses related to the perquisite. Lee et al. (2018) compare the association between business flights and possible leisure flights with firm performance by collecting data on the arrival information of corporate jets in different destinations. They find that business related trips are linked with better firm performance.

Luo et al. (2011) investigate the effect of bank ownership on Chinese firms between 1999 and 2006. They find that at the firm level, the total amount of perquisite offers are positively related to the interest and financial expenses in the accounting statement. Gul et al. (2011) also apply perquisite data at the firm level in China between 2001 and 2005 to test the linkage between perquisite expenditure and the quality of firm information. They report a negative association between perquisites and the quality of information, and this relation can be less significant when firms have better quality financial reporting. Xu et al. (2014) further examine the association between excess perquisites and the crash risk of state-owned corporations in China. They use firm-level perquisites between 2003 and 2010 to show that CEOs tend to withhold bad news in order to enjoy more perquisites that thus induces a higher crash risk. This relation is particularly strong when the CEO is close to retirement.

The majority of perquisites were reported in the “other annual compensation” and “all other compensation” section.

We also test for the Ln(Perquisite) (the natural log of perquisite), the result is available on request.

The choice of our control variables follows the studies in the bank loan literature (Graham et al. 2008; Hasan et al. 2014, 2017).

In addition, Henry (2008) indicates that institutional investors play an important role in externally monitoring managers such that firm valuation is positively related to institutional ownership. Bhojraj and Sengupta (2003) also find that greater institutional ownership and outside directors are associated with lower bond yields and higher credit ratings.

Demerjian et al. (2012) develop a DEA approach to establish an efficient frontier for firm efficiency. The input variables that contribute to firm efficiency and revenue production process are: Net Property, Plant, and Equipment (PP&E); Net Operating Leases; Net R&D; Purchased Goodwill; Other Intangible Assets; Cost of Inventory; and Selling, General, and Administrative Expenses (SG&A), and revenue is the output of DEA vector. Furthermore, a Tobit model is designed to regress the firm efficiency on factors that generate revenue for the firm (including size, market share, free cash flow, complex multi-segment, business cycle, and international operations) within each industry. They note that the residual of the Tobit model represents managerial ability.

Both the compensation and the governance literature have widely discussed the issue of gender, but it is also related to CEO behavior. Faccio et al. (2016) discuss impact of gender on CEO risk-taking tendency, Huang and Kisgen (2013) identify the relation between gender and degree of overconfidence, Sila et al. (2016) discuss the effect of gender on board governance. Since in this paper we investigate the association between loan spreads and CEO perquisites from an agency cost view, we consider gender is also a pertinent control variable for corporate governance and hence include it as a control.

To save space, we do not report the coefficients for the loan’s purpose and type nor for industry and year dummies in the tabulated table.

For consistency in the dependent and independent variables in the SEM model, we use one standardized measure of CEO perquisites, Log(Perk), in the model estimation.

The top three types of perquisites by percentage in total perquisite amount are (a) financial, (b) other, (c) air expenses, which are 49.08%, 23.91%, and 17.71%, respectively.