Printer Friendly

Policy Uncertainty and the Dual Role of Corporate Political Strategies.

"This country has come to feel the same when Congress is in session as
when the baby gets hold of a hammer."
--Will Rogers, New York Times, July 5th, 1930.


In spite of ample prior empirical findings that support the notion that corporate political strategies have value relevance, there is very little direct evidence with regard to the underlying mechanism, which in the framework of a discounted cash flow valuation model could be a numerator and/or a denominator effect. Essentially, corporate political strategies can reduce the impact of political uncertainty on systematic risk, thereby reducing the cost of capital and/or propel favorable changes in industry dynamics thus enhancing growth opportunities. The relative importance of these two effects has not been explicitly addressed at the firm level in prior studies. Boutchkova et al. (2012) in a cross-country study examine at the industry level how local and global political risks affect the systematic and unsystematic components of industry return volatility. We intend to fill this void in the literature by empirically investigating at the firm level: 1) how corporate political strategies, political uncertainty, and their interactions affect firms' systematic risk, and 2) whether corporate political strategies enhance the value relevance of firms' real options and if this effect is more pronounced when political uncertainty is high.

Since political uncertainty can take many different forms, we choose to focus on the type of uncertainty about future cash flows that emanates from legislative activity (i.e., policy uncertainty). (1) Policy uncertainty implies that there is a greater array of both threats and opportunities for affected firms (Kim, Pantzalis, and Park, 2012). We proxy policy uncertainty here by the number of bills that have the potential to affect a firm's future business landscape.

We account for three distinct types of corporate political strategies: 1) appointing ex-politicians on corporate boards, 2) making political action committee (PACs) contributions, and 3) lobbying. For each of these types of strategies, we use several measures that have been previously employed in the literature. In addition, we devise a composite measure of corporate political strategies, which we label the Political Strategy Index (PSI). Our investigation spans 15 years, from 1994 to 2008, utilizing over 66,000 firm-year observations. Consistent with the notion that exposure to uncertainty emanating from legislative activity is a source of systematic risk, we find that beta increases with the number of value relevant bills introduced over the past year. This result is also in line with the industry-level findings of Boutchkova et al. (2012) who find that domestic political uncertainty is positively correlated with the systematic component of industries' return volatility. More importantly, we find that corporate political strategies are associated with lower betas, and that this relationship is more pronounced when political uncertainty is higher. Our beta regressions' results are in line with the view that stocks by firms that have access to political intelligence co-vary less with the market. (2)

Contrary to the beta regressions findings, our political connection measures and their interactions with policy uncertainty are shown to be significantly and positively correlated with idiosyncratic risk. Given that idiosyncratic risk can arise from the way innovation affects the uncertainty of expected future profits (Shiller, 2000; Campbell et al., 2001; Mazzucato and Tancioni, 2008), we interpret this result as suggesting that firms that have the means to mitigate political uncertainty are better at innovation (Ovtchinnikov, Reza, and Wu, 2015). A potentially important aspect of innovation can be found in corporate political participation. The extraordinary growth of corporate lobbying and other forms of corporate political participation over the past few decades can be viewed as the result of a path-dependent learning process (Drutman, 2015). Companies may initially be reluctant to become politically active, but once they start doing so, they can gain more confidence in their ability to not only protect themselves from government actions, but also to expand their growth opportunities in business environments increasingly affected by political uncertainty. Through their political activism, firms can gather valuable political intelligence, thereby lowering political uncertainty. As such, corporate political activism propels innovation (Ovtchinnikov et al., 2015).

Alternatively, our idiosyncratic risk findings could be viewed as in line with the argument made in Chun et al. (2008) that firm-specific performance heterogeneity (i.e., idiosyncratic risk) may be a "finer and more nuanced metric" of the intensity of creative destruction that economic growth theorists envision as the process wherein creative innovators dominate laggards. Thus, in a creative destruction framework (Chun et al., 2008; Chun, Kim, and Morck, 2011), it is possible that political connections are accentuating firm heterogeneity within industries (consisting of a mix of early adopters of political strategies and laggards) making firms' portfolios of real options more value relevant as potential drivers of competitive advantages in uncertain environments (Trigeorgis, 1996; Trigeorgis and Lambertides, 2014). (3)

We demonstrate empirically that politically connected firms possess more value relevant real options than non-connected firms. Specifically, we find that the stock returns of connected firms increase (decrease) more than those of non-connected firms when their stock return volatility increases (decreases). This effect is significantly more pronounced among firms operating in more uncertain policy environments, consistent with the notion that real options become more valuable in such environments as the connected firm is in a better position to exploit the extra managerial flexibility that comes with being connected.

We examine whether the aforementioned effects change as predicted after exogenous shocks proxied here by cases of a politically connected board member's sudden death. Our results provide confirmation of the notion that causality runs from corporate strategies to risk and/or returns. We obtain similar results when we repeat our earlier tests using three different measures of corporate political strategies and alternative measures of policy uncertainty. Furthermore, our results are robust to the exclusion of low priced (less than $5 per share) stocks and the use of weekly returns. Overall, our evidence supports the notion that corporate political strategies can both alleviate the impact of political uncertainty on systematic risk and serve as facilitators of competitive advantage in uncertain times.

Our paper contributes to the recent policy uncertainty and corporate political connection literature. First, prior literature provides evidence that political or policy-related uncertainty affects country- or industry-wide stock market volatility (Bialkowski, Gottschalk, and Wisniewski, 2008; Fuss and Bechtel, 2008; Boutchkova et al., 2012). Our investigation is done at the firm level by exploiting each firm's political activities. In addition, our findings contribute to our better understanding of the mechanisms through which corporate political strategies create/destroy value. Politically active firms, measured as firms with PAC contributions, increase innovation prior to industry deregulation (Ovtchinnikov et al. 2015) and delay investment in anticipation of future lucrative tax incentives (Wellman, 2017) as connected firms are capable of timing industry deregulation and tax regulation changes in advance. Our paper complements these papers that identify the mechanisms through which political activism creates value. We formally provide the underlying mechanism of corporate political strategies' value relevance by investigating systematic and idiosyncratic risks of connected firms and their real options in the valuation framework. Moreover, most political connections studies suffer from endogeneity issues. We use ex-politicians' sudden deaths as an exogenous shock to establish a causal relationship. We confirm that causality runs from corporate political strategies to risk and growth opportunities. Finally, we employ a comprehensive sample of corporate political activism, such as firms with politically connected boards, lobbying, and political contributions, as opposed to relying on the use of a single political connection measure that only represents one tactic among a menu of choices available to politically active firms. In doing so, we mitigate the measurement errors of corporate political connectedness.

