Executive compensation among Australian mining and non-mining firms: Risk taking, long and short-term incentives

Economic Modelling 64 (2017) 211–220

Contents lists available at ScienceDirect

Economic Modelling journal homepage: www.elsevier.com/locate/econmod

Executive compensation among Australian mining and non-mining firms: Risk taking, long and short-term incentives

MARK

⁎

Subba Reddy Yarram , John Rice UNE Business School, University of New England, Armidale NSW 2351, Australia

A R T I C L E I N F O

A BS T RAC T

JEL codes: G3 M520 C33 Q33

How firms determine the pay of their executive employees is a vital research area. In the Australian context, mining firms form a large portion of listed companies. These miners tend to have more volatile earnings, operate with less certainty and higher risk in relation to capital investment. We look at a sample of ASX listed miners and non-miners from 2005 to 2013. We note that miners pay their CEOs less (AUD 1 m vs AUD 1.5 m for non-miners) overall. However, we also note that miners tend to use enhanced contingent long-term remuneration arrangements to significantly boost the pay-performance relationship compared to non-miners particularly during the pre-GFC period. Curiously, non-miners tend to have more generous short-term contingent arrangements linking executive pay and performance. The GFC, as an event, has adversely impacted these arrangements, lessening the generosity of pay-performance among miners, while enhancing these arrangements among non-miners. Overall, the results of the study provide support for optimal contracting theory and do not generally support the managerial power approach for both mining and non-mining firms.

Keywords: Corporate governance Executive compensation Panel data models Resource booms

1. Introduction Executive pay continues to attract the attention of both the general public and policy makers. Old questions relating to the alignment of shareholder and managerial interests have been expanded to take into account how to achieve the best alignment of the risk preferences of these stakeholders (Rustam et al., 2013). The Global Financial Crisis (GFC), as an event, crystallised the concerns of many that executives were over-incentivised to take risks with shareholder owned assets. Hence, the effective design of compensation policies was seen as an important means of reining in pay structures and incentives that led to excessive risk-taking by chief executive officers (CEOs) (Bahaji and Casta, 2016; Chiara et al., 2016; Citci and Inci, 2016; Yeoh, 2015). A significant body of literature exists relating to CEO compensation and risk preference (Coles et al., 2006; Guay, 1999), with much of this work focusing on banking and manufacturing firms. Guay (1999) contends that firms with growth opportunities can gain more if executive compensation can be designed to motivate the risk-averse managers to undertake investments in high risk but positive net present value projects. Consistent with this, (Coles et al., 2006) find that compensation linked to higher sensitivity to stock price volatility encourages managers to undertake investments that are risky. Many other studies find important differences in pay structures in Australia, the US and other Organisation of Economic Cooperation and

⁎

Corresponding author. E-mail address: [email protected] (S.R. Yarram).

http://dx.doi.org/10.1016/j.econmod.2017.03.034 Received 3 June 2016; Received in revised form 15 March 2017; Accepted 31 March 2017 0264-9993/ © 2017 Elsevier B.V. All rights reserved.

Development (OECD) countries (Chalmers et al., 2006; Izan et al., 1998; Matolcsy and Wright, 2007, 2011; Matolcsy, 2000; Merhebi et al., 2006). An example is that fixed pay continues to be a larger part of total pay in the Australian corporate sector compared to the US, where incentive pay - particularly in the form of options - has been more widely employed. Most of the earlier studies on the Australian context analyse executive compensation prior to the onset of Global Financial Crisis (GFC). This study fills the gap in the literature by considering a more recent period of 2005 to 2013 when many Australian firms started including short and long-term incentives in their pay structures. Other sources of variance in executive compensation relate to the nature and economic context of the Australian corporate landscape. Australia's economy hosts a greater proportion of resource and mining firms compared to most other markets (Roarty, 2010). Such resource firms represent a significant proportion of the Australian Securities Exchange (ASX). Generally, the products of these firms have greater price volatility than manufactured good, accentuating the earnings volatility of the sector which includes very large firms such as BHP Billiton and Rio Tinto, two of the largest resources firms globally. According to the Australian Bureau of Statistics (ABS), the mining sector contributes over 8 percent to the GDP of Australia. As resource exports are a significant component of Australia's total exports (accounting for over 40 percent of total exports according to the

Economic Modelling 64 (2017) 211–220

S.R. Yarram, J. Rice

mining firms in the literature. This study therefore examines the executive pay of mining firms and compares it with non-mining firms in the Australian context. The findings of this study help inform corporate finance theory particularly in terms of designing compensation policies and identifies if there are differences in the pay levels and structure of pay of firms operating in resources and non-resources sectors of the economy. For a sample of 129 mining firms and 332 non-mining firms for the study period of 2005 to 2013, this study finds that mining firms on average pay their CEOs approximately $1 million dollar a year as total salary compared to $1.5 million paid in non-mining firms. While twothirds or more of the total compensation is fixed in nature there are important differences in the incentive pay structures of mining and non-mining firms. While mining firms pay a relatively higher proportion of long-term incentives, non-mining firms pay a relatively higher level of short-term incentive pay. Mining firms also show higher payperformance sensitivity in terms of long-term incentive pay, nonmining firms show higher sensitivity of short-term incentive pay. We find that the economic variables identified in previous literature have significant influences on the pay levels of Australian mining and nonmining firms. Overall, our results do not provide a strong evidence in favour of managerial power approach in Australia for both mining and non-mining firms. The rest of the paper is organized as follows. In the next section a review of relevant literature on executive compensation is provided followed by empirical analysis and discussion of findings in section three. The last section summarises and concludes the study.

ABS), the sector often exerts a significant impact on Australia's financial and foreign exchange markets. Driven by stronger resource prices before the onset of the GFC, Australia's economy experienced significant economic growth. The resources sector generated considerable trade and budget surpluses during this period. The mining sector has also made contributions to the development of physical infrastructure in remote areas of Australia and has also made significant investments for community welfare in the form of indigenous health and well-being and many parts of the country. Foreign investment in the mining sector is considerable and the foreign direct investment (FDI) from China alone stood at $11 billion in 2011 (Huang, 2015). During the greatest part of the period examined here, the Australian resources sector did relatively well financially. Key export prices for coal and iron ore (two major exports) rebounded from GFC lows due to policies, especially in China, aimed at enhancing infrastructure investment. This allowed the mining sector to thrive until resource prices eventually fell sharply from mid-2013 onwards. Hosseinzadeh et al. (2016), for example, find that the majority of mining firms in Australia have improved their overall efficiencies during recent years and suggest that there is scope for efficiency gains for the remaining firms. Australia is often described as having a two-speed economy (Jayasuriya and Cannon, 2015) where the fortunes of mining and non-mining firms are often countercyclical. This is partially explained by the Dutch Disease conundrum (Corden, 2012), where high commodity prices drive the currency up, reducing the competitiveness of other sectors of the economy. Given this, and also given the innate risk involved in exploration and extraction of mining reserves, executive pay structures and their relationship to risk are likely to differ significantly between mining and non-mining firms. Corporate taxation in Australia differs from other major OECD countries in relation to the presence of a tax imputation system. This arrangement allows firms to declare franked dividends to domestic investors who then are not subject to double taxation. The tax imputation may be of particular importance as companies have the ability to pay franked dividends and thereby the ability to attract equity investors. This particular feature is of considerable interest to resident investors who could use franking credits to offset their tax obligations. Similarly, given the relative tax-advantage of dividends, executives in Australia may derive more benefits from compensation arrangements that include shares. Corporate disclosure and financial reporting has undergone changes in the last two decades particularly with the adoption of Australian version of International Financial Reporting Standards from the beginning of 2005. Australia also boasts one of the most transparent executive compensation disclosure regimes based on the introduction of AASB 1046 Director and Executive Disclosures by Disclosing Entities in January 2004 and the release of AASB 124 Related Party Disclosures in 2009 (Walker, 2010). The Corporate Law and Economic Reform Program (CLERP) 9 introduced in 2004 has further strengthened executive compensation disclosures in Australia. The Australian financial markets are in general broad, deep and highly efficient with active participation of institutional investors as well as investor associations. Australia's corporate sector is highly professionalized with boards of directors generally appropriately trained and often professionally qualified. The ASX issued Principles of Good Governance and Best Practice Recommendations in 2003, with subsequent amendments made in 2007, 2009, 2010 and 2014 further strengthening the governance provisions in the Australian corporate sector. The highly developed nature of the financial markets and corporate disclosure regimes makes Australia an ideal context to analyse the levels and structure of pay. Further given the significance of the resource sector, it is important to consider separately the executive compensation issues in mining and non-mining firms in Australia. Despite this structural variance relating to risk and financial performance between the mining and non-mining sectors in Australia there has been scant attention paid to the issue of executive pay in

