CEOs Inside Debt and Firm Innovation. Abstract. In the environment of high technology industries, innovation is one of the most

CEOs Inside Debt and Firm Innovation Abstract In the environment of high technology industries, innovation is one of the most important element to help firm stay competitive and to promote core value. We get information of patent war and lawsuits very often, and it is always linked to massive amount of money. Therefore we try to study how the firms will choose their innovation strategies and what are the consequences. Our results show that firms with high inside debt would invest less innovation activities. When we further decompose innovation activities into in-house patenting activities and patent purchasing activities, we find that firms with high inside debt would invest less in-house patenting activities. Finally, when we separately look at the effects of inside debt and innovation strategies on firm performance, we find that inside debt negatively affects Tobin s Q and ROE, and in-house patenting activities positively affect Tobin s Q and ROE. These findings indicate that in-house patenting activities twist the side effect of inside debt. Keywords: Inside Debt Innovation CEOs Compensation In-house Patenting Patent Purchasing 1

CEOs Inside Debt and Firm Innovation Introduction In this paper we examine whether CEOs inside debt will affect firm innovation. Consider the environment of high-technology firms in recent years, the patent fighting war is the main issue of firm operation especially smart phone companies. To investigate the factors that influence firms innovation strategies is the important contribution for practitioners and literatures. Previous literatures have documented that managerial compensation is the main factor for inducing innovation activities (i.e. Choy, Lin, and Officer, 2014; Cassell, Huang, Sanchez, and Stuart, 2012; Cole, Daniel, and Naveen, 2006; Ryan and Wiggins, 2002). However, when discussing this relationship, previous literatures only define in-house patenting activities as innovation activity. But, they ignore that patent purchasing activity is one of the reasons for patent fighting. Cole, Daniel, and Naveen (2006), Choy, Lin, and Officer (2014) indicate that managers risk preference is the main factor for the development of firm innovation. Therefore, the compensation scheme is the way to direct managerial risk preference. Edmans and Liu (2011) indicate that the financial crisis shows that the substantial debt holders loss from agency problem between debt holders and managers are not completely solved. In other words, debt-based compensation may play an important role on alleviate the agency problem between insiders and debt holders. Edmans and Liu (2011) also indicate that the compensation scheme typically include both equity-based and debt-based compensation tools. However, these two types of compensation tools have different impact on managers desire for enhancing firm value. The goal of this paper is to get further evidence and contribute literature by considering the whole picture for the relationship between compensation and innovation activities. 2

Equity-based compensation motivates managers risk-taking behavior. We believe managers avoid investing in innovation activities is because of the uncertainty of the future cash flow, even innovation activities contribute the competition advantage of firms (Brown and Petersen, 2011). Jensen and Meckling (1976) argue that managers risk-averse behavior induce them to skip the risky but value-enhancing investment projects. To change managers risk-averse behavior, granting managerial compensation scheme to managers could be the possible solution. (Cole, Daniel, and Naveen, 2006). Ryan and Wiggins (2002) indicate that investing in innovation activities brings long-term horizon and high degree of uncertainty. If managers are compensated on short-term accounting measures (Jensen, 1986), they will avoid investing in long-horizon innovation projects (i.e. R&D projects) for the downside risk. Equity-based compensation induces manages risk taking behavior, at meantime, the agency problem between managers and debt holders arise. Jensen and Meckling (1976) argue that when managers benefit shareholders, they also expense at the benefit of debt holders. Debt holders are like to see that managers behave more conservative on managers investment and financing strategies in order to secure debt holders benefits. Jensen and Meckling (1976) indicate that inside debt holdings induces managers to take less-risky investment and financing projects and to mitigate the agency problem between managers and debt holders. Managers are more inclined to make conservative investment projects when they hold excess inside debt. Cassell, Huang, Sanchez, and Stuart (2012) find that when firms with high proportion of managerial inside debt have lower research and development expenditure (R&D) and higher diversification. They document that managers reduce research and development expenditure and diversify firm segments for enhancing payoff in bankruptcy and liquidation value, which is much align with debt holders interests. 3

When previous literatures examine the relationship between the managerial compensation and innovation activities, they focus on in-house patenting activities only (i.e. Choy, Lin, and Officer, 2014; Cassell, Huang, Sanchez, and Stuart, 2012; Cole, Daniel, and Naveen, 2006; Sheikh, 2002; Holthausen, Larcker, and Sloan, 1995). However, in-house patenting activity is not the only way for innovation. Patent purchasing activity is the other way. Narula (2001) indicate that comparing with non-in-house patenting activities, in-house patenting activities help firms to take background competences and marginal competences. O Regan and Kling (2011) also argue that in-house patenting activities increase firms sale, market share, and profitability. In contrast to the in-house patenting activities, non-in-house patenting activities have several disadvantages. Narula (2001) illustrates that firms have to pay negotiating and enforcing costs for negotiating contract. Firms also have to bear the disability of accumulating know-how and take the risk for leaking know-how to potential competitors by the collaborators. Furthermore, firms have highly possibility of loss of key assets, because appropriateness of innovation varies widely by country or by industry. However, when firms invest in-house patenting activities, they have to consider high adjustment costs and the uncertainty of future cash flow. Brown and Petersen (2011) document that firms have to pay high cost on hiring skilled technology works and training them. Once firms terminate in-house patenting activities, they have to pay high cost on lying off or transfer these workers. Therefore, when firms invest in-house patenting activities, they have to consistently invest in-house patenting activities for saving the substantial adjustment costs. Brown, Fazzari, and Petersen (2009) indicate that equity financing is the only way to finance in-house patenting activities when firms exhausted their internal fund. Although inhouse patenting activities can bring the competition advantage in the long-term, managers have to consider the uncertainty of future cash flow by in-house patenting activities. 4