I. Literature Review and Hypotheses Development

Prior studies have found that political connections are value relevant. However, there are conflicting views as to how corporate political strategies affect the future stock performance of connected firms. One stream of the literature supports the political capital view that political connections enhance firm value through political rent-seeking behaviors by a connected firm (Faccio, 2006; Goldman, Rocholl, and So, 2009). Additionally, political connections have been shown to help firms access cheaper financing through equity (Boubakri et al., 2012), public debt (Bradley, Pantzalis, and Yuan, 2016), and bank loans (Houston et al., 2014), as well as through initial public offerings (IPOs) (Francis, Hasan, and Sun, 2009).

Although much of the literature supports the political capital view, there is also ample evidence (Chen, Ding, and Kim, 2010; Asian and Grinstein, 2011; Chaney, Faccio, and Parsley, 2011; Kim and Zhang, 2016) that political connections can be associated with riskier corporate behavior and, in a sense, render firms distinct from their non-connected peers. (4) Furthermore, Aggarwal, Meschke, and Wang (2012) find that PAC contributing firms underperform non-contributing firms as campaign donations are more likely a symptom of agency problems. Kang and Zhang (2015) uphold Aggarwal et al.'s (2012) findings by providing evidence in line with the view that politically connected directors are not as effective as other outside directors when monitoring and advising managers. In sum, this strand of the literature supports the view that the complicated and opaque nature of political connections makes firms riskier and increases the level of information asymmetry.

Overall, the long list of papers documenting the value relevance of political connections has still not provided exhaustive evidence regarding the specific channels through which connections can affect firm valuation. We intend to contribute to the literature by unraveling this controversial matter and providing evidence that helps reveal certain channels through which political connections are expected to affect valuation. Our first goal is to demonstrate that political connections can function as a hedging mechanism that can be moderated by policy uncertainty.

Earlier studies (Alesina and Rodrik, 1994, among many others) have provided theoretical arguments and confirming evidence that broad economic factors, like inflation and unemployment, are affected by political developments. Several more recent studies explore whether policy uncertainty affects asset value. Croce et al. (2012) investigate whether uncertainty about tax policy affects both bond and equity prices and report that it is indeed the case. Cohen, Coval, and Malloy (2011) examine the effect of uncertainty induced by "changes in congressional committee chairmanship" on economic activities including corporate investment, employment, and productivity. They find that following the appointment of a new chairman of a congressional committee, the politician's home state obtains additional federal outlays, government fund transfers, and government procurement contracts. This increase in available state funds discourages local corporate investment, employment, and productivity (i.e., government spending "crowds out" corporate economic activities). Belo, Gala, and Li (2013) argue that government policy is primarily shaped by the level of partisanship. Cohen, Diether, and Malloy (2013) find that after the passage of bills, firms headquartered in a legislator's home state experience positive abnormal returns. The phenomenon is more pronounced for an interested group comprised of firms belonging to a specific industry corresponding to each bill. Pastor and Veronesi (2012) theoretically analyze the impact of uncertainty about government policy on stock prices. A key feature of their model is dividing uncertainty arising from government policy into two parts: "political uncertainty" associated with changes in policy and "impact uncertainty" associated with the magnitude of the effect on stock price when a policy is implemented. They find that both types of uncertainty affect stock prices. Kim et al. (2012) measure uncertainty about future policies for different areas of the political map. Their proxy for policy uncertainty is constructed after general elections held every two years in the US using the degree of different state politicians' partisan alignment with the incumbent president (a measure they label as political alignment index or PAI). Firms whose headquarters are located in high PAI areas experience higher positive abnormal returns than those located in low PAI areas in both time series and cross-sectional tests, consistent with the notion that policy risk, as reflected in a dynamically changing political map, affects stock returns. Overall, regardless as to which policy uncertainty proxy has been used, the notion that it has value implications has recently gained strong scholarly support.

This notion is also complemented by the evidence of studies examining how politics affect stock market volatility. Most of this evidence comes from studies examining the country-wide effects of political uncertainty, either by focusing on a single country (Bailey and Chung, 1995; Herron, 2000; Leblang and Mukherjee, 2005; Fuss and Bechtel, 2008) or in a cross-country setting (McGillivray, 2003; Bialkowski et al., 2008). In a more recent study, Boutchkova et al. (2012) focus on industries and demonstrate that some sectors are more susceptible to political uncertainty than others. Boutchkova et al. (2012) also determine that industries that are sensitive to political factors are more volatile during periods of higher political uncertainty.

In contrast to the aforementioned studies, our empirical investigation is done at the firm level and entails testing 1) whether corporate political strategies on the systematic and unsystematic components of firm risk and 2) whether the effect is moderated by policy uncertainty. Motivated by Cohen et al.'s (2011, 2013) work, we use the number of legislative bills that are linked to the industry in which a firm belongs and that are introduced by state congressmen (either Senators or House Representatives) as a proxy for policy uncertainty. Given the legislature's ability to affect share prices through its policy actions (Ferguson and Wittee, 2006; Kim et al., 2012; Pastor and Veronesi, 2012), we hypothesize that firms' exposure to policy uncertainty emanating from legislative activity will affect firms' systematic risk. Moreover, if firms' political participation is motivated by gaining protection from government action (Drutman, 2015), corporate political strategies should alleviate the impact of exposure to systematic risk. Alternatively, the magnitude of the effect of connected firms' access to political intelligence can be dependent upon policy uncertainty. To test the aforementioned hypotheses empirically, we set a regression as follows:

[Beta.sub.i,t+1] = [[beta].sub.0] + [[beta].sub.1] Ln([Bills.sub.i,t]) + [[beta].sub.2] Political [strategy.sub.i,t] + [[beta].sub.3] Ln([Bills.sub.i,t]) * Political [strategy.sub.i,t] + [SIGMA] [beta][[CHI].sub.i,t] + [SIGMA] Year + [SIGMA] Indmtry + [[epsilon].sub.i,t], (1)

where Beta is systematic risk, computed from the market model using daily returns over the year. Bills is equal to the number of legislative bills linked to the industry in which a firm belongs and that are introduced/sponsored by either the home state Senators or House Representatives. As mentioned previously, we use Bills as a proxy for uncertainty regarding the impact of policy initiatives on future cash flows (i.e., exposure to policy uncertainty). Boutchkova et al. (2012) find that domestic political uncertainty is positively related to systematic volatility. Thus, we expect a positive [[beta].sub.1] coefficient. Further, a negative [[beta].sub.3] coefficient would lend support for the hypothesis that corporate political strategies can act as a hedging mechanism that can mitigate policy uncertainty's effects on systematic risk. In our robustness tests, we also use two alternative measures of policy uncertainty by decomposing Bills into the annual bills that are sponsored by the firm's home state politicians (hereafter [Bills.sup.loc]) and the bills linked to the industry in which a firm belongs (hereafter [Bills.sup.ind]). For corporate political strategies (Political strategy), we use several alternative measures pertaining to three such strategies: 1) having former politicians on corporate boards, 2) making PAC contributions, and 3) lobbying. We include a set of control variables representing factors that past studies (Hamada, 1972; Subramanyam and Thomadakis, 1980; Hong and Sarkar, 2007) have argued should be correlated with beta. Specifically, the control variables are as follows. Size is the market value of equity at the end of year y. BM is the ratio of the book-to-market value of equity. Leverage is the total long-term debt divided by total assets. ROA is net income divided by total assets. Tangibility is property, plant, and equipment divided by assets. HHI is the Herfindahl index based on sales of the first three digits of the standard industrial classification (SIC) code. Firm age is the years since a firm is first listed in Compustat.

In the regression, we use standard errors that are robust to clustering at both the firm and year levels following the procedure of Thompson (2011). The estimate of the variance-covariance matrix is: [V.sub.FIRM&Time] = [V.sub.Firm] + [V.sub.Time] - [V.sub.White]. This method combines the standard errors clustered by firm with the standard errors clustered by time. In the computation, the White (1980) variance-covariance matrix is subtracted to prevent double counting the diagonal of the variance-covariance matrix.

Approximately, 22.9% of our sample firms are engaged in at least one of three political strategies (through directors, PACs, or lobbying). These firms tend to be particularly large. To ensure that our political strategy variables measure political connections independent of size, we use the size-orthogonal political strategy measures throughout our empirical analyses. These orthogonal measures are the residuals obtained from regressing each political strategy variable on market capitalization. The results of these regressions are reported in Appendix A. Consistent with our expectations, firm size is strongly and positively related to all of the political strategy variables. In four regressions, the t-statistic value on the estimated coefficient of firm size ranges from 23.22 to 38.45. Our results are qualitatively equivalent when we use the raw values of the political strategies (untabulated).

We also argue that the interplay of policy uncertainty with political connections should be associated with greater levels of idiosyncratic volatility. This hypothesis is formed by combining two pieces of evidence. First, Mazzucato and Tancioni (2008) demonstrate that idiosyncratic volatility is higher in innovative industries and among firms characterized by greater uncertainty about their future earnings. In a similar vein, Shiller (2000) argues that idiosyncratic risk is much higher during periods of technological revolutions when investors are more likely to exhibit behavioral biases. Campbell et al. (2001) find that the steady increase in idiosyncratic risk since the 1960s can be partly attributed to the effect of the IT revolution. In addition, Ovtchinnikov et al. (2015) confirm that politically active firms successfully time future legislation and set their innovation strategies in expectation of future legislative changes. Their findings support the notion that political activism can help reduce political uncertainty, which, in turn, fosters firm innovation. Thus, in the face of policy uncertainty, firms that adopt corporate political strategies should exhibit greater levels of idiosyncratic volatility, ceteris paribus.

We proceed to test this by estimating the following model:

[IV.sub.i,t+1] = [[beta].sub.0] + [[beta].sub.1] Ln([Bills.sub.i,t]) + [[beta].sub.2] Political [strategy.sub.i,t] + [[beta].sub.3] Ln([Bills.sub.i,t]) * Political [strategy.sub.i,t] + [SIGMA][[beta].sub.t][[CHI].sub.i,t] + [SIGMA] Year + [SIGMA] Indmtry + [[epsilon].sub.i,t] (2)

where the dependent variable, firm-specific performance heterogeneity (i.e., idiosyncratic volatility) is defined as the variance of residuals obtained from the Carhart (1997) four-factor model in which a firm's daily returns are regressed on the size, value, and momentum factors along with the market risk premium over the year. The aforementioned hypothesis implies that the coefficient [[beta].sub.3] should be positive. Furthermore, if the adoption of corporate political strategies can cause industries to experience creative destruction-like effects and greater levels of firm heterogeneity, then we expect a positive coefficient [[beta].sub.1]. (5)

Our next goal is to demonstrate whether or not political connections can also function as a source of growth opportunities. Trigeorgis (1993) provides examples of real options and points to the managers' ability to defer, expand, contract, abandon, or otherwise alter a project at different stages during its useful operating life. We argue that the size of a firm's real options portfolio should expand as the firm becomes more politically active. In addition, political connections should allow managers more flexibility in reacting to changes in the environment the firm operates in. As in Croci et al. (2017), we view corporate political strategies as intangible-type assets, a source of political intelligence that should enhance management's ability to better time its responses to political developments. This effect should be particularly valuable when firms face high levels of uncertainty. Thus, we posit that corporate political strategies can boost growth opportunities by serving as a facilitator of valuable real options. (6)

In order to assess whether corporate political strategies are associated with valuable real options at the firm-specific level, we utilize a well-established fact from the option literature. Specifically, we exploit the fact that because their asymmetric payoff profile options have valuations that are strictly increasing in the volatility of the underlying asset. Thus, if a connected firm holds more real options than a comparable non-connected firm, it must be the case that the value of the connected firm increases (decreases) more than the value of non-connected firm when volatility increases (decreases).

Grullon, Lyandres, and Zhdanov (2012) hypothesize and determine that the positive relation between firm-level stock returns and firm-level stock return volatility documented in Duffee (1995) can be driven by firms' real options. They find strong empirical support for this notion. Firms with abundant investment opportunities (small firms, young firms, research and development (R&D) firms, and high growth firms) and high operational flexibility (firms in non-unionized industries and firms with high earnings and sales convexity) have a stronger positive relationship between firm stock returns and changes in firm stock volatility than firms with less investment opportunities and less operational flexibility. (7)

We measure firm i's volatility during month t as the standard deviation of the firm's daily returns during month t.