2. Literature review and theoretical framework Jensen and Murphy (1990), in their pioneering study on executive compensation, estimate the pay-performance sensitivities for US firms. They conclude that CEOs of US firms were paid like bureaucrats during the duration of their study – essentially remuneration was a function of scale. Subsequent studies found that pay-performance sensitivities generally increased in the US before the GFC (Murphy, 2013) with greater contingent rewards made available for corporate performance above expectations. Broadly there are two strands of literature dealing with executive compensation. Supporters of optimal contracting theory contend that the market for managerial labour market is competitive and executive compensation is determined by a set of economic factors relating to the businesses and individual CEOs (Core et al., 1999; Core and Larcker, 2002; Gabaix and Landier, 2008; Jensen and Murphy, 1990; Murphy, 1985, 1997, 2013). Supporters of the managerial power approach, on the other hand, argue that CEOs control the nomination process of boards of directors and as such often exert indirect influence on their own compensation levels (Bebchuk et al., 2011; Bebchuk and Fried, 2003; Yermack, 2006). Chalmers et al. (2006) find evidence of both approaches in Australia. They find that fixed salary component and share-based incentive components are explained by optimal contracting theory, while bonus payments and options grants are explained by the managerial power approach. As such, Murphy (2013) suggests that both the optimal contracting theory and managerial power approaches are relevant explanators for the compensation levels and structure observed in corporate firms. Prior literature identifies a variety of economic and governance factors that influence the compensation levels and structure of corporate firms. For example, the size of a firm and the nature of its business activity may influence the compensation levels as increased complexity requires sophisticated functional and managerial skills for executives (Core et al., 1999). Similarly, the performance of a firm is likely to influence the pay levels given the ability to attract high quality executives. Merhebi et al. (2006) find positive influence of size and performance on pay levels of CEOs in Australia. Lee (2009) also finds that size has a significant positive effect on pay levels of CEOs. For the 212

Economic Modelling 64 (2017) 211–220

S.R. Yarram, J. Rice

period 1999 to 2001, Matolcsy and Wright (2007) find that size and industry affect the level of executive compensation in Australian firms. They highlight the significant differences in compensation contracts between firms and find that a considerable number of companies do not include share based payments. Matolcsy and Wright (2011) find that performance has a significant influence on the compensation structures of Australian firms and that firms that performed poorly have inconsistent compensation structures whereas those firms that did well have more consistent compensation structures. Somewhat in contrast to these prior findings, Izan et al. (1998) find no relation between performance and CEO pay in the Australian context. Lee (2009) also finds that performance has no influence on the pay levels of CEOs in Australia. It should be noted, however, that these results pertain to an earlier period when disclosures relating to executive pay were poor and that governance framework was yet to formally evolve. Cairns and Davis (2001), Samis et al. (2005), Zhang (2014), Zhang and Kleit (2016) and Zhang et al. (2015) consider the issues involved in using traditional discounted valuation when analysing the firms operating in the resources sector and highlight the need for considering the uncertainties in commodity prices and the real options in investment analysis. Given the increased complexity involved in managing businesses in the recent period, and generally increasing global uncertainty and financial market volatility, this study hypothesises that size and performance have a significant positive influence on the CEO pay levels and structure in both mining and non-mining firms in Australia. A positive influence of size is justified as most mining firms in Australia do not have dominant shareholders and thus the competition for external CEO talent among medium to large firms tends to bidup overall CEO compensation. Similarly, a positive influence of performance is consistent with most compensation literature that seeks to align CEO interests with shareholders – and shareholder returns are optimised by positive short, medium and long term returns in general, and especially where there is no clear trade-off between these time periods. Growth opportunities may require careful management of new products or new markets and often, die to the managerial talent required, influence pay levels positively. Walker (2010) finds that size of firms and growth opportunities have a significant positive effect on the pay levels of Australia firms for the period 2004 and 2005. We therefore hypothesise that growth opportunities have a positive influence on the executive pay in the Australian context. External economic circumstances may clearly play a role in determining pay levels, as well as compensation structure, of firms (de La Bruslerie, 2016). Matolcsy (2000) finds that performance has a positive relationship on pay during economic booms, while performance has no influence on pay when the economy experiences recessionary conditions. Furthermore, Rankin (2010) finds that the GFC had a significant impact on the compensation structures of Australian firms. Rankin (2010) and Doucouliagos et al. (2007) analyse executive compensation levels and structure in financial sector and find that these firms typically pay a higher level of pay compared to nonfinancial firms. Governance factors are expected to influence the pay levels and structure of corporate firms (Rustam et al., 2013). The general principal-agent dilemma is clearly evident in the mining sector. Higher volatility of earnings and capital growth tend to coincide with a strong risk tolerance, and indeed risk preference, among shareholders and managers (Wang et al., 2013). However, managers of these firms tend to prefer a balance towards short term incentives to incentivise their behaviour, while shareholders and their Board representatives tend to prefer the balance to be greater in respect to longterm incentives (Gomez-Mejia et al., 1987; Kim, 2013; Zajac and Westphal, 1994). Large boards may at times be unwieldy, with this in turn leading to inefficiencies. Small boards on the other hand may not have the necessary expertise to effectively monitor the executives and

their performance. Given the complexity of operating environment and volatile commodity markets, larger boards may not lend to effective monitoring and governance particularly in mining firms and therefore we hypothesise that board size has a negative influence on the pay levels and incentives that place emphasis on long-term objectives. For similar reasons, we also hypothesise that board size has a negative influence on the pay-performance sensitivities of mining and nonmining firms. Similarly, board independence may have a significant influence on pay levels and incentive structures that are deployed. If boards are truly independent this may lead to effective monitoring. Conversely, if independent directors are nominated or have social connections with executives, this may result in effective monitoring through these connections. Given the developments in corporate governance and the principle rather than rule based governance frameworks in Australia, we expect a positive influence of board independence on pay levels and compensation structures that favour long-term value creation. We also hypothesize that board independence has a significant positive influence on the pay-performance sensitivities of mining and non-mining firms in Australia. CEO duality, on the other hand, may help a CEO entrench and therefore lead to higher pay levels and incentive structures that are short-term oriented particularly in nonmining firms. Armstrong et al. (2012); Beatty and Zajac (1990); Zajac and Westphal (1994) argue that powerful CEOs who hold a dual position as CEO and chairperson or having a better negotiation power over the board may mould their compensation to reflect their preferences by minimizing the long-term component of pay. We also consider the tenure of CEOs, the existence of a remuneration committee and the membership of the CEO on the remuneration committee as we anticipate that these factors have a bearing on the pay levels as well structure of pay. CEO tenure may have a positive influence on the pay levels given the possibility of CEO entrenching and controlling the boards and their activities. The mining sector in Australia has generally been more volatile in relations to earnings due, in part, to the strong cyclicality in the prices for minerals. We anticipate that CEOs that are recently appointed to mining firms may more aggressively negotiate short-term incentive-based pay as a means of securing a real option against any cyclical upside in the firms’ earnings. This situation may, in turn, increase the risk tolerance and indeed risk preference of CEOs appointed to mining firms. This may exhibit through the observation of decisions that enhance the potential for earnings upside by making investment decisions that expose their firms to higher risks and higher potential returns. Nagar et al. (2003) show that share based incentives reduce agency costs relating to managerial disclosures. Remuneration committees of firms are generally charged with better aligning the interests of shareholders and managers. They do this through the effective and deliberate structuring of executive compensation to incentivise managers to act in the long term best interests of shareholders – especially in relation to capital growth. CEOs, however, may try to minimize longterm incentives as they can only benefit from such arrangements only if performance improves on a long-term basis (Gomez-Mejia et al., 1987; Zajac and Westphal, 1994). As such, we anticipate that stronger CEOs, who are abler in exerting influence on remuneration arrangements, will tend to have a greater emphasis on short term incentives over longer term incentives. We anticipate that this is especially true in the mining sector where risk preference is higher than is otherwise the case. Setting up a stand-alone remuneration committee may lead to better pay sensitivities while it may or may not influence those CEOs’ pay levels. Similarly, when the CEO is a member of the remuneration committee it is possible that they may or may not influence the pay levels and structure of their respective firms. Substantial shareholders may have a similar long-term view as that of the shareholders and may have a significant negative impact on the pay levels of executives. This view is consistent with (David et al., 1998) who find that institutional investors with no conflicts of interests 213