According to the previous literatures, we find that firms investing the in-house patenting activities come up with the competition advantages, but the uncertainty of future cash flow as well. Therefore, managers treat the investments of in-house patenting activities as the valuable, but risky projects. Since managers holding high inside debt will align with debt holders interests and prefer investing conservative projects (Eisdorfer, Giaccotto, and White, 2013; Choy, Lin, and Officer, 2014), investing in-house patenting activities would not be the main innovation activities for managers. In contrast, when investing patent purchasing activities, managers have transferred the uncertain of future cash flow to the developers. Therefore, managers treat patent purchasing activity investment as physical assets investment in terms of relatively conservative investment. Whereby, managers holding high portion of inside debt are more inclined to invest patent purchasing activities than in-house patenting activities. To examine our hypothesis, we collect in-house patenting data from Compustat. Our patent purchasing data of S&P 500 firms are collected from LexisNexis Academic database. Both of these data are over 2006-2014. We start from year 2006, because the SEC s requires firms to expand executive compensation disclosure from year 2006. We also exclude the firms not in the industries where are the pharmaceutical products, chemicals, machinery, business services, computers, and electronic equipment. According to our observation, these industries are intellectual intensive industries. We abstract compensation variables from the Standard & Poor s Execucomp database. By following Cassell, Huang, Sanchez, and Stuart (2012), we use two measures for catching the effect of inside debt, including: the natural log of the ratio of the CEO s debt-to-equity ratio to the firm s debt-to-equity ratio; an indicator variable equals to one when CEO to firm debt/equity ratio greater than one. When we navigate patenting purchasing data from LexisNexis Academic database, we have key words "Firm Name AND Patent AND Acquisition" and research it within major world 5

publications. Our patenting purchasing information is including the announcement date of patent purchasing, the amount of patent purchasing, the number of patent purchasing, and the target firm information. We also obtain firm-specific information from Compustat from CRSP. To examine our research question, first, we examine whether inside debt affect innovation activities. When we measure innovation activities, we sum up the amounts of inhouse patenting activities and patent purchasing activities. Second, we further examine whether firms with relatively low portion of inside debt will invest in-house patenting activities more than patent purchasing. For examine this hypothesis, we decompose the amount of innovation activities into in-house patenting activities and patent purchasing activities. Finally, we examine whether the inside debt and innovation strategies simultaneously affect on firm performance. Our evidences support the hypothesis. We find that firms with high inside debt would invest less innovation activities. When we further decompose innovation activities into inhouse patenting activities and patent purchasing activities, we find that firms with high inside debt would invest less in-house patenting activities. Finally, when we look at the synergy of inside debt and innovation strategies on firm performance, we find that firms with high inside debt and in-house patent activities will have highest Tobin s Q and returns on equity (ROE). When we separately look at the effects of inside debt and innovation strategies on firm performance, we find that inside debt negatively affects Tobin s Q and ROE, and in-house patenting activities positively affect Tobin s Q and ROE. These findings indicate that inhouse patenting activities twist the side effect of inside debt. This paper contributes literatures and practitioners for several aspects. First, Edmans and Liu (2011) indicate that the financial crisis shows that the substantial bound holder losses 6

from agency problem between debt holders and managers are not completely solved. Inside debt may play an important role on alleviate the agency problem between insiders and debt holders. This paper tries to examine how the inside debt encourage managers to make innovation activities. Furthermore, we will examine whether aligning managers interests with debt holders is good strategy for firm innovation policies. Second, previous literatures examine the relationship between managerial compensation and innovation activities, they only consider in-house patenting activities as innovation activities. Now a day, patent purchasing activities are the main innovation activity, especially for the industries with intense competition, short product life-cycle, and demanding innovation, such as high technology industry and pharmaceutical industry. This paper not just only considers in-house patenting activities, but also considers patent purchasing activities. We believe that our findings can shed light on the debate on how managerial compensation scheme enhances or impedes long-term innovation activities. Third, we further look at the synergy of inside debt and innovation activities on firm performance. Our evidences can contribute the literature and practitioners to see how to develop compensation scheme and innovation strategies enhance firm value and long-term firm performance. The paper process is as follows. Section 2 develops the hypotheses. Section 3 describes the research design. Section 4 is the conclusion. Section 5 presents our reference. 7

Literature Review How Does Inside Debt Affect Innovation Strategies? Agency theory (Jensen and Meckling, 1976) states that managers may pursue their selfinterests and damage the shareholders wealth by conducing the second-best operation decisions. In order to alleviate this principle-agent interest conflict problem, shareholders may consider granting equity based compensation. The purpose of granting equity based compensation is to alleviate the agency problem between managers and shareholders, and to motivate managers to take risky but increasing the equity value projects. However, when shareholders alleviate the agency problem between managers and shareholders, at meantime, the benefits of alleviating agency problems between managers and shareholders are expensing at debt holders benefits. Jensen and Meckling (1976) indicate that granting the manager the mixed equity- and debt- based compensation might reduce the shareholder-debt holder agency problem caused by purely equity-based compensation. The debt-based compensation includes managerial pensions and deferred compensation (Liu, Mauer, and Zhang, 2014; Cassell, Huang, Sanchez, and Stuart, 2012). Edmans and Liu (2011) argue that managers holding high inside debt may increase their effort on increasing the payoff in bankruptcy or enhancing the liquidation value of firms. They also argue that if the amount of inside debt excesses the amount of equity based compensation in the whole managerial compensation, the inside debt can improve the managerial efforts on alleviating the agency cost of debt. Liu, Mauer, and Zhang (2014) document that inside debt is characterized as unfunded and unsecured. They state that Unfunded means that although pension and deferred compensation accounts accumulate and grow in accordance with the the stated asset allocation, the firm never actually allocates money to the pension or the deferred compensation plan of the employee. The firm is obligate only to make payments at 8