[mathematical expression not reproducible]

where, [R.sub.i,t,[tau]] is the firm i's excess return ([r.sub.i,t,[tau]] - [r.sub.f,t,[tau]]) on day [tau] in month t and [n.sub.t] is the number of trading days in month t. [mathematical expression not reproducible] is the mean excess return of firm i in month t. Then, we estimate the following Fama-MacBeth (1973) cross-sectional regression, which is similar to the models found in Grullon et al. (2012):

[R.sub.i,t] = [[beta].sub.0] + [[beta].sub.1] [DELTA][Volatility.sub.i,t] + [[beta].sub.2] Political [strategy.sub.i,t] + [[beta].sub.3] [DELTA][Volatility.sub.i,t] * Political [strategy.sub.i,t] + [SIGMA][[beta].sub.t][[CHI].sub.i,t] + [[epsilon].sub.i,t], (4)

where, [R.sub.i,t] is firm i's stock excess return in month t, [DELTA][Volatility.sub.i,t] is the month-to-month change in the volatility of the stock's daily excess returns, and [[CHI].sub.i,t] indicates the controlling variables.

If corporate political strategies provide the firm with growth opportunities that enhance the firm's flexibility (real options), we expect to find a positive and significant coefficient on the interaction term, [[beta].sub.3]. We will perform the above tests using several alternative measures of corporate political strategies and for subsamples formed after sorting on alternative measures of policy uncertainty measured at the industry and state levels. More specifically, we measure policy uncertainty based on: 1) the number of congressional bills related to the industry in which a firm belongs, 2) the number of bills drafted and introduced in Congress by local (i.e., from the state the firm headquarters is located) congressmen, or 3) the number of congressional bills that are related to the industry in which a firm belongs to and that are drafted and introduced in Congress by local congressmen of a firm. Information on each congressional bill will be collected from the Congressional Bills Project (http://www.congressionalbills.org/index.html).

II. Data Selection and Variable Description

An important contribution of our paper is that it provides evidence from a large, comprehensive political connections' dataset. We construct a fairly large and diverse set of political variables at the firm and state levels and utilize them in our investigation of political connections and its effect on stock returns. We will introduce them with detailed information on data sources and construction in the following subsections.

A. Directors' Political Experience

To identify political connections that are based on the composition of a firm's board of directors, we use the electronic data gathering, analysis, and retrieval (EDGAR) database that can be downloaded from the Securities and Exchange commission (SEC) website (ftp://ftp.sec.gov/edgar/full-index/) and search Form 10-K filings for board information including a firm's name, filing dates, types of filing, the central index key (CIK), and every director's name and their short biography. While we are able to determine a director's political experience by reading their individual biography, we also account for the many cases where the biographical information is either missing or incomplete. Thus, we collect lists of the US politicians from various sources that provide information on a politician's former or incumbent political position, party affiliation, and years in the position and resigning from the position. (8) We then use the politicians' names from our lists of the US politicians to link them with the board of directors' information extracted from EDGAR. This procedure enables us to construct a rich dataset that measures various ways a firm's board can provide the firm with political connectedness. (9)

As in Houston et al. (2014) and Kim and Zhang (2016), we use three main political connection measures that are continuous variables, as opposed to discrete, using corporate board information: 1) the number of politically connected directors, 2) a board's political experience (i.e., the average tenure of past political activities of a board of directors), and 3) political freshness measured by the elapsed period from the year a connected director left politics to the year they serve as a corporate director. These variables are intended to capture the degree and nature of the board's connectedness. (10) Further detailed definitions of the variables are reported in Appendix C.

B. Corporate Political Contributions and Lobbying Expenditures

We also devise measures of alternative corporate political strategies based on two types of politically-related corporate expenditures that are publicly available: 1) corporate contributions to US political campaigns and 2) lobbying expenditures. We extract the corporate contributions data from the Federal Election Commission (FEC) summary files on political contributions to US House and Senate election campaigns. Following Cooper, Gulen, and Ovtchinnikov (2010), we construct four different measures of political connectedness using corporate political contributions: 1) N. of supported candidates measured as the number of politicians running for office supported by the firm, 2) Strength of relationships measured as the total length of relationships between the firm and the candidates, 3) Supported candidates' ability measured by the home state of the firm and the candidate, and 4) Supported candidates' power measured by the candidate's committee ranking.

We collect corporate lobbying expenditures from 1998 to 2008 from the OpenSecrets website (http://www.opensecrets.org) that keeps track of the influence of money on US elections and public policy. After passage of the Lobbying Disclosure Act of 1995, the Secretary of the Senate and the Clerk of the House of Representatives are required to disclose lobbying-related information, verify its accuracy, and compile lobbying data. Data includes filing dates for lobbying activities, lobbying amounts, registrant name and address, client's name, as well as the address and industry classification related to a bill involving lobbying by a firm. For instance, 3M Co. filed its year-end report on March 7, 2002 that accounts for lobbying activities d from January 1, 2001 to December 31, 2001. The company's total lobbying expenditures were $877,100 spent on 27 different industry-specific bills. Since the data does not allow us to track how much money has been spent on a specific bill, we can only measure aggregate corporate lobbying expenditures by adding up all of the reported expenses by firm and year.

C. Policy Uncertainty

Recent studies (Kim et al., 2012; Cohen et al., 2011, 2013) indicate that a major source of policy risk is uncertainty surrounding legislative activity. A widely shared popular view is that congressional activity interferes with the markets and injects uncertainty about the future. We argue that legislators often draft, sponsor, and/or amend bills with an eye on firms located in the geographic area that constitutes their political home, especially those firms with whom they are connected (Roberts, 1990; Jayachandran, 2006). This legislative activity creates uncertainty regarding the redistribution of future growth opportunities among firms within an industry and/or a state and can generate the perception of higher risk among investors (Kim et al., 2012).

In sum, we expect that policy uncertainty arises from high levels of legislative activity. However, unless they are intended to produce economy-wide effects, bills typically tend to have either a specific industry focus or to be targeting a specific geographic area promoting a specific policy. Thus, we do not expect that all legislative activities are equally important in terms of injecting uncertainty regarding the future cash flows of a particular firm. Therefore, we utilize three measures: 1) the number of bills related to the industry in which a firm belongs ([Bills.sup.ind]), 2) the number of bills drafted by local congressmen ([Bills.sup.loc]), and 3) the number of bills related to the industry in which a firm belongs and that are introduced by local politicians (hereafter Bills). We trace the information on each congressional bill that is obtained from the Congressional Bills Project (http://www.congressionalbills.org/index.html), and link the bills to one of 49 industries based on the bill's subject categorization developed by the Library of Congress and the Fama-French classification. The detailed mapping is available in Appendix B. The higher the number of [Bills.sup.ind], [Bills.sup.loc], and Bills, the greater the level of legislative activity-induced uncertainty surrounding firms.