Economic Modelling 64 (2017) 211–220

S.R. Yarram, J. Rice

transparent during the study period compared to the earlier period. All compensation and governance data is collected from the Governance Module of SIRCA for the period 2005 to 2013. Similarly, information on financial variables is collected from the Worldscope / Thomson Reuters Datastream Database. Table 1 provides selected descriptive statistics for mining and nonmining firms for the sample period spanning 2005 to 2013. CEOs of mining firms on average were paid $1,012,723 a year compared to $1,515,973 for CEOs in non-mining firms. Fixed pay accounted for over two-thirds of total pay in both mining and non-mining firms during the study period. Similarly, CEOs in both mining and nonmining firms received higher levels of long-term incentive pay rather than short-term incentives. Mining firms specifically favoured employing long-term incentives rather than short-term incentives during the study period. Mining firms typically have smaller boards that are less independent compared to non-mining firms. Both mining and non-mining firms have largely separated the roles of CEO and Chairperson, with only 9 percent of firms having a sole occupant of these two roles. Mining firms also had weaker governance structures compared to non-mining firms in terms of having a separate remuneration committee and independence of the remuneration committees, while nonmining firms included CEOs on their remuneration committees more often. Similarly, substantial shareholders own a greater share of issued capital in non-mining firms compared to the mining firms. Mining firms on average are smaller, less profitable and have lower indebtedness compared to non-mining firms. On the other hand, during the study period, mining firms had higher growth, market returns and returned higher performance as measured by Tobin's Q.

influence the overall pay levels negatively. We therefore propose that substantial shareholding has a negative influence on the pay levels and pay-performance sensitivities of both mining and non-mining firms. Given the earlier literature, and our propositions specifically relating to pay levels and structure in the previous discussion, this study analyses the influences of economic and governance factors on the pay levels and pay structure of mining and non-mining firms in Australia. Further, this study estimates the pay-performance sensitivities separately for each component of pay and also in the pre, during and post GFC periods. Lastly this study considers the influence of governance variables on the pay performance sensitivities of mining and non-mining firms by incorporating interaction variables that relate changes in market capitalization and specific governance variables. 3. Empirical analysis and discussion This study employs a sample of firms listed on the ASX and members of the All Ordinaries Index for the period 2005 to 2013. All financial firms are excluded given the highly regulated nature of the financial sector. The final sample consists of 129 mining firms and 332 non-mining firms, with a total of 862 mining-firm-years and 2373 nonmining-firm-years for the study period of 2005 to 2013. The study period of 2005 to 2013 is chosen as this period represents an era of a new corporate governance regime in Australia. The ASX issued Principles of Good Corporate Governance and Best Practice Recommendations in 2003, with companies adopting these principles soon after. Similarly, Australian firms started adopting International Financial Reporting Standards (IFRS) in 2005 and the level of disclosure relating to executive compensation is considerably more Table 1 Descriptive statistics for mining and non-mining firms. Mining firms

FixedSalary STI LTI TotalSalary %FixedSalary %STI %LTI BrdSize BrdInd CeoDuality CeoTenure RemComDummy RemComInd CEORemCom %SubSH Tobin's Q TotalReturn Ln(TotalAssets) Growth Leverage EBIT(000 s) Mcap(000 s)

Non-mining firms

No of firms

Average

Median

SD

No of firms

Average

Median

SD

862 862 862 862 862 862 862 862 862 862 862 862 862 862 862 812 813 820 561 815 803 812

$508,148 $137,216 $445,780 $1,012,723 0.69 0.08 0.20 6.21 0.71 0.08 6.54 0.52 0.48 0.09 0.28 2.29 0.26 11.74 0.59 0.07 $35,483 $734,832

$411,846 $0 $55,350 $589,806 0.73 0.00 0.11 6.00 0.75 0.00 5.00 1.00 0.67 0.00 0.26 1.83 0.00 11.80 0.31 0.00 -$1844 $215,448

$410,973 $343,383 $1,080,764 $1,156,475 0.28 0.12 0.23 2.34 0.15 0.27 5.08 0.50 0.48 0.29 0.21 1.59 0.81 1.69 0.70 0.12 $132,872 $1,571,766

2373 2373 2373 2373 2373 2373 2373 2373 2373 2373 2373 2373 2373 2373 2373 2310 2330 2326 1887 2310 2303 2312

$808,493 $395,116 $423,426 $1,515,973 0.66 0.14 0.17 7.07 0.75 0.08 8.29 0.68 0.62 0.20 0.33 1.74 0.15 12.85 0.23 0.18 $106,654 $1,329,171

$601,501 $95,000 $106,246 $951,073 0.65 0.12 0.12 7.00 0.80 0.00 7.00 1.00 0.80 0.00 0.31 1.24 0.02 12.86 0.12 0.14 $28,004 $375,325

$600,702 $993,608 $849,234 $1,432,157 0.24 0.15 0.18 2.30 0.14 0.27 6.14 0.47 0.44 0.40 0.26 1.39 0.59 1.81 0.41 0.18 $189,962 $2,117,864

This table provides average, median and standard deviation variables included in the study for the period 2005 to 2013. These statistics are calculated separately for mining and nonmining sample firms. FixedSalary is the sum of base salary, superannuation, value of non-pecuniary benefits, other compensation and value of long-service leave. Short-term incentives (STI) is computed as the sum of cash bonus and short-term compensation. LTI is long-term incentives and is computed as the sum of long-term compensation and other long-term compensation. TotalSalary is the sum of FixedSalary, STI and LTI. BrdSize is measured as the number of board of directors. BrdInd is measured as the proportion of non-executive directors to total number of directors. CeoDuality is a dummy variable that takes a value of 1 when the CEO and Chairperson roles are performed by the same person otherwise it is set to equal 0. CeoTenure is the length of service of a CEO. RemComDummy takes a value of 1 if a firm has a remuneration committee otherwise it is set to equal 0. RemComInd is measured as the proportion of non-executive directors to the total number of directors on the remuneration committee. CEORemCom is a dummy variable that takes a value of 1 if CEO is on the remuneration committee otherwise it is assigned a value of 0. %SubSH is the proportion of shareholding held by the substantial shareholders who hold 5% or more shareholding in a firm. Tobin's Q is computed as the ratio of the sum of market value of equity and the book value of liabilities to total assets. Profitability is measured as a proportion of earnings before interest and taxed to total assets. TotalReturn is measured as the sum of capital gains yield plus the dividend yield. Mcap measures the market capitalization of a firm. Ln(TotalAssets) is the natural logarithm of total assets. Growth is measured as average sales revenue growth over the last 5 years. Leverage is measured as a proportion of the book value of total debt to the sum of book value of total debt and market value of equity.

214

Economic Modelling 64 (2017) 211–220

S.R. Yarram, J. Rice

Table 2 Correlations.