retirement CEOs and other top executives as beneficiaries of these plans are classified by bankruptcy courts as just another unsecured creditor of firms. Accordingly, the characteristics of inside debt, unfunded and unsecured, have managers be like outside creditors and force managers to align with the debt holders interests. In addition, risk-aversion theory of inside debt indicates that the benefits of inside debts are unsecured, managers will behave more conservative to secure the liquidation value of assets when firms file bankruptcy (Liu, Mauer, and Zhang, 2014). Sundaram and Yermack (2007) also indicate that high leverage firms need more inside debt to encourage managers to improve liquidity value of assets. In other words, risk-shifting considerations are the important issue for the high leverage firms. Therefore, inside debts treat manages as one of debt holders and induce managers behave conservatively on investment and financing decisions. Inside debt induces managers to make investment and financing decision conservatively (Jensen and Meckling, 1976; Sundaram and Yermack, 2007; Cassell, Huang, Sanchez, and Stuart, 2012; Choy, Lin, and Officer, 2014). Choy, Lin, Officer (2014) examines the impact of inside debt on firm risk. They find that after firms freezes inside debt, managers shift their investment from capital expenditures to research and development projects (R&D) and increase leverage level. They indicate that inside debt induces manages to behave risk-averse. Their findings also support Sundaram and Yermack s (2007) argument that when managerial pension value increases relatively to equity value, managers will behave risk averse. Cassell, Huang, Sanchez, and Stuart (2012) also argue that since characteristics of inside debt have managers face the similar default risk as outside debt holders face, managers hold high inside debt will act risk-averse behavior. Whereby, Cassell, Huang, Sanchez, and Stuart (2012) find a negative association between managerial inside debt holdings and the volatility of future 9

firms stock return, R&D expenses, and financial leverage. They also find that a positive relationship between managerial inside debt holdings and the size of diversification and assets liquidity. Cassell, Huang, Sanchez, and Stuart (2012) document that managers hold more inside debt will be more incline to make conservative investment and financing policies. Patent investment improves firms growth potential, but brings highly uncertainty of future return. Hsu (2009), and Brown, Fazzari, and Petersen (2009) indicate that patent investment is an important factor for the firms future grow. Hsu (2009) find that patent investment can increase future stock returns and premiums, since patenting activities are the main driving force for economic growth and fluctuations. Brown, Fazzari, and Petersen (2009) indicate that the uncertain and volatile return of patent investments limit firms to finance their patent investment from external capital market. One of considerations by external financing providers for patent investments is the uncertainty of patent investment. They indicate that even though equity financing is the main funding resource of patent investment, several disadvantages of equity financing still restrict firms to invest patenting activities. When firms issue new shares, they have to pay a lemon premium to shareholders due to information asymmetry. Shareholders averse selection problem downwards share price when firms announce to issue new shares. According to previous literatures, investing innovation activities come up with the uncertainty of future cash flow and increasing the information asymmetry between firm and investors. For the managers holding high amount of inside debt, they will behave conservatively on investment policies. We can set up the first hypothesis: H1: CEOs holding low inside debt are inclined to invest innovation activities. 10

How Does Inside Debt Affect In-House Patenting/Patent Purchasing Activities? When the previous literatures examine how managerial inside debt affects innovation activities, they only focus on the in-house patenting activities. However, not only in-house patenting is an innovation investment, but also patenting purchasing is the other important innovation investment. There are several differences between in-house patenting and patent purchasing. First, they have different competition advantage. For investing in-house patenting activities, O Regan and Kling (2011) argue that in-house patenting activities increase firms sale, market share, and profitability. Narula (2001) indicate that comparing with non-in-house patenting activities, in-house patenting activities help firms to take background competences and marginal competences. In contrast to in-house patenting activities, Narula (2001) illustrates the several disadvantages of patent purchasing activities. First, the uncertain of innovation process of non-in-house patenting increases costs in negotiating and enforcing contracts. Second, firms only get results of non-in-house patenting activities, but disable to accumulate their own know-how. Third, collaborators have highly possibility to leakage the know-how to the potential competitors. Fourth, firms have highly possibility of loss of key assets, because appropriateness of innovation varies widely by country or by industry. Accordingly, when we look at the advantage of these two innovation activities, the managers of firms will prefer investing in-house patenting activities more than patent purchasing activities for keeping their long-term competition advantage. Since high stock liquidity enhances managers investment efficiency (Fang, Noe, and Tice, 2009), we expect that high stock liquidity will positively affect in-house patenting activities. Second, they have different cost drivers. Brown and Petersen (2011) indicate that the adjustment costs from in-house patenting investment are quantitatively important and 11