D. Descriptive Statistics

Table I presents the descriptive statistics of the sample that includes 66,059 firm years from 1994 to 2008. We find that about 68 bills sponsored by local politicians and 124 industry-related bills are drafted each year in Congress. As reported in Table I, only 5.68 of them are both industry-related and introduced by local politicians, which are used as our policy risk (Bills). Our sample firms have 0.14 connected directors, make contributions to 9.63 candidates, and spend $ 126K for lobbying. In addition, the median market value of equity is $ 181 million with a median book-to-market ratio of 0.51.

III. Results

A. Political Strategy and a Firm's Systematic Risk

We begin our empirical tests by examining the relationship between a firm's political strategy and the systematic portion of its risk (Beta). In the cross-sectional tests presented in Table II, we regress a firm's beta on the policy uncertainty measure [Ln(Bills)], the residual political strategy index ([PSI.sup.R]), and their interaction along with other control variables. [PSI.sup.R] is the residual value of PSI, which combines three yearly rank-based political strategy variables: 1) B-index on the corporate board's political connectedness, 2) P-index on PAC contributions, and 3) L-index on lobbying expenditures.

Model (1) provides direct evidence as to how uncertainty about the impact of new policies manifests itself in a firm's systematic risk. Specifically, we find that in the cross-section of firms, greater policy uncertainty is associated with larger betas, a result consistent with the evidence in Boutchkova et al. (2012) that the systematic component of return volatility increases in domestic political uncertainty. In Model (2), we find that political connections are negatively correlated with a firm's beta. The coefficient of -0.23 indicates that a change in residual [PSI.sup.R] by 0.19 (equal to one standard deviation of [PSI.sup.R] as reported in Table I) is associated with a decrease by 0.04 in beta.

Our main focus is whether the hedging effect of political connections is dependent upon policy uncertainty. To address this issue, we add the interaction term between Ln(Bills) and [PSI.sup.R] to the regression model. In Model (3), the coefficient of the interaction term is -0.0481 and is significant at the 1 % level implying that the hedging effect of political connections on systematic risk is more pronounced when the firm's business environment is more uncertain. This moderating effect is economically large. An increase by one standard deviation (11.99) from the mean (5.68) in Bills enlarges the negative effect of political strategies by -0.06, which is about 35% (-0.06/-0.17) of the hedging effect without any industry uncertainty.

In Table II, we produce evidence based on [PSI.sup.R], an aggregated index of different political connections. From our sample, we find that only 22.9% of firms employ at least one of the three previously mentioned individual corporate political strategies. In the following tests presented in Table III, however, we separately explore the effectiveness of each political strategy as a hedging mechanism.

First, we consider directors' political ties. As previously mentioned in Section II, we construct four distinct measures to calibrate the degree and nature of corporate board political connectedness: N. of connected directors, Board's political experience, and Board's political freshness, and a composite measure, the B-index, which combines the yearly ranks of all three of the aforementioned variables. Next, we turn our focus to corporate contributions to PACs. The literature has provided evidence that firms benefit from maintaining ties to politicians through campaign contribution programs (Roberts, 1990; Jayachandran, 2006; Knight, 2006; Cooper et al, 2010; Shon, 2010). In a similar vein with the B-index, we create the P-index value using the ranks of the four measures of corporate political contributions developed by Cooper et al. (2010): 1) N. of supported candidates, 2) Strength of relationships, 3) Supported candidates' ability, and 4) Supported candidates' power. The earlier political economy literature's focus on political connections through corporate PAC contributions can be primarily attributed to the public availability of PAC donations data. In recent years, there has been increased attention on corporate lobbying activities. Some recent studies find that lobbying firms do better than non-lobbying firms in terms of both operating performance (Chen, Parsley, and Yang, 2015) and stock performance (Hill et al., 2013). To measure a firm's lobbying engagement, we create the L-index using the rank of corporate lobbying expenditures. The detailed definitions of all of the aforementioned political connections variables are included in Appendix C.

We regress firm beta on the residual values of B-index, P-index and L-index and report the results in Table III. Consistent with the PSI-based findings in the previous table, we find significant negative relationships between all three measures and firm risk [see Models (1), (3), and (5)]. The coefficients of the connectedness measures' interactions with the number of newly introduced relevant bills are negative and significant at conventional levels in all three regressions [see Models (2), (4), and (6)] indicating that each political strategy is an effective hedging mechanism on systematic risk.

B. Economic Significance of Political Strategy

Although the results in the past two tables imply that political strategy is an effective hedging tool for firms, it is not easy to interpret the economic significance of the results. Consequently, in the following tests, we examine a firm's risk-adjusted stock returns and the cost of equity.

First, we examine risk-adjusted returns for the portfolios formed by policy uncertainty and a corporate board's political connectedness. A firm is classified into the low (high) policy uncertainty group if the total number of bills in the firm's state is lower (higher) than the median value in year y. A firm is classified into the low (high) political strategy group if the firm's political strategy index is lower (higher) than the median value in year y. We calculate the risk-adjusted abnormal stock returns using the Carhart (1997) four-factor model.

[R.sup.P.sub.m] - [R.sup.F.sub.m] = [[beta].sub.0] + [[beta].sub.1] ([R.sup.M.sub.m] - [R.sup.F.sub.m]) + [[beta].sub.2][SMB.sub.m] + [[beta].sub.3][HML.sub.m] + [[beta].sub.4][UMD.sub.m] + [e.sub.m], (5)

where [R.sup.P.sub.m] is a particular portfolio's monthly return, [R.sup.F.sub.m] is the one-month Treasury bill rate, and [R.sup.M.sub.m] is the value-weighted market return. SMB (small minus big) is the difference each month between the return on small and big firms, while HML (high minus low) is the monthly difference of the returns on a portfolio of high book-to-market and low book-to-market firms. UMD (up minus down) is the momentum factor computed on a monthly basis as the return differential between a portfolio of winners and a portfolio of losers.

The results are reported in Table IV. When we compare abnormal monthly returns that are obtained from the Carhart (1997) four-factor model in Panel A, we find several important results. First, we determine that the low political strategy group presents high future stock returns when compared to the high political group even though they are in low or high policy uncertainty areas. In addition, the difference in stock returns between the high and low political strategy groups is 33% (1.26/0.95) larger when they are in high policy uncertainty areas suggesting that political connections are beneficial when firms face a high degree of policy risk. Moreover, returns are not different between firms in the low and high policy risk areas when they are highly connected, while the return difference is sizeable and significant when they are less connected (0.46% with a t-statistic of 2.97). This evidence implies that distinctly high stock returns caused by the high costs of capital for non-hedged firms in high policy risk areas make a significant differences in stock returns. Alternatively, highly connected firms successfully hedge policy risk, presenting normal stock returns even if they are dealing with high levels of policy risk.