TotalSalary BrdSize BrdInd CeoDuality CeoTenure RemComDummy RemComInd CEORemCom %SubSH Tobin's Q Profitability TotalReturn Mcap(000 s) Ln(TotalAssets) Growth Leverage

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

1 0.52* 0.30* −0.06* 0.18* 0.22* 0.24* 0.07* −0.08* −0.11* 0.67* −0.03 0.73* 0.67* −0.14* 0.17*

1 0.33* −0.13* 0.04 0.25* 0.27* 0.06* 0.06* −0.22* 0.49* −0.15* 0.52* 0.61* −0.15* 0.28*

1 −0.15* −0.01 0.23* 0.29* 0.06* −0.02 −0.20* 0.23* −0.13* 0.24* 0.35* −0.14* 0.17*

1 0.15* −0.11* −0.14* 0.04 0.04 0.10* −0.08* 0.03 −0.09* −0.11* 0.01 −0.06*

1 0.09* 0.08* 0.10* −0.05* −0.04 0.14* −0.01 0.11* 0.14* −0.08* 0.03

1 0.97* 0.34* 0.09* −0.23* 0.15* −0.09* 0.15* 0.36* −0.03 0.18*

1 0.15* 0.08* −0.23* 0.16* −0.11* 0.17* 0.37* −0.04 0.18*

1 0.06* −0.09* 0.06* −0.01 0.05* 0.10* −0.04 0.08*

1 −0.04 −0.11* −0.01 −0.12* 0.05* −0.08* 0.12*

1 −0.08* 0.46* 0.01 −0.39* 0.10* −0.47*

1 0.01 0.87* 0.66* −0.10* 0.13*

1 0.03 −0.13* 0.04 −0.28*

1 0.69* −0.09* 0.07*

1 −0.07* 0.47*

1 −0.11*

1

This table shows the correlations between variables employed in the study for the overall study period of 2005–2013. Profitability is measured as a proportion of earnings before interest and taxes to total assets. Descriptions for all other variables are provided in Table 1. * indicates significance at 1%.

measured as the sum of capital gains yield plus the dividend yield. Mcap measures the market capitalization of a firm. Ln(TotalAssets) is the natural logarithm of total assets. Growth is measured as average sales revenue growth over the last 5 years. Leverage is measured as a proportion of the book value of total debt to the sum of book value of total debt and market value of equity. Table 3 presents the results of panel random effects models. Total compensation and its components are separately regressed on a set of economic and governance variables. Average industry pay levels have a significant positive influence on pay levels in individual firms, with this true for both mining and non-mining firms. This finding is consistent with Chen (2010) who found that the external labour market plays an important role in the determination of pay of corporate executives. Economic variables as identified in previous literature have significant influences on the pay levels of Australian mining and non-mining firms. Firm size has a significant positive influence on pay levels. Similarly performance as measured by Tobin's Q has a significant positive influence on pay in both mining and non-mining firms. These findings are consistent with earlier studies (Lee, 2009; Matolcsy and Wright, 2007; Merhebi et al., 2006). Leverage have a significant negative influence on the total compensation in both mining and non-mining firms. These findings are consistent with earlier research in the area (Schultz et al., 2013). Growth, on the other hand, has a significant negative influence on the pay level of both mining and non-mining firms in the Australian context. This finding contradicts the finding of many previous studies including Walker (2010). Bizjak et al. (1993) find that high-growth firms employ lower salary or bonus incentives and lower total incentives compared to low-growth firms. The negative influence of growth opportunities on executive pay is similar to the finding of (Yermack, 1995). Broadly, the role of economic factors on the level and structure of compensation is consistent with optimal contracting theory for both mining as well as non-mining firms. Governance factors influence pay levels and structure in varying ways in both mining and non-mining firms. Board size has a significant negative influence on the fixed pay of both mining and non-mining firms, but has no influence on the short-term incentive pay of both mining and non-mining firms. Board size has a significant negative influence on the long-term incentive pay of non-mining firms while it has no significant influence on the long-term incentive pay of mining firms. Board independence has a significant positive influence on the fixed pay of both mining and non-mining firms. Board independence also has a significant positive influence on the short-term incentive pay of mining firms while it has a significant positive influence on the longterm incentive pay of non-mining firms. CEO duality has a significant negative influence on the fixed pay of mining firms, while at the same

Table 2 shows the correlations among variables employed in the study. Larger firms have higher compensation. Similarly, profitability is positively correlated to total compensation. Growth, performance and higher levels of substantial shareholding on the other hand are negatively related to total compensation. Firms that have remuneration committees also show that these committees are highly independent. The econometric analytical framework starts with modelling pay levels as a function of previously identified economic characteristics, ownership and governance variables. Consistent with (Core et al., 1999; Schultz et al., 2013), we estimate the following random effects panel data model with cluster robust standard errors

ln(Payit ) = Xit′β + (αi + εit ) Where, Payit alternatively includes log transformed fixed pay, shortterm incentive pay, long-term incentive pay and total pay, Xit′ in a set of log transformed economic, ownership and governance variables, αi are firm-specific effects that are assumed to be uncorrelated with the regressors and εit is an idiosyncratic error. A Hausman test (results not reported) showed that there is no evidence that the firm-specific effects are correlated with the independent variables and therefore we use the random effect panel data models of compensation. Consistent with earlier literature (see, for example, Core et al., 1999; Schultz et al., 2013), we employ board composition, CEO tenure and membership of the CEO on remuneration committee as explanatory variables. We specifically define these variables in a manner consistent with earlier literature. BrdSize is measured as the number of board of directors. BrdInd is measured as the proportion of non-executive directors to total number of directors. CeoDuality is a dummy variable that takes a value of 1 when the CEO and Chairperson roles are performed by the same person otherwise it is set to equal 0. CeoTenure is the length of service of a CEO. RemComDummy takes a value of 1 if a firm has a remuneration committee otherwise it is set to equal 0. RemComInd is measured as the proportion of non-executive directors to the total number of directors on the remuneration committee. CEORemCom is a dummy variable that takes a value of 1 if CEO is on the remuneration committee otherwise it is assigned a value of 0. %SubSH is the proportion of shareholding held by the substantial shareholders who hold 5% or more shareholding in a firm. We also employ a set of firm-specific economic variables identified in previous literature (see, for example, Core et al., 1999; Schultz et al., 2013) as having influence on pay levels. Tobin's Q is computed as the ratio of the sum of market value of equity and the book value of liabilities to total assets. Profitability is measured as a proportion of earnings before interest and taxed to total assets. TotalReturn is 215

Economic Modelling 64 (2017) 211–220

S.R. Yarram, J. Rice

Table 3 Pay and governance. Variables

Ln(FixedSalary-IndAvg)

Ln(Fixed Pay) Mining Model 1

Non-mining Model 2

0.582*** (3.91)

0.366*** (5.40)

Ln(STI-IndAvg)

Ln(Short-term incentives)

Ln(Long-term incentives)

Ln(Total Pay)

Mining Model 3

Non-mining Model 4

Mining Model 5

Non-mining Model 6

Mining Model 7

Non-mining Model 8

0.584** (2.05)

0.347*** (4.29) 1.848** (2.02)

1.930*** (5.92) 0.597*** (3.86) 0.392*** (13.11) 0.138*** (6.51) −0.126*** (−2.77) −0.239 (−1.00) −0.053*** (−2.71) 0.738** (2.55) −0.047 (−0.37) 0.008 (1.24) 0.182*** (2.61) 0.078 (0.77) −0.720*** (−4.19) 560 0.526

0.364*** (5.48) 0.381*** (22.40) 0.074*** (5.15) −0.091** (−2.25) −0.566*** (−4.90) −0.027** (−2.56) 1.185*** (7.47) −0.197*** (−2.64) 0.019*** (6.20) 0.048 (1.23) −0.020 (−0.47) −0.178** (−2.20) 1883 0.540

Ln(LTI-IndAvg) Ln(TotalSalary-IndAvg) Ln(TotalAssets) Tobin's Q Growth Leverage BrdSize BrdInd CeoDuality CeoTenure RemComDummy CEORemCom %SubSH Firm-years R2 Overall

0.349*** (14.47) 0.062*** (3.52) −0.037 (−1.01) −0.021 (−0.11) −0.062*** (−3.84) 0.950*** (4.02) −0.183* (−1.70) 0.030*** (5.45) 0.067 (1.15) 0.135 (1.59) −0.320** (−2.24) 550 0.575