substantially larger than those of physical investment. Firms need to maintain their in-house patenting investment to avoid the substantial adjustment costs. When firms invest their inhouse patenting activities, they have to pay cost on hiring skilled technology workers and training these workers. If firms terminate their in-house patenting activities, they have to pay adjustment costs, including the costs of transferring skilled technology workers or patenting facilities. Brown and Petersen (2011) argue that firms need a consistent financing resource to support in-house patenting activities for avoiding the potential adjustment cost. In contrast, patent purchasing brings minor adjustment cost. Narula (2001) indicates that there are several advantages of non-in-housing patenting activities. First, non-in-housing activities have less capital needed. Second, patent purchasing helps firms to transfer the risk of failed developing patents. Comparing with in-house patenting activities, firms can avoid adjustment costs when they consider patent purchasing. Therefore, when firms consider patent purchasing strategy, the consistency of investing for avoiding adjustment costs will be the minor consideration. Since investing patent purchasing activities can transfer the uncertainty of future cash flow to the developers, investing the patent purchasing activities will be treat as the physical assets investment activities. Therefore, we can expect that manages holding high inside debt would like to invest patent purchasing activities, instead of in-house patenting activities. We set up the second hypothesis as follow: H2A: CEOs holding low inside debt are inclined to invest in-house patenting activities. H2B: CEOs holding high inside debt are inclined to invest patent purchasing activities. 12

How Does Inside Debt and Innovation Strategies Affect the Value of Firm? Implementing in-house R&D or patent acquisition strategy has different implications to firm value. As O Regan and Kling (2011) indicate, firms made significant investments in inhouse innovations tend to outperform their competitors in sales revenue, market share, and profitability. Comparing with acquiring patents from other entities, firms implementing inhouse R&D have more patent background competences and better technological skills (Narula, 2001). On the other hand, obtaining patents from third parties relies on sellers willingness to provide patent-related support after acquisitions. Since technological skills are controlled by the external parties, patents acquiring firms probably would encounter difficulties to bring their workforce up to speed after purchasing patents. Hence, patent acquisition strategy brings high level of systematic risk to firms. As Paster and Veronesi (2009) argue, new technologies bring high level of uncertainty of future productivity. On the other hand, when firms develop technologies in-house, the uncertainty of the future productivity increases idiosyncratic risk, because small scale of production and a low probability of a large-scale adoption of the developed technology in the industry. Taking together, investing in in-house R&D increases idiosyncratic risk; while patent purchasing increases systematic risk. Applying the capital assets pricing model (CAPM) to the context of innovation strategy, an increasing in systematic risk leads to a higher beta, thus raising the required rate of return. As the required rate of return increases, the firm value goes down, because a higher discount factor has to be used to derive firm value. As our inference, when firms invest patent purchasing activities, it come our high systematic risk and high require of return. If that is the case, managers will face the low market value of firms. Therefore, we form the following hypothesis: 13

H3: Firms with low CEOs inside debt and high in-house patenting activities have highest market value among those firms in the other sets. Data and Methodology Data To examine our hypothesis, we use Standard & Poor (S&P) 500 firms over 2006-2014 as our sample firms. We start from year 2006, because the SEC s requires firms to expand executive compensation disclosure from year 2006. We also exclude the firms not in the industries where are the pharmaceutical products, chemicals, machinery, business services, computers, and electronic equipment. According to our observation, these industries are intellectual intensive ones. We abstract compensation variables from the Standard & Poor s Execucomp database. We subtract in-house patenting data from Compustat. For patent purchasing data, we hand collect data through LexisNexis Academic database. We use Standard & Poor s Execucomp database for data on CEO compensateon. We obtain firmspecific information from Compustat, industrial segment data from the Compustat Industrial Segment Tapes, and stock return information from CRSP. Methodology Innovation activities, in-house patenting and patent purchasing We use three measurements to capture firms investments in product innovation: (1) patent acquisition activities (PURCHASER), (2) in-house R&D activities (IN-HOUSE), and (3) innovation activities (INNOVATION). For patent purchasing data, we collect it from LexisNexis Academic over the period 2006-2012. When we hand collect patent purchasing data from LexisNexis Academic database, we have key words Firm Name AND Patent AND Acquisition and research it within major world publications. Our patenting purchasing 14

information is including the announcement date of patent purchasing, the amount of patent purchasing, the number of patent purchasing, and the target firm information. We measure patent acquisition activities (PURCHASER) by calculating the ratio of the amount of patent acquired to the amount of lagged total assets. To measure in-house R&D activities (IN- HOUSE), we calculate the ratio of R&D expenses (R&D, COMPUSTAT data#46) according to data in the COMPUSTAT to the amount of total lagged assets. Regarding the innovation activities (INNOVATION), we aggregate the in-house patenting activities (IN-HOUSE) and patent acquisition activities (PURCHASER). CEO s Inside Debt By following Cassell, Huang, Sanchez, and Stuart (2012), we use an indicator variable (INSIDE) equals to one when CEO to firm debt-to-equity ratio is greater than one. To calculate the CEO to firm debt-to-equity ratio, we have the ratio of sum of the present value of accumulated pension benefits and deferred compensation to sum of equity based compensation be scaled by the ratio of firm total debt to the market value of equity. Control Variable For examining hypothesis 1 and hypothesis 2, we follow Brown, Fassari, and Petersen (2009) and control factors that will affect managerial innovation strategies. MARKETBOOK is Market-to-Book ratio. SGWTH is Sales growth. CASHFLOW is equal to sum of cash and short-term investments divided by total assets. CASHFLOW is equal to sum of change in cash and short-term investments divided by total assets. STKISSUES is equal to net cash raised from stock issued divided by total assets. DBTISSUES is equal to net new long-term debt issued divided by total assets. Moreover, profitability (PROFITABILITY), and capital expenditure (CAPXR) may also affect investment strategies on tangible or intangible assets. Year is to control year fixed effect. 15