In addition, we examine the relationship between political strategies and the cost of equity. Similar to other tests, we include policy risk, political strategy, their interaction variables, and the controlling variables in the regression model. We find that a firm's cost of equity is positively associated with policy risk. Their political strategies reduce the cost of capital and the effect is stronger when policy risk is high. Overall, the results in Table IV are compatible with the previous results and suggest that the hedging effect of corporate political strategies is economically large. (11)

C. Political Strategy and Firm's Idiosyncratic Risk

The previous subsection indicates that policy uncertainty's effects on a firm's systematic risk can be alleviated by implementing various corporate political strategies. Next, we investigate whether these corporate political strategies and policy uncertainty also affect firm-specific risk. We define idiosyncratic volatility (IV) as the variance of residuals obtained from the Carhart (1997) four-factor model in which a firm's daily returns are regressed on size, value, and momentum factors along with the market risk premium over the year. We regress IV on all individual political strategy variables previously introduced, as well as policy uncertainty, their interactions, and the set of control variables introduced in the previous table.

The results are reported in Table V. We find that idiosyncratic volatility (IV) is higher among politically active firms consistent with the notion that access to political intelligence generates greater firm-specific performance heterogeneity. (12) Interestingly, the significant interaction terms imply that the aforementioned effect gets stronger with policy uncertainty. Model (3) indicates that the magnitude of the effect on a firm's idiosyncratic volatility generated by political strategies is increased by 16% (0.02/0.14) when there is an increase in Bills by one standard deviation (11.99) from the mean (5.68). We interpret the results presented in Tables II, III, and V as evidence that political connections can entail a shift of the firm's risk landscape from systematic to firm-specific risk.

D. Political Strategy and Real Options

In the previous test, we find that politically connected firms coping with policy uncertainty become more idiosyncratic. We now argue that corporate political strategies can serve as facilitators of valuable real options. In a sense, we posit that political connections accentuating firm heterogeneity can also increase the value relevance of a firm's portfolio of real options. These real options can become potential drivers of competitive advantages in uncertain environments (Trigeorgis, 1996; Trigeorgis and Lambertides, 2014) as political connections can yield intelligence that enhances a firm's operating flexibility.

We employ the approach of Grullon et al. (2012) who hypothesize and confirm that a positive relation between firm stock returns and changes in firm stock volatility is due to a firm's real options. This is the case as firms can change their operating and investment decisions in a way that both mitigates the effects of bad news (reducing the downside) and amplifies the effects of good news (making the best case even better). In Table VI, following the original method of Grullon et al. (2012), we estimate Fama-MacBeth (1973) cross-sectional regressions of a firm's excess returns on the change in volatility, political connection measures, the interacted terms between the change in volatility and political connection measures, and other controlling variables. In Model (1), we confirm the evidence of Grullon et al. (2012) by demonstrating that the change in volatility is positively and significantly related to excess returns. In Model (2), we find a positive relation between political connections and excess returns, consistent with evidence in the prior literature (Cooper et al., 2010). Model (3) includes the political strategy interaction with changes in volatility. Our focus is on this interaction term, which displays a positive and significant coefficient. A one standard deviation increase in residual [PSI.sup.R] (0.19) is associated with a 16% ((1.01*0.19)/1.19) increase in the return's sensitivity to changes in volatility. This result implies that political strategies can boost the value relevance of the firm's real options. In an untabulated test, we confirm that our results remain the same when we use an ordinary least squares (OLS) model.

We also examine whether the aforementioned effects become more pronounced in uncertain policy environments. We use the number of recently introduced relevant bills (Bills) to gauge the level of uncertainty. To determine whether the value relevance of real options changes with the level of uncertainty emanating from legislative activity, we construct two subsamples by classifying firms into low and high policy uncertainty groups depending upon whether they rank in the top or bottom tercile of Bills, respectively.

We then replicate the regression models that include political strategy, changes in stock return volatility and their interactions, as first shown in Table VI, for the two groups and report the results in Table VII. The real option effect always remains significant for the high Bills group, (i.e., when policy uncertainty is high), while it is significant in only two of four regressions using the low Bills subsample. The difference between the interaction coefficients from the high and low Bills groups' regressions is significant at the 5% level in two of four cases. Thus, it seems that the degree to which political connections boost the value relevance of real options is marginally different between periods of high and low legislative activity.

IV. Empirical Identification

The evidence produced from regressing risk measures on the lagged variables of policy risk and political strategies cannot firmly establish causal relationship. Responding to this concern, this section demonstrates the mechanism of political hedging strategies in a dynamic test setting. To operationalize this, we consider political events that could significantly change the impact of a firm's political strategies on stock returns and the value of real options. These events should be random and not endogenously determined. We argue that ex-politicians' sudden deaths that reduce a board's connectedness can meet this requirement. We define Sudden death as a variable with a value of one for the years after an ex-politician on the board suddenly dies and zero for the years before the death.

We revisit our main tests by considering this exogenous shock. Since this event is directly related to a sudden change in political connections through corporate boards, we conduct the test with the residual value of B-index, where the index combines the yearly ranks of N. of connected directors, Board's political experience, and Board's political freshness.

In Panel A of Table VIII, we estimate systematic risk and unsystematic risk in Model (1) and Model (2), respectively. Model (1) reports the consistent patterns found in Table III. Policy risk proxied by the number of bills is negatively related to beta. The interaction between policy risk and B-index is negative implying that political strategy via ex-politicians is an effective hedging mechanism that can reduce the impact of policy uncertainty on systematic risk. The positive and significant coefficient of the triple interaction (Ln([Bills.sub.t]) x Sudden death x [B-index.sup.R]) suggests that a firm's ability to hedge against the impact of policy uncertainty on systematic risk is substantially weakened after the sudden death of an ex-politician serving on the board. In Model (2), we also confirm that an exogenous blow to the board's connectedness reduces the IV impact of political strategies and policy uncertainty. In effect, this implies that active corporate political strategies boost firm heterogeneity.