0.303*** (19.97) 0.043*** (3.26) −0.115*** (−3.09) −0.239** (−2.24) −0.025*** (−2.59) 1.164*** (7.96) −0.067 (−0.96) 0.024*** (8.37) 0.040 (1.10) −0.016 (−0.43) −0.173** (−2.32) 1861 0.521

0.341*** (4.83) 0.130** (2.37) −0.303*** (−2.85) −0.812 (−1.37) 0.024 (0.53) 1.198* (1.69) −0.216 (−0.57) 0.015 (0.95) 0.254 (1.33) 0.380 (1.63) −0.909** (−2.14) 250 0.367

0.508*** (16.68) 0.057** (1.97) −0.185** (−2.32) −0.912*** (−4.11) −0.019 (−0.96) −0.177 (−0.59) 0.406** (2.34) 0.011** (2.01) −0.062 (−0.83) 0.124* (1.72) −0.175 (−1.22) 1190 0.449

0.795*** (2.98) 0.508*** (2.74) −1.018** (−2.53) −1.093 (−0.52) −0.095 (−0.55) 2.861 (1.13) −0.317 (−0.28) −0.149** (−2.49) 1.537** (2.54) −1.651* (−1.85) −4.407*** (−2.92) 560 0.137

1.123*** (8.50) 0.377*** (3.16) −0.213 (−0.64) −2.475*** (−2.59) −0.160* (−1.82) 4.680*** (3.57) −2.734*** (−4.50) −0.045* (−1.79) 0.182 (0.56) −0.143 (−0.41) −1.276* (−1.92) 1883 0.165

This table provides the results relating to analysis of the determinants of CEO pay and its components. Ln(FixedSalary-IndAvg) is natural logarithm of industry average FixedSalary. Ln (STI-IndAvg) is natural logarithm of industry average STI. Ln(LTI-IndAvg) is natural logarithm of industry average LTI. Ln(TotalSalary-IndAvg) is natural logarithm of industry average TotalSalary. Descriptions for other variables are provided in Table 1. Note: t-statistics are provided in parentheses. ***, ** and * denote significance at 0.01, 0.05 and 0.10 respectively.

(see, for example, Core et al., 1999; Schultz et al., 2013), we estimate the following random effects panel data model

time it has a significant positive effect on the short-term incentive pay of non-mining firms. We observe that it has a significant negative influence on the long-term incentive pay of non-mining firms. CEO tenure has a mixed influence on the pay levels of both mining and non-mining firms. While it has a significant positive influence on the fixed pay of both mining and non-mining firms, it has only a positive influence on the short-term incentive pay of non-mining firms. In contrast, it has a significant negative influence on the long-term incentive pay of both mining and non-mining firms. A stand-alone remuneration committee has a significant positive influence on the long-term incentive pay of mining firms, while it has no influence on the fixed pay as well as short-term incentive pay of both mining and non-mining firms. Membership of a CEO on the remuneration committee has no influence on the fixed pay, while it has varying influences on incentive pay. When a CEO is a member of a remuneration committee, he / she exerts a positive influence on the short-term pay of non-mining firms, while at the same time, he / she seems to exert a negative influence on the long-term incentive pay of mining firms. Substantial shareholders, in general, have a significant negative influence on the pay levels of both mining and non-mining firms. An exception to this is a lack of significant influence on the short-term incentive pay of non-mining firms. Broadly, these results provide a mixed evidence on the role of ‘managerial power’ in influencing compensation levels and structure in both mining and non-mining firms in Australia. We also estimate pay-performance sensitivities and pay-performance elasticities using panel random effects models. Consistent with

∆(Payit ) = ∆MCapit′β + (αi + εit ) Where, Payit alternatively includes fixed pay, short-term incentive pay, long-term incentive pay and total pay, MCap is market capitalization, αi are firm-specific effects that are assumed to be uncorrelated with the regressors and εit is an idiosyncratic error. We also estimate payperformance elasticities by regressing changes in logarithmic transformed pay on changes in logarithmic transformed market capitalization. Table 4 presents the pay-performance elasticities for both mining and non-mining firms. For a $1000 change in market capitalization for mining firms, CEO pay changes by 31.3 cents while it changes by 18.2 cents for non-mining firms. These pay-performance-sensitivities (PPS) are consistently higher for mining firms compared to non-mining firms for all types of pay. Mining firms, in particular, show high PPS in the case of long-term incentive pay. Overall, PPS is low for Australian firms compared to the levels of PPS shown in earlier studies for the US. Further, mining firms have exhibited higher PPS compared to nonmining firms and their PPS is predominantly in the form of long-term incentive. Conversely, for non-mining firms changes in performance pay is largely in the form of short-term incentive pay. Panel B in Table 4 captures pay-performance elasticities. Nonmining firms have a higher degree of pay-performance elasticity both in terms of fixed pay and long-term incentive pay, while mining firms

216

Economic Modelling 64 (2017) 211–220

S.R. Yarram, J. Rice

Table 4 Pay-performance sensitivities and elasticities. Δ(FixedSalary) Mining Model 1 Panel A: Pay-performance sensitivities Δ(Mcap) 0.031** (2.40) Firm-years 765 R2 Overall 0.007 Δln(FixedSalary) Mining Model 1 Panel B: Pay-performance elasticities Δln(Mcap) 0.081*** (3.03) Firm-years 745 R2 Overall 0.012

Δ(STI)

Δ(LTI)

Δ(TotalPay)

Non-mining Model 2

Mining Model 3

Non-mining Model 4

Mining Model 5

Non-mining Model 6

Mining Model 7

Non-mining Model 8

0.029*** (2.98) 2225 0.004

0.088*** (7.54) 2225 0.025

Non-mining Model 6

0.313*** (7.00) 765 0.060 Δln(TotalPay) Mining Model 7

0.182*** (7.76) 2225 0.026

Non-mining Model 4

0.187*** (5.22) 765 0.035 Δln(LTI) Mining Model 5

0.065*** (4.79) 2225 0.010

Non-mining Model 2

0.095*** (5.66) 765 0.040 Δln(STI) Mining Model 3

Non-mining Model 8

0.110*** (5.25) 2173 0.013

0.343*** (3.73) 197 0.067

0.255*** (4.90) 1058 0.022

0.656** (2.37) 765 0.007

1.173*** (6.36) 2225 0.018

0.230*** (7.03) 765 0.061

0.220*** (9.95) 2225 0.043

This table reports results relating to pay-performance sensitivities and elasticities for mining and non-mining firms the period 2005 to 2013. Changes in pay as a function of changes in market capitalization is captured in pay-performance sensitivities. Similarly, changes in natural logarithm of each pay variables as a function of changes in natural logarithm of market capitalization is captured in terms of pay-performance elasticities. Variable descriptions are provided in Table 1. Note: t-statistics are provided in parentheses. *** and ** denote significance at 0.01 and 0.05 respectively.

pay-performance sensitivities of mining and non-mining firms using panel random effects models. Consistent with (Core et al., 1999; Schultz et al., 2013), we estimate the following random effects panel data model

have a higher degree of pay-performance elasticity in terms of shortterm incentive pay. Table 5 captures the influence of the GFC on PPS of both mining and non-mining firms. For the purpose of this analysis, the periods of 2005 to 2007, 2008 to 2010 and 2011 to 2013 are considered as pre, during and post GFC periods respectively. The results show some interesting variations across the 3 different sub-periods. While the mining firms exhibited relatively higher levels of PPS in the pre and during GFC sub-periods, non-mining firms registered relatively higher levels of PPS in the post GFC period. During the pre-GFC period, nonmining firms have shown higher sensitivities in terms of short-term incentive pay. On the other hand, mining firms have shown higher degree of sensitivity in long-term incentive pay. During the GFC, both the mining and non-mining firms have shown higher PPS both in terms of short-term incentives as well as long-term incentives, but the mining firms in general have higher PPS. During the post-GFC period, nonmining firms have shown higher PPS both in terms of fixed pay as well as short-term incentive pay. During this period mining firms have shown higher PPS in long-term incentive pay. We also examine the role of governance and economics factors on