For examining our third hypothesis, we follow Billet and Mauer (2003), modify their model, and control factors that will affect firm performance. SIZE is the value of the natural logarithm of total assets. We use this variable to proxy size effect. LEVERAGE is the ratio of total debt in period t divided by the book value of the lagged total assets. Firms with high leverage would have growth constraint, due to the limitations of debt covenant. PROFITABILITY is the ratio of EBITDA to total assets. Firms with high profitability would present high earnings performance in the current year. CAPXR (capital expenditure) is the ratio of capital expenditure to the lagged total assets. Firms with high capital expenditure may have high growth potential. TANGIBILITY is the ratio of fixed assets to lagged total assets. We use this variable to proxy bankruptcy cost. CONSTRAINT is an indicator variable that takes the value of 1 if the firm didn t pay a cash dividend. We use this variable to proxy external financing constraint. Models Our first hypothesis examines whether inside debt (INSIDE) decrease innovation activities. We follow and modify the model from Brown, Fassari, and Petersen (2009) as follow: INNOVATION ' = F +β. INSIDE ' + β 4 CASHHOLDING ;, ' (1) + +β = CASHHOLDING ;, '>. + β? MARKETBOOK ;, ' + β E SGWTH ;, '>. + β G STKISSUES ;, ' + β I STKISSUES ;, '>. + β J DBTISSUES ;, ' + β K DBTISSUES ;, '>. + β.l CASHFLOW ;, ' + β.. CASHFLOW ;, '>. + β.4 PROFITABILITY ;, ' +β.= CAPXR ;, ' + YEAR + ε ;,' For measuring CEO s inside debt, we consider an indicator variable, that equals to one when CEO to firm debt-to-equity ratio is greater than one. We use the ratio of sum of the in-house patenting activities and patent purchasing activities to lagged assets (INNOVATION) as the dependent variables. 16

We further to decompose the innovation activities into two activities, including in-house patenting activates and patent purchasing activities. We set the second hypothesis and examines whether CEOs holding high inside debt (INSIDE) will decrease their investment in in-house patenting activities. We follow and modify the model from Brown, Fassari, and Petersen (2009) as follow: RNDR (PURCHASER) ' = F +β. INSIDE ' + β 4 PURCHASER (RNDR) ' (2) + β = CASHHOLDING ;, ' + β? CASHHOLDING ;, '>. + β E MARKETBOOK ;, ' + β G SGWTH ;, '>. + β I STKISSUES ;, ' + β J STKISSUES ;, '>. + β K DBTISSUES ;, ' + β.l DBTISSUES ;, '>. + β.. CASHFLOW ;, ' + β.4 CASHFLOW ;, '>. + β.= PROFITABILITY ;, ' +β.? CAPXR ;, ' + YEAR + ε ;,' Next, we try to examine how the synergy of innovation activities and CEOs inside debt affect on firm performance. We set Tobin s Q and 5-year average ROE for measuring the market value of firms and long-term accounting performance. PERFORMANCE (3) = F +β. INSIDE INNOVATION ' + β = INNOVATION ' + β? INSIDE ' +β E CASHHOLDING ;, ' + β G SIZE ;, ' + β I LEVERAGE ;, ' + β J PROFITABILITY ;, ' + β K CAPXR ;, ' + β.l TANGIBILITY ;, ' + β.. CONSTRAINT ;, ' + YEAR + ε ;,' As argued, firms invest in in-house patenting activities leading to the uncertainty of the future cash flow increases idiosyncratic risk, because of the small scale of production and a low probability of a large-scale adoption of the developed technology in the industry. Since technological skills are controlled by the external parties, patent purchasing firms are highly likely to encounter difficulties to efficiently tune their technological workforce after purchasing patents. Hence, patent purchasing strategy generates a high level of systematic risk to firms. Since Capital assets pricing model (CAPM) indicates that an increase in systematic risk leads to a higher beta, the higher required rate of return is asked. As the 17

required rate of return increases, the firm value decreases as well, because a higher discount factor has to be used to derive firm value. Empirical Results Descriptive Statistics The sample including Standard & Poor (S&P) 500 firms database from 2006 to 2014, where data is available to compute CEOs inside debt and compensation incentives and where accounting data is available on Compustat. Table 1 shows the descriptive statistic results. For INSIDE we set as a dummy variable and the mean for INSIDE is 0.476. We can tell from the mean that the amount of firm CEO s inside debt holding ratio is greater firm s debt/equity ratio is more than firm CEO s inside debt holding ratio is less than firm s debt/equity ratio. Next we look at the innovation activities. We find that mean values of in-house patenting activities (RNDR) and patent purchasing activities (PURCHASER) are 0.071 and 0.015, separately. We find that firms tend to make effort in in-house patenting innovation activities rather than patent purchasing. However, when we look at the firms with patent purchasing activities, table 2 shows that the mean value of patent purchasing activities (0.252) is relatively larger than in-house patenting activities (0.100). On average, in-house patenting activity is the major innovation activities. However, patent purchasing activities is a significant investment project, while firms consider to invest in it. [Insert table 1 here] [Insert table 2 here] Correlation Test 18