Panel B examines the value of real options. Consistent with the findings in Table VI, the results indicate that the board-based political connection works as stimulator of real options. Interestingly, this positive effect is weaker after the sudden death of an ex-politician member of the board of directors only when firms face high levels of policy uncertainty. This evidence, once again, supports the notion that causality runs from corporate political strategies to risk and growth opportunities.

V. Robustness

A. Alternative Measures of Policy Uncertainty

In this section, we construct two alternative policy uncertainty measures as a robustness check. The main variable used to proxy for policy risk is the number of bills linked to the industry in which a firm belongs and that are introduced by either home state Senators or home state House Representatives. We decompose it into two broader measures: 1) the bills that are sponsored by the firm's home state politicians ([Bills.sup.loc]) and 2) the bills that target the firm's industry ([Bills.sup.ind]). More specifically, [Bills.sup.loc] is the annual number of bills introduced by either the home state Senators or House Representatives, while ([Bills.sup.ind]) is the annual number of bills linked to the industry in which a firm belongs. Using these broader policy uncertainty measures, we find that the results are generally similar to the findings in the previous tables.

B. Other Robustness Checks

We also conduct various other robustness tests. First, we determine whether our findings are driven by micro-cap stocks. Following prior research (Cohen and Lou, 2012), we exclude those stocks that are priced below $5 a share from our sample. In addition, we recognize that the use of daily returns to calculate beta in our main tests may be problematic as missing observations from non-trading occurrences could affect the beta estimates (Conrad and Kaul, 1988). Accordingly, Conrad and Kaul (1988) choose weekly data as a compromise solution to the twin problems associated with the relatively low number of monthly observations and non-trading occurrences in the daily data. Thus, we calculate beta using weekly returns over the year. Moreover, we exclude any financial and utility firms and retest the model. Finally, we construct a narrower sample by requiring the matching of bills with industries resulting in a loss of about half of the observations in the tests. The results are qualitatively similar to those obtained in our previous tables.

For the sake of brevity, we keep all of the test results out of the paper. The complete set of the results is available from the authors upon request.

VI. Conclusions

The fast growing literature on the links between politics and the financial markets contains considerable evidence that political connections add value to firms. Yet, to date, there has been no study that examines the risk and growth opportunity implications of this relation. Essentially, if political strategies add value, they may either reduce the cost of capital (hedging mechanism) or boost future cash flow expectations (growth opportunities mechanism).

We fill this gap in the literature by first examining the systematic risk impact of the interaction of policy uncertainty induced by politicians' legislative activities and multi-dimensional corporate political strategies involving contributions to political campaigns, adding ex-politicians to their boards of directors, and incurring expenditures for lobbying activities. We find that all of the three aforementioned political strategies, individually and collectively, can serve as effective hedging mechanisms on systematic risk. We demonstrate that the hedging effect of political connections on systematic risk is more pronounced when a firm's business environment is more uncertain.

Interestingly, we determine that firms with political connections actually experience a partial shift of the systematic portion of their risk to the firm-specific portion. The magnitude of the effects on a firm's idiosyncratic volatility generated by policy uncertainty is about seven times larger when a firm's political connectedness strengthens by one standard deviation from the bottom. We also find that in addition to accentuating firm heterogeneity, political connectedness renders firms' portfolio of real options more value relevant. These findings are consistent with the notion that political intelligence obtained through various corporate political activities can improve firms' operating flexibility and act as a potential driver of competitive advantages in uncertain environments.

Overall, the results from our various tests are consistent with the notion that corporate political activities have a dual role. They can be employed as hedging tools that can potentially reduce cost of capital or, alternatively, corporate political activities can boost a firm's growth opportunities and future cash flow expectations.
Appendix A: Political Strategy and Firm Size

This table reports results of OLS regressions that examine the relation
between firm size anc political strategies.

     Dependent    Constant        Ln(Size)       N        Avg. [R.sup.2]
     Variable

(1)  PSI            -0.54 (***)     0.05 (***)   66,059   0.26
                  (-35.18)        (38.45)
(2)  B-index        -0.36 (***)     0.04 (***)   66,059   0.08
                  (-19.79)        (23.22)
(3)  P-index        -0.63 (***)     0.06 (***)   66,059   0.20
                  (-30.03)        (31.37)
(4)  L-index        -0.71 (***)     0.07 (***)   47,607   0.20
                  (-32.48)        (35.58)

(***) ()Significant at the 0.01 level.

Appendix B: Mapping Bills to Fama-French 49 Industry Classification

Major  Major Description      Fama-French 49 Industry Classification

 1     Macroeconomics
 2     Civil rights,
       minority issues,
       and civil liberties
 3     Health                 11   12   13    2     3   4   5
 4     Agriculture             1
 5     Labor, employment,
       and immigration
 6     Education
 7     Environment
 8     Energy                 29   30   31
10     Transportation         41   23   24   25
12     Law, crime,
       and family
       issues
13     Social welfare
14     Community              17   18
       development and
       housing issues
15     Banking, finance, and  45   46   47   48
       domestic commerce
16     Defense                26
17     Space, science,        32   35   36   37    22
       technology and
       communications
18     Foreign trade
19     International
       affairs and
       foreign aid
20     Government
       operations
21     Public lands
       and water
       management
24     State and local
       government
       administration
26     Weather and
       natural
       disasters
27     Fires
28     Arts and                7
       entertainment
29     Sports and              6
       recreation
30     Death notices
31     Churches and
       religion
99     Other,
       miscellaneous,
       and human interest