∆(Payit ) = Xit′β + (αi + εit ) Where, Payit alternatively includes log transformed fixed pay, shortterm incentive pay, long-term incentive pay and total pay, Xit′ is a set of log transformed economic, ownership and governance variables in addition to changes in market capitalization, αi are firm-specific effects that are assumed to be uncorrelated with the regressors and εit is an idiosyncratic error. Table 6 presents the results relating to influence of governance variables on PPS. Board size and CEO duality show no significant influence on the PPS related to incentive pay while they have a negative influence on the PPS relating to fixed pay. This lack of association between incentive PPS and board size and CEO duality suggests the complexity of the compensation design process. Mining firms with larger boards exert a negative influence on long-term incentives employed, while CEOs that perform the dual roles exert a positive

Table 5 Pay-performance sensitivities and GFC. Δ(FixedSalary)

Panel A: Pre-GFC Δ(Mcap) Firm-years R2 Overall Panel B: GFC Δ(Mcap) Firm-years R2 Overall Panel C: Post-GFC Δ(Mcap) Firm-years R2 Overall

Δ(STI)

Δ(LTI)

Δ(TotalPay)

Mining Model 1

Non-mining Model 2

Mining Model 3

Non-mining Model 4

Mining Model 5

Non-mining Model 6

Mining Model 7

Non-mining Model 8

0.091*** (3.86) 249 0.057

0.053*** (3.56) 782 0.016

0.039 (1.30) 249 0.007

0.089*** (5.60) 782 0.039

0.219*** (2.97) 249 0.034

0.040* (1.68) 782 0.004

0.349*** (3.97) 249 0.060

0.181*** (5.12) 782 0.032

0.016 (0.72) 274 0.002

−0.008 (−0.56) 751 0.000

0.123*** (5.70) 274 0.107

0.075*** (3.96) 751 0.021

0.194*** (3.55) 274 0.044

0.070*** (3.31) 751 0.014

0.333*** (4.83) 274 0.079

0.137*** (3.71) 751 0.018

0.018 (0.81) 242 0.003

0.063** (2.44) 692 0.009

0.072* (1.82) 242 0.014

0.139*** (4.37) 692 0.027

0.124* (1.81) 242 0.013

0.067** (2.06) 692 0.006

0.215** (2.46) 242 0.025

0.269*** (4.31) 692 0.026

This table reports results relating to pay-performance sensitivities for mining and non-mining firms. Results are separately reported for the pre-GFC, GFC and post-GFC periods. Changes in pay as a function of changes in market capitalization is captured in pay-performance sensitivities. Variable descriptions are provided in Table 1. Note: t-statistics are provided in parentheses. ***, ** and * denote significance at 0.01, 0.05 and 0.10 respectively.

217

Economic Modelling 64 (2017) 211–220

S.R. Yarram, J. Rice

Table 6 Pay-performance sensitivities and governance. Δ(FixedSalary)

Δ(Mcap) BrdSize BrdInd CeoDuality CeoTenure RemComDummy CEORemCom %SubSH Δ(Mcap) * Larger Board Δ(Mcap) * BrdInd Δ(Mcap) * CeoDuality Δ(Mcap) * RemComDummy Δ(Mcap) * CeoRemCom Δ(Mcap) * %SubSH Firm-years R2 Overall

Δ(STI)

Δ(LTI)

Δ(TotalPay)

Mining Model 1

Non-mining Model 2

Mining Model 3

Non-mining Model 4

Mining Model 5

Non-mining Model 6

Mining Model 7

Non-mining Model 8

0.006 (0.16) −14,134.582*** (−3.48) 23,2853.622*** (3.73) −20,328.570 (−0.75) 247.402 (0.17) 1239.710 (0.08) 35,083.718 (1.43) 36,985.864 (1.00) 0.005 (0.16) 0.040 (1.00) −0.063 (−1.52) 0.034 (1.04) 0.026 (0.74) −0.040 (−0.88) 765 0.048

0.034 (0.95) −7767.946** (−2.30) 15,7271.931*** (2.82) −6611.225 (−0.27) 1841.349* (1.80) 8736.203 (0.59) 2849.467 (0.17) 3208.944 (0.11) −0.014 (−0.51) 0.029 (1.09) −0.009 (−0.18) 0.019 (0.73) −0.014 (−0.58) −0.030 (−1.35) 2225 0.013

0.061 (1.28) −4149.746 (−0.79) 36,340.490 (0.45) 7680.120 (0.22) −1219.027 (−0.66) 17,175.656 (0.84) 14,616.693 (0.45) −42,648.528 (−0.89) 0.050 (1.15) 0.074 (1.39) −0.079 (−1.46) 0.011 (0.25) 0.017 (0.37) −0.016 (−0.26) 765 0.056

0.073* (1.72) 530.714 (0.13) 20,749.819 (0.31) 20,752.336 (0.70) −2157.221* (−1.76) 5908.364 (0.33) −5898.589 (−0.30) −6509.419 (−0.19) −0.031 (−0.93) −0.017 (−0.52) −0.049 (−0.88) 0.022 (0.72) 0.034 (1.15) 0.076*** (2.88) 2225 0.034

0.445*** (4.37) −15,429.741 (−1.38) 95,773.873 (0.55) 49,704.737 (0.67) −4527.183 (−1.15) 88,163.524** (2.04) −40,967.673 (−0.60) −10,2247.728 (−1.00) −0.160* (−1.75) 0.014 (0.12) −0.104 (−0.91) −0.239*** (−2.65) 0.088 (0.89) 0.149 (1.17) 765 0.061

0.151*** (3.04) 612.048 (0.13) 14,6830.323* (1.88) −32,361.725 (−0.93) 485.441 (0.34) −14,266.848 (−0.69) −10,370.872 (−0.45) −7200.189 (−0.18) −0.041 (−1.06) 0.043 (1.15) 0.150** (2.30) −0.055 (−1.52) −0.068** (−1.97) −0.027 (−0.88) 2225 0.024

0.514*** (4.04) −33,232.415** (−2.37) 35,9270.461* (1.66) 38,141.808 (0.41) −5570.734 (−1.14) 10,6514.077** (1.97) 7863.795 (0.09) −10,7574.506 (−0.84) −0.107 (−0.93) 0.128 (0.91) −0.248* (−1.72) −0.196* (−1.74) 0.132 (1.07) 0.094 (0.59) 765 0.085

0.258*** (3.01) −6625.183 (−0.81) 32,4852.072** (2.40) −18,220.613 (−0.30) 169.568 (0.07) 377.719 (0.01) −13,419.994 (−0.34) −10,500.664 (−0.16) −0.086 (−1.28) 0.055 (0.86) 0.092 (0.82) −0.014 (−0.22) −0.049 (−0.81) 0.019 (0.36) 2225 0.031

This table reports results relating to pay-performance sensitivities for mining and non-mining firms the period 2005 to 2013. Changes in pay as a function of changes in market capitalization is captured in pay-performance sensitivities. Interactions between market capitalization and each governance variable are also incorporated in this analysis. Variable descriptions are provided in Table 1. Note: t-statistics are provided in parentheses. ***, ** and * denote significance at 0.01, 0.05 and 0.10 respectively.

influence on the long-term incentives of non-mining firms. When CEOs are on the remuneration committees, they exert a negative influence on the long-term incentive pay of non-mining firms. Broadly, these results do not provide a strong evidence of the managerial power theory on the incentive pay setting of Australian mining as well as non-mining firms. To check the robustness of our findings relating to pay-performance sensitivities of mining and non-mining firms, we estimate joint models for both mining and matched non-mining firm sub-samples with an interaction variable. Similar to treatment effects models, the interaction term in these models captures the significance of a firm's status as a mining or non-mining firm. Using propensity score matching, we identify two sets of non-mining firms that are separately matched on size and on stock market returns. Table 7 presents results relating to pay-performance sensitivities of our mining and matched non-mining firm subsamples for the study period. Panel A shows that mining firms have statistically significant greater pay-performance sensitivities with regard to long-term incentive pay compared to size-matched non-mining firms. Pay-performance sensitivities of mining firms are statistically significantly different from that of the matched non-mining firms. Furthermore, these sensitivities are higher for mining firms as indicated by the positive sign of the interaction term. Pay-performance sensitivities relating to short-term incentives are lower for mining firms compared to non-mining firms, however the differences in these pay-performance sensitivities are not statistically different for mining and matched non-mining firms. We come to similar conclusions when using an alternative matching of firms by stock market returns as shown in Panel B. Table 8 provides results of difference-in-differences analysis of mining and propensity score matched non-mining firms for the pre-