In Table 3 present the correlation of the main variables used in this study. The relation between RNDR and INSIDE come out negative and statistically significant. This result matches our expectation, because when debt base compensation outweighs equity base compensation. We believe the CEOs are more likely to be more conservative and less risk taking when it comes to innovation investment. [Insert table 3 here] Univariate Analysis In Table 4, we compare the empirical results of how inside debt will affect innovation activities, including in-house patenting and patent purchasing activities. In the high inside debt group, we find that innovation activities, patent purchasing activities, and in-house patenting activities are all relatively lower than those in low inside debt group. These findings support our argument CEO s inside debt induces CEO to do conservative investment strategies. We also find that firms with high inside debt have relatively lower Tobin s Q and long-term ROE than those with low inside debt. Accordingly, the results support our H1 that CEOs holding low inside debt will invest more innovation activities. Also, the results support our H2A that CEOs holding low inside debt will invest more in-house patenting activities than those holding high inside debt. [Insert table 4 here] Multivariate Analysis Effects of CEOs Inside Debt on Firm Innovation Strategies According to the results from the univariate analysis, we conclude that CEOs granted high inside debt would behave more conservative on their innovation strategies than those 19

granted low inside debt. To verify the findings, we conduct multivariate analysis and examine our first and second hypothesis. The first section of multivariate analysis, we examine whether inside debt is associated with innovation activities. In the second section of multivariate analysis, we further decompose innovation activities into in-house patenting activities and patent purchasing activities. Table 5 shows the results of multivariate analysis. The result in the first column of table 5 indicates that the coefficient on INSIDE (- 0.015) is negative and at 5% significant, when dependent variable is innovation. This finding support our first hypothesis that CEOs holding more inside debt would invest less innovation activities. The results in the second column of table 5 indicate that the coefficient on INSIDE (-0.013) is negative and at 5% significant, when dependent variable is RNDR. This finding supports our hypothesis 2A that CEOs holdings more inside debt would invest less in-house patenting activities. The results in the third column of table 5 indicate that the coefficient on INSIDE (-0.002) is negative but not significant, when dependent variable is PURCHASER. This finding does not support our hypothesis 2B that CEOs holdings more inside debt would invest more patent purchasing activities. Overall, the results in table 5 show that granting CEOs inside debt to CEOs would alliance with their self-interests with debtholders and try to avoid downside risk from investing intellectual investment projects. [Insert table 5 here] Effects of CEOs Inside Debt and Firm Innovation Strategies on Firm Performance This section tries to answer a question that how CEOs inside debt and firm innovation activities affect firm performance. To examine this question, we look at two aspects of firm performance. First, we set Tobin s Q to measure the market value of firms. We try to see how market look at the strategies of CEOs compensation scheme and firm innovation activities. 20

The first column of table 6 shows that the coefficient (0.724) on the interaction term of inside debt (INSIDE) and innovation (INNOVATION) is positive and at 10% significant, when dependent variable is Tobin s Q. The second column of table 6 shows that the coefficient (1.395) on the interaction term of inside debt (INSIDE) and in-house patenting activities (RNDR) is positive and at 10% significant, when dependent variable is Tobin s Q. In the same column, we also find that the coefficient (1.183) on RNDR is positive and at 1% significant. In the third column of table 6, we find that the coefficient (0.839) on the interaction term of inside debt (INSIDE) and patent purchasing activities (PURCHASER) is positive but not significant, when dependent variable is Tobin s Q. Accordingly, we find that in-house patenting activity is positively associated with market value of firms. When firms with high inside debt and in-house patenting activities, market still grants the positive valuation on the firms. The explanation for this result is that investing in-house patent activities positively affects firm value and outweigh negative effect from granting inside debt to CEOs. [Insert table 6 here] Second, we set 5-year average return on equity (ROE) to measure the accounting performance of firms. We try to see how the strategies of CEOs compensation scheme and firm innovation activities affect on firms long-term accounting performance. The first column of table 7 shows that the coefficient (0.122) on the interaction term of inside debt (INSIDE) and innovation (INNOVATION) is positive but not significant, when dependent variable is ROE. The second column of table 7 shows that the coefficient (0.721) on the interaction term of inside debt (INSIDE) and in-house patenting activities (RNDR) is positive and at 1% significant, when dependent variable is ROE. In the same column, we also find that the coefficient (0.302) on RNDR is positive and at 10% significant. 21

In the third column of table 6, we find that the coefficient (-0.005) on the interaction term of inside debt (INSIDE) and patent purchasing activities (PURCHASER) is negative but not significant, when dependent variable is ROE. We also find that the coefficient (0.313) on patent purchasing activities (PURCHASER) is also positive and at 1% significant. Accordingly, we find that in-house patenting activity is positively associated with long-term accounting performance of firms. When firms with high inside debt and in-house patenting activities, firms still have high earnings performance. The explanation for this result is that investing in-house patent activities positively affects firm value and outweigh negative effect from granting inside debt to CEOs. [Insert table 7 here] Conclusion Nowadays company with better profit spend millions on their innovation activities, innovation is one of the most important factors to enhance firm core value. And recent patent fighting and lawsuits among high technology companies indicate that using different innovation strategies comes out different economic consequences. To investigate the factors that influence firms innovation strategies is the important contribution for practitioners and literatures. Earlier literatures have documented that managerial compensation is the main factor for inducing innovation activities. Our study try to get further evidence by separating innovation activities to two parts: In-house patenting and Patent purchasing, and how CEOs compensation (inside debt) will affect their behavior to innovation and firms performance. According to previous literatures, innovation activities come up with the uncertainty of future cash flow and increasing the information asymmetry between firm and investors. As the managers holding high amount of inside debt, they are likely to behave more conservative on investment policies. Earlier literatures indicate that in-house patenting has larger 22