Appendix C: Variable Definitions

Variables              Definitions

Policy risk
variables
Bills                  The annual number of bills linked to the industry
                       in which a firm belongs and that are introduced
                       by either the home state Senators or House
                       Representatives. The data on bill information are
                       collected from the Congressional Bills Project
                       (http://www.congressionalbills.org/index.html).
Corporate political
strategy
variables
N. of connected        The number of board members who are politically
directors              connected. To be considered as politically
                       connected, the board member's party on the former
                       political position must be same as the incumbent
                       President's party. If a firm does not have any
                       politically connected member, a value of zero is
                       assigned.
Board's political      The average tenure of past political activities
experience             of the boards of directors in a calendar year y.
Board's political      It is computed by subtracting elapsed period from
freshness              50, where the elapsed period is from the year a
                       politically connected director left the political
                       position to the year they serve as a corporate
                       director. After collecting the freshness scores
                       from all directors, we compute the average of the
                       directors' freshness for each firm. Again, we
                       require that the board member's party on the
                       former political position be same as the
                       incumbent President's party to be considered as
                       politically connected.
B-Index                The political strategy index that combines the
                       yearly ranks of N. of connected directors (the
                       number of board members who are politically
                       connected), Board's political experience (the
                       average tenure of past political activities of
                       the boards of directors) and Board's political
                       freshness (the board's political freshness based
                       on the directors' elapses period).
N. of supported        The number of candidates supported by the firm.
candidates             The data comes from the Federal Election
                       Commission (FEC) summary files on political
                       contributions to House and Senate elections. In
                       the regressions, it is transformed by adding one
                       and taking the natural log.
Strength of            The strength of the relationships between
relationships          candidates and the contributing firm. It is
with supported         measured by the total length of relationships
candidates             between the firm and the candidates. The data
                       come from the Federal Election Commission (FEC)
                       summary files on political contributions to House
                       and Senate elections. In the regressions, it is
                       transformed by adding one and taking the natural
                       log. Refer to Cooper et al. (2010) for the
                       detailed description and computation of this
                       variable.
Supported candidates'  The ability of the politicians to help the firm.
ability                It is measured by the home state of the firm and
                       the candidate. The data come from the Federal
                       Election Commission (FEC) summary files on
                       political contributions to House and Senate
                       elections. In the regressions, it is transformed
                       by adding one and taking the natural log. Refer
                       to Cooper et al. (2010) for the detailed
                       description and computation of this variable.
Supported candidates'  The power of the candidates. It is measured by
power                  the candidate's committee ranking. The data come
                       from the Federal Election Commission (FEC)
                       summary files on political contributions to House
                       and Senate elections.
                       In the regressions, it is transformed by adding
                       one and taking the natural log. Refer to Cooper
                       et al. (2010) for the detailed description and
                       computation of this variable.
P-lndex                The political strategy index that combines the
                       yearly ranks of N. of supported candidates (the
                       number of supported candidates), Strength of
                       relationships (the strength of the relationships
                       between candidates and the contributing firm),
                       Supported candidates' ability (the ability of
                       the candidates to help the firm), and Supported
                       candidates' power (the power of the candidates).
Corporate lobbying     It is measured by aggregating all reported
expenditures           expenses. The lobbying information is collected
                       from the OpenSecrets (http://www.opensecrets.org)
                       of the Center for Responsive Politics (CRP).
L-lndex                The political strategy index that measures the
                       yearly rank of Lobbying expenditures.
PSI                    The political strategy index that combines the
                       yearly standardized ranks of N. of politically
                       connected board members, N. of supported
                       candidates, and Lobbying expenditures.
                       [mathematical expression not reproducible], where
                       [Rank.sub.k] (Political [strategy.sub.ik]) is the
                       rank function that assigns rank for each
                       observation. Political [strategy.sub.ik] is the
                       kth measure of political strategy measures for
                       firm i in our sample, and K is the dimensions of
                       the measures. For each information variable, the
                       firm with the highest value in the measure is
                       ranked as [N.sub.k], while the firm with the
                       lowest value is ranked as one. The denominator
                       ([K.sub.i]) averages the ranks regardless as to
                       the number of values of the firm in the sample.
                       For example, the firm that has only two measures
                       in the records is divided by [K.sub.i] = 2. A
                       firm with all three measures is divided by
                       [K.sub.i] = 3. This construction scales the
                       variable PSI to a value between zero (weakest
                       political strategy) and one (strongest political
                       strategy possible).
Firm characteristics
Ret                    A firm i's monthly return.
Size                   The natural log of one plus the market value of
                       common equity that is computed by the number of
                       common shares times the share price at the end of
                       the calendar year.
BM                     The ratio as the book value to the market value
                       of equities for the firm. The market equity value
                       of the firm is the value of all common stock
                       outstanding.
Beta                   Systematic risk computed from the market model
                       using daily returns over the year.
Leverage               A proxy for a firm's leverage measured as total
                       long-term debts divided
                       by assets [(dltt+dlc)/at].
Tangibility            A proxy for a firm's tangibility measured by the
                       net of properties, plant, and equipment divided
                       by assets [ppe/at].
HHI                    Herfindahl index using a firm's sales based on
                       the first three digits of their SIC code.
Firm age               Years since listed in Compustat.
R&D                    A firm's R&D expenditures normalized by assets
                       [xrd/at].
FCF                    Free cash flow normalized by assets
                       [(oibdp---xint--txt--dvp--dvc)/at].
Foreign                Foreign is a dummy that is equal to one if a
                       firm's foreign sales are greater than zero in a
                       given calendar year and zero otherwise.
UNCOV                  A percentage of union coverage in a given
                       four-digit SIC industry code
                       (http://www.unionstats.com)


References

Aggarwal, R.K., F. Meschke, and T.Y. Wang, 2012, "Corporate Political Donations: Investment or Agency?" Business and Politics 14, Article 3.

Alesina, A. and D. Rodrik, 1994, "Distributive Politics and Economic Growth," Quarterly Journal of Economics 109, 465-490.

Asian, H. and Y. Grinstein, 2011, "Political Contributions and CEO Pay," Georgia State University Working Paper.

Bailey, W. and Y.P. Chung, 1995, "Exchange Rate Fluctuations, Political Risk, and Stock Returns: Some Evidence from an Emerging Market," Journal of Financial and Quantitative Analysis 30, 541-561.

Belo, F., Y.D. Gala, and J. Li, 2013, "Government Spending, Political Cycles and the Cross Section of Stock Returns," Journal of Financial Economics 107, 305-324.

Bialkowski, J., K. Gottschalk, and T.P. Wisniewski, 2008, "Stock Market Volatility Around National Elections," Journal of Banking and Finance 32, 1941-1953.

Boubakri, N., O. Guedhami, D. Mishra, and W. Saffar, 2012, "Political Connections and the Cost of Equity Capital," Journal of Corporate Finance 18, 541-559.

Boutchkova, M., H. Doshi, A. Durnev, and A. Molchanov, 2012, "Precarious Politics and Return Volatility," Review of Financial Studies 25, 1111-1154.

Bradley, D.J., C. Pantzalis, and X. Yuan, 2016, "Policy Risk, Corporate Political Strategies, and the Cost of Debt," Journal of Corporate Finance 40, 254-275.

Bulan, L.T., 2005, "Real Options, Irreversible Investment and Firm Uncertainty: New Evidence from US Firms," Review of Financial Economics 14, 255-279.

Campbell, J.Y., M. Lettau, B.G. Malkiel, and Y. Xu, 2001, "Have Individual Stocks