Table 7 Pay-performance sensitivities for matched mining and non-mining firms. Δ(FixedSalary) Model 1s

Δ(STI) Model 2s

Panel A: Size-matched mining and non-mining firms Δ(Mcap) 0.017 0.125*** (0.95) (5.42) Δ(Mcap) * Matched0.011 −0.030 Firm (0.47) (−1.04) Firm-years 1536 1536 R2 Overall 0.005 0.039 Δ(FixedSalary) Model 1r

Δ(STI) Model 2r

Panel B: Return-matched mining and non-mining firms Δ(Mcap) 0.057*** 0.105*** (4.36) (6.45) Δ(Mcap) * Matched−0.029 −0.007 Firm (−1.47) (−0.26) Firm-years 1523 1523 2 R Overall 0.017 0.042

Δ(LTI) Model 3s

Δ(TotalPay) Model 4s

0.092** (2.40) 0.096**

0.233*** (4.43) 0.079

(1.98) 1536 0.029

(1.19) 1536 0.050

Δ(LTI) Model 3r

Δ(TotalPay) Model 4r

0.104*** (3.87) 0.081*

0.269*** (7.11) 0.046

(1.94) 1523 0.031

(0.79) 1523 0.061

This table reports results relating to pay-performance sensitivities mining and matched non-mining firms the period 2005 to 2013. Changes in pay as a function of changes in market capitalization is captured in pay-performance sensitivities. Variable descriptions are provided in Table 1. Note: t-statistics are provided in parentheses. ***, ** and * denote significance at 0.01, 0.05 and 0.10 respectively.

218

Economic Modelling 64 (2017) 211–220

S.R. Yarram, J. Rice

mining firms show higher sensitivity of short-term incentive pay. Payperformance sensitivities have undergone significant changes for mining and non-mining firms during the study period. While mining firms have shown reduced pay-performance sensitivity in the post-GFC period compared to the pre-GFC period, non-mining firms have shown increased pay-performance sensitivities during the post-GFC period. Given the higher degree of volatility in earnings due to the strong cyclicality in the prices for minerals, CEOs that are recently appointed to mining firms tend to aggressively negotiate incentive-based pay as a means of securing a real option against any cyclical upside in the firms’ earnings. This situation in turn increases the risk tolerance and indeed risk preference of CEOs appointed to mining firms as they seek to enhance the potential for earnings upside by making investment decisions that expose their firms to higher risks and higher potential returns. Overall, the results provide support for optimal contracting theory for both mining and non-mining firms in Australia. Governance factors have some degree of influence on the payperformance sensitivities of mining and non-mining firms during the study period, however these influences are not consistent with the predictions of managerial power approach. Large boards in mining firms reduce the pay-performance sensitivity, whereas CEOs that perform the dual roles improve the pay-performance sensitivity of long-term incentives of non-mining firms. When CEOs are on the remuneration committees, they reduce the pay-performance sensitivity of long-term incentive pay of non-mining firms. Overall, the results do not provide a strong evidence in favour of managerial power approach in Australia for both mining and non-mining firms. However, an important limitation of our study is that we do not model possible differences in the executive traits and investor characteristics of mining and non-mining firms. We essentially consider that pay levels and sensitivities are influenced by economic and governance characteristics of firms and that changes in pay may not determine the changes in market capitalization. We confirm this by testing for endogeneity using WuHausman F test and Durbin-Wu-Hausman chi-square test. However, it is true that CEO traits and investor characteristics may differ significantly between mining and non-mining firms and this remains a limitation of this study and perhaps these differences are part of an unobserved heterogeneity. Another limitation that potentially limits the generalisability of our results is that the study period of 2005 to 2013 may not capture a complete business cycle. Overall, the findings of this study have implications for theory and practice in relation to executive pay, corporate finance and resource economics. Economic and business factors play an important role in determining the risk and size of cash flows for mining and non-mining firms and these factors need to be considered carefully when designing compensation policies. Similarly, policy makers tasked with designing and overseeing corporate governance frameworks need to consider the influence exerted by members of boards and board sub-committees in improving corporate performance and aligning the pay and performance incentives for senior managers. Further, major economic upheavals such as the GFC may have significant impact on the resource and non-resource sectors and their financial policies. Future studies may consider the role of executive traits and investor characteristics in determining pay levels and pay-performance sensitivities of mining and non-mining firms. Future studies may consider a much longer study period that includes different phases of a complete economic cycle.

Table 8 Difference-in-differences regressions.

Treatment GFC Treatment * GFC Firm-years R2 Overall

D(FixedSalary) DiD-Model 1

D(STI) DiD-Model 2

D(LTI) DiD-Model 3

D(TotalPay) DiD-Model 4

0.082** (2.47) 0.085*** (2.93) −0.145***

0.034 (0.73) 0.186*** (4.44) −0.137*

0.245*** (3.63) 0.054 (0.88) −0.160

0.376*** (3.77) 0.325*** (3.70) −0.456***

(−2.86) 1008 0.018

(−1.88) 1008 0.028

(−1.53) 1008 0.021

(−2.97) 1008 0.040

This table reports results of difference-in-differences regressions. Treatment captures the interaction between changes in market capitalization and the dummy variable that captures mining and propensity score matched non- mining firms. GFC refers to the interaction between changes in market capitalization and the dummy variable that equals 0 if the financial year is prior to 2008 and 1 if the financial year is after 2010. Variable descriptions are provided in Table 1. Note: t-statistics are provided in parentheses. ***, ** and * denote significance at 0.01, 0.05 and 0.10 respectively.

GFC and the post-GFC period. Non-mining firms are matched using propensity score that considered size, return on assets, growth, board size, board independence, CEO duality and year. Treatment variable captures the interaction between changes in market capitalization and a dummy variable that captures mining and propensity score matched non-mining firms. GFC refers to the interaction between changes in market capitalization and the dummy variable that considers the preGFC period (i.e., financial year is prior to 2008) and the post-GFC period (financial year is after 2010). Results from diff-in-diff analysis show that pay-performance sensitivities of mining firms are significantly higher than pay-performance sensitivities of propensity score matched non-mining firms in terms of long-term incentive pay as well as fixed pay. Mining firms seek to lock in senior executives with greater long-term contingent remuneration. The GFC has had a significant impact on the incentive structures of both mining and non-mining firms. Pay-performance sensitivities relating to short-term incentives as well as fixed pay have significantly increased in the post-GFC period compared to the pre-GFC period. Pay-performance sensitivities of mining firms have declined in the post-GFC period compared to the matched non-mining firms and that the declines are significant in the fixed pay and short-term incentives perhaps caused by lower profits. Overall, the GFC has an adverse impact on the incentives structures of mining firms.

4. Summary and conclusion This study considers the levels and sensitivities of executive compensation in mining and non-mining firms that were listed on the Australian All Ordinaries Index for the period 2005 to 2013. Mining firms on average pay their CEOs approximately $1 million dollar a year as total salary compared to an average $1.5 million in non-mining firms. While two-thirds or more of the total compensation is fixed in nature, mining firms pay a relatively higher proportion of long-term incentives whereas non-mining firms pay a relatively higher level of short-term incentive pay. We find that the economic variables identified in previous literature have significant influences on the pay levels of Australian mining and non-mining firms. Pay-performance sensitivities, in general, are low for Australian firms compared to the US, as reported in earlier studies. Pay-performance-sensitivities are consistently higher for mining firms compared to non-mining firms. Mining firms show higher pay-performance sensitivity in terms of long-term incentive pay, while non-

Acknowledgement We thank the editor, referees, Robert Faff and Chandrasekhar Krishnamurti for their constructive feedback on this paper. Usual disclaimers apply.