substantial cost than patent purchasing, firms need to maintain their in-house patenting investment to avoid the substantial adjustment costs. Firms need a consistent financing resource to support in-house patenting activities for avoiding the potential adjustment cost. Since CEOs inside debt outweigh the equity based compensation, they could be more conservative. Therefore, we can expect that managers holding high inside debt would like to invest patent purchasing activities, instead of in-house patenting activities. Our first hypothesis is to examine whether CEOs holding low inside debt would invest more innovation activities. The second hypothesis is to examine whether CEOs holding low (high) inside debt would invest more in-house patenting activities (patent purchasing activities). The third hypothesis is to examine how CEOs inside debt and innovation activities affect on firm performance. To examine our hypothesis, we use Standard & Poor (S&P) 500 firms over 2006-2014 as our sample firms. We also exclude the firms not in the industries where are the pharmaceutical products, chemicals, machinery, business services, computers, and electronic equipment. According to our observation, these industries are intellectual intensive industries. The results from univariate analysis shows that innovation activities, in-house patenting activities, patent purchasing activities in the firms with low CEOs inside debt are significantly higher than those with high CEOs inside debt. We also find that market value of firms and long-term accounting performance in the firms with low CEOs inside debt are significantly higher than those with high CEOs inside debt. The results from multivariate analysis shows that granting more inside debt to CEOs encourage CEOs to avoid to invest innovation activities, in terms of in-house patenting activities. When we look at the synergy of CEOs inside debt and innovation strategies, we find that in-house patenting activities positively affect on market value of firms and long-term accounting performance. However, 23

granting CEOs inside debt to CEOs will reduce the long-term performance, in terms of longterm accounting performance. We have several research limitations. First, due to the limitation the collecting data for patent purchasing activities, we focus on the S&P 500 listed firms only. Second, our results not fully support our hypothesis. For the further research, we will reconsider the theoretical problem behand the hypothesis and see whether we should revise the hypothesis. Especially, we should look at the relationship between CEO s conservative behavior and investment strategies. References Brown, J. and B.C. Petersen, 2011, Cash holdings and R&D smoothing. Journal of Corporate Finance 17: 694-709. Brown, J., S.M. Fazzari, and B.C. Petersen, 2009, Financing innovation and growth cash flow, external equity, and the 1990s R&D boom. Journal of Finance 64: 151-185. Cassell, C.A., S.X. Huang, J.M. Sanchez, and M.D. Stuart, 2012, Seeking safety the relation between CEO inside debt holdings and the riskiness of firm investment and financial policies, Journal of Financial Economics 103: 588-610. Choy, H., J. Lin, and M.S. Officer, 2014, Does freezing a defined benefit pension plan affect firm risk?. Journal of Accounting and Economics 57:1-21. Coles, J.L., N.D. Daniel, and L. Naveen, 2006, Management incentives and risk-taking. Journal of Financial Economics 79: 431-468. Edmans A. and Q. Liu, 2011, Inside Debt. Review of Finance 15:75-102. 24

Eisdorfer A., C. Giaccotto, and R. White, 2013, Capital structure, executive compensation, and investment efficiency. Journal of Banking and Finance 37: 549-562. Holthausen, R.W., D.F. Larcker, and R.G. Sloan, 1995, Business unit innovation and the structure of executive compensation. Journal of Accounting and Economics 19: 279-313. Hsu P.H., 2009, Technological innovations and aggregate risk premiums. Journal of Financial Economics 94: 264-279. Jensen, M. and W. Meckling, 1976, Theory of the firm: managerial behavior, agency costs, and ownership structure. Journal of Financial Economics 3: 305-360. Liu Y., D.C. Mauer, and Y. Zhang, 2014, Firm cash holdings and CEO inside debt. Journal of Banking and Finance 42: 83-100. McAlister, L., R. Srinivasan, and M.C. Kim, 2007, Advertising, research and development, and systematic risk of the firm. Journal of Marketing 71: 35-48. Narula, R., 2001, Choosing between internal and non-internal R&D activities: some technological and economic factors. Technology Analysis & Strategic Management 13: 365-386. O'Regan, N. and G. Kling, 2010, Technology outsourcing in manufacturing small- and medium-sized firms: another competitive resource? R&D Management 41: 92-105. Pastor, L. and P. Veronesi, 2009, Technological revolutions and stock prices. American Economic Review 99: 1451-1483. 25

Ryan Jr., H.E. and R.A. Wiggins III, 2002, The interactions between R&D investment decisions and compensation policy. Financial Management 31: 5-29. Sundaram R.K. and D.L. Yermack, 2007, Pay me later: inside debt and its role in managerial compensation. Journal of Finance 62: 1551-1588. 26

Table 1 Descriptive Statistic-Full Sample This table shows the full sample of the descriptive statistics for innovation activity measurements, inside debt measurements, performance measurement, and control variables. Variables Mean Median Max Min Q1 Q3 STD N INSIDE 0.476 0.000 1.000 0.000 0.000 1.000 0.500 649 CEO TO FIRM 6.378 0.842 170.741 0.000 0.000 3.499 22.457 649 RNDR 0.071 0.047 0.741 0.004 0.024 0.104 0.074 649 XRD 1,008.130 300.000 8,168.000 6.502 98.900 939.918 1,740.090 649 PURCHASER 0.015 0.000 1.575 0.000 0.000 0.000 0.100 649 PATENT PURCHASE 194.280 0.000 7,800.000 0.000 0.000 0.000 1,078.970 649 INNOVATION 0.086 0.049 1.908 0.004 0.024 0.111 0.136 649 TOBINSQ 1.922 1.694 5.608 0.854 1.346 2.285 0.842 649 ROE t=1~5 0.512 0.480 1.280-0.221 0.361 0.678 0.249 648 CASHHOLDING 0.219 0.176 0.824 0.005 0.092 0.287 0.174 649 MARKETBOOK 3.233 2.641 52.761-23.021 1.699 4.033 5.559 649 SGWTH 0.042 0.039 0.483-0.494-0.020 0.099 0.139 649 STKISSUES -0.033-0.020 0.179-0.295-0.055 0.001 0.066 601 DBTISSUES 0.025 0.000 0.362-0.158-0.009 0.037 0.084 598 rcashholding 0.000 0.000 0.000 0.000 0.000 0.000 0.000 648 PROFITABILITY 0.155 0.149 0.460-0.029 0.102 0.193 0.085 649 CAPXR 0.036 0.027 0.177 0.005 0.017 0.045 0.030 649 SIZE 8.964 8.759 12.240 6.109 8.091 9.738 1.232 649 LEVERAGE 0.240 0.205 0.948 0.003 0.123 0.306 0.170 649 PROFITABILITY 0.155 0.149 0.460-0.029 0.102 0.193 0.085 649 TANGIBILITY 0.412 0.306 1.460 0.072 0.200 0.536 0.301 648 CONSTRAINT 0.396 0.000 1.000 0.000 0.000 1.000 0.489 649 INSIDE, an indicator variable for measuring CEOs inside debt. We set as an indicator equal to one if CEO_TO_FIRM is greater than one; CEO_TO_FIRM, the ratio of CEO to firm debt/equity ratio. We have the ratio of CEOIDH to CEOEH be scaled by the ratio of FD to FE; CEOIDH (CEO inside debt holdings), the sum of the present value of accumulated pension benefits and deferred compensation; CEOEH (CEO equity holdings), the sum of the market value of stock (including restricted shares) held by the CEO at the fiscal year-end, the estimated value of in-the-money unexercised exercisable options, and the estimated value of in-the-money unexercised unexercisable options; FD, the sum of total current liabilities and long-term debt; FE is the market value of stock at fiscal year end; RNDR, ratio of research and development expense to lagged total assets; XRD, the amount of research and development expense; PURCHASER, the ratio of patent purchasing to lagged total assets; PATENT_PURCHASE, the amount of patent purchase; INNOVATION, the ratio of the sum of the amount of patent purchasing and the amount of research and development expense to lagged total assets; TOBINSQ, the ratio of the market value of assets to the book value of assets; ROE t=1~5, five year average return on equity; CASHHOLDING, the ratio of the sum 27

of cash and short-term investments to lagged total assets; MARKETBOOK, market-to-book ratio; SGWTH, the ratio of the change in sale revenue; STKISSUES, the ratio of net cash raised from stock issued to lagged total assets; DBTISSUES, the ratio of net new long-term debt issued to lagged total assets; rcashholding, the ratio of the sum of change in cash and short-term investments to lagged total assets. PROFITABILITY is the ratio of EBITDA to total assets. CAPXR is the ratio of capital expenditure to the lagged total assets. SIZE is the value of the natural logarithm of total assets. LEVERAGE is the ratio of total debt in period t divided by the book value of the lagged total assets. TANGIBILITY is the ratio of fixed assets to lagged total assets. CONSTRAINT is an indicator variable that takes the value of 1 if the firm didn t pay a cash dividend. 28

Table 2 Descriptive Statistic-Patent Purchasing Sample Only This table shows the patent purchasing sample of the descriptive statistics for innovation activity measurements. This table represents that if firms invest patent purchasing activities, the amount will be higher than their in-house patenting activities. Variables Mean Median Max Min Q1 Q3 STD N RNDR 0.100 0.073 0.626 0.000 0.032 0.117 0.117 54 XRD 2,919.770 1,367.550 8,168.000 1.541 170.725 5,271.000 3,039.970 54 PURCHASER 0.252 0.125 1.575 0.002 0.024 0.363 0.322 54 PATENT PURCHASE 3,324.900 1,900.000 7,800.000 10.000 500.000 7,300.000 3,079.190 54 a. RNDR, ratio of research and development expense to lagged total assets; XRD, the amount of research and development expense; PURCHASER, the ratio of patent purchasing to lagged total assets; PATENT_PURCHASE, the amount of patent purchase. Table 3 Correlation Table This table shows the results of correlation test for main variables. Variables INSIDE RNDR PURCHASER RNDR -0.158 *** PURCHASER -0.046 0.206 *** INNOVATION -0.120 ** 0.697 *** 0.845 *** a. INSIDE, an indicator variable for measuring CEOs inside debt. We set as an indicator equal to one if CEO_TO_FIRM is greater than one; RNDR, ratio of research and development expense to lagged total assets; PURCHASER, the ratio of patent purchasing to lagged total assets; INNOVATION, the ratio of the sum of the amount of patent purchasing and the amount of research and development expense to lagged total assets; b. ***, **, * indicates significance at the 1%, 5%, and 10% level, respectively. 29