219

Economic Modelling 64 (2017) 211–220

S.R. Yarram, J. Rice

Jayasuriya, K., Cannon, D., 2015. State transformation and resource politics: australia and the regional political economy. Pac. Rev. 28, 391–410. Jensen, M.C., Murphy, K.J., 1990. Performance pay and top-management incentives. J. Polit. Econ. 98, 225–264. Kim, H.-S., 2013. Executive bonus compensation when abnormal earnings and the state of the economy are correlated. Econ. Model. 32, 58–65. Lee, J., 2009. Executive performance-based remuneration, performance change and board structures. Int. J. Account. 44, 138–162. Matolcsy, Z., Wright, A., 2007. Australian CEO compensation: the descriptive evidence. Aust. Account. Rev. 17, 47–59. Matolcsy, Z., Wright, A., 2011. CEO compensation structure and firm performance. Account. Financ. 51, 745–763. Matolcsy, Z.P., 2000. Executive cash compensation and corporate performance during different economic cycles. Contemp. Account. Res. 17, 671–692. Merhebi, R., Pattenden, K., Swan, P.L., Zhou, X., 2006. Australian chief executive officer remuneration: pay and performance. Account. Financ. 46, 481–497. Murphy, K.J., 1985. Corporate performance and managerial remuneration: an empirical analysis. J. Account. Econ. 7, 11–42. Murphy, K.J., 1997. Executive compensation and the modern industrial revolution. Int. J. Ind. Organ. 15, 417–425. Murphy, K.J., 2013. Chapter 4 – Executive compensation: where we are, and how we got there. In: Handbook of the Economics of Finance 2. Elsevier, Amsterdam, 211–356, (Part A). Nagar, V., Nanda, D., Wysocki, P., 2003. Discretionary disclosure and stock-based incentives. J. Account. Econ. 34, 283–309. Rankin, M., 2010. Structure and level of remuneration across the top executive team. Aust. Account. Rev. 20, 241–255. Roarty M., 2010. The Australian Resources Sector: Its Contribution to the Nation, and a Brief Review of Issues and Impacts. Parliament of Australia, Parliamentary Library. Rustam, S., Rashid, K., Zaman, K., 2013. The relationship between audit committees, compensation incentives and corporate audit fees in Pakistan. Econ. Model. 31, 697–716. Samis, M., Davis, G.A., Laughton, D., Poulin, R., 2005. Valuing uncertain asset cash flows when there are no options: a real options approach. Resour. Policy 30, 285–298. Schultz, E., Tian, G.Y., Twite, G., 2013. Corporate governance and the CEO pay– performance link: Australian evidence. Int. Rev. Financ. 13, 447–472. Walker, J., 2010. The use of performance-based remuneration: high versus low-growth firms. Aust. Account. Rev. 20, 256–264. Wang, J., Sheng, J., Yang, J., 2013. Optimism bias and incentive contracts in portfolio delegation. Econ. Model. 33, 493–499. Yeoh, P., 2015. Corporate governance in the UK: a time for reflection. Bus. Law Rev. 36, 130–135. Yermack, D., 1995. Do corporations award CEO stock options effectively? J. Financ. Econ. 39, 237–269. Yermack, D., 2006. Flights of fancy: corporate jets, CEO perquisites, and inferior shareholder returns. J. Financ. Econ. 80, 211–242. Zajac, E.J., Westphal, J.D., 1994. The costs and benefits of managerial incentives and monitoring in large US corporations: when is more not better? Strateg. Manag. J. 15, 121–142. Zhang, K., 2014. Essays on Mining Economics and System Modeling. The Pennsylvania State University, University Park, PA 16802, United States, 2014. Zhang, K., Kleit, A.N., 2016. Mining rate optimization considering the stockpiling: a theoretical economics and real option model. Resour. Policy 47, 87–94. Zhang, K., Nieto, A., Kleit, A.N., 2015. The real option value of mining operations using mean-reverting commodity prices. Miner. Econ. 28, 11–22.

References Armstrong, C.S., Ittner, C.D., Larcker, D.F., 2012. Corporate governance, compensation consultants, and CEO pay levels. Rev. Account. Stud. 17, 322–351. Bahaji, H., Casta, J.-F., 2016. Employee stock option-implied risk attitude under RankDependent Expected Utility. Econ. Model. 52, 144–154. Beatty R., Zajac E., 1990. Top management incentives, monitoring, and risk-bearing: a study of executive compensation, ownership, and board structure in initial public offerings. Academy of Management Proceedings, Academy of Management Proceedings. Academy of Management, vol. 1990, pp. 7–11. Bebchuk, L.A., Cremers, K.M., Peyer, U.C., 2011. The CEO pay slice. J. Financ. Econ. 102, 199–221. Bebchuk, L.A., Fried, J.M., 2003. Executive compensation as an agency problem. J. Econ. Perspect. 17, 71–92. Bizjak, J.M., Brickley, J.A., Coles, J.L., 1993. Stock-based incentive compensation and investment behavior. J. Account. Econ. 16, 349–372. Cairns, R.D., Davis, G.A., 2001. Adelman's rule and the petroleum firm. Energy J., 31–54. Chalmers, K., Koh, P.-S., Stapledon, G., 2006. The determinants of CEO compensation: rent extraction or labour demand? Br. Account. Rev. 38, 259–275. Chen, M.-Y., 2010. Managerial pay adjustments: decomposition and impact on firm productive efficiency. Econ. Model. 27, 196–207. Chiara, M., Marco, F., Antonio, R., 2016. Owners' preferences for CEOs characteristics: did the world change after the global financial crisis? Corp. Gov.: Int. J. Bus. Soc. 16, 116–134. Citci, S.H., Inci, E., 2016. The masquerade ball of the CEOs and the mask of excessive risk. Econ. Model.. Coles, J.L., Daniel, N.D., Naveen, L., 2006. Managerial incentives and risk-taking. J. Financ. Econ. 79, 431–468. Corden, W.M., 2012. Dutch disease in Australia: policy options for a three‐speed economy. Aust. Econ. Rev. 45, 290–304. Core, J.E., Holthausen, R.W., Larcker, D.F., 1999. Corporate governance, chief executive officer compensation, and firm performance. J. Financ. Econ. 51, 371–406. Core, J.E., Larcker, D.F., 2002. Performance consequences of mandatory increases in executive stock ownership. J. Financ. Econ. 64, 317–340. David, P., Kochhar, R., Levitas, E., 1998. The effect of institutional investors on the level and mix of CEO compensation. Acad. Manag. J. 41, 200–208. de La Bruslerie, H., 2016. Does debt curb controlling shareholders' private benefits? Modelling in a contingent claim framework. Econ. Model. 58, 263–282. Doucouliagos, H., Haman, J., Askary, S., 2007. Directors' remuneration and performance in Australian banking. Corp. Gov. 15, 1363–1383. Gabaix, X., Landier, A., 2008. Why has CEO pay increased so much? Q. J. Econ. 123, 49–100. Gomez-Mejia, L.R., Tosi, H., Hinkin, T., 1987. Managerial control, performance, and executive compensation. Acad. Manag. J. 30, 51–70. Guay, W.R., 1999. The sensitivity of CEO wealth to equity risk: an analysis of the magnitude and determinants. J. Financ. Econ. 53, 43–71. Hosseinzadeh, A., Smyth, R., Valadkhani, A., Le, V., 2016. Analyzing the efficiency performance of major Australian mining companies using bootstrap data envelopment analysis. Econ. Model. 57, 26–35. Huang, X., 2015. The impact of strategic market entry considerations on the financial performance of Chinese investment in the Australian Mining Industry. Thunderbird Int. Bus. Rev. 57, 323–339. Izan, H.Y., Sidhu, B., Taylor, S., 1998. Does CEO pay reflect performance? some Australian evidence. Corp. Gov.: Int. Rev. 6, 39–47.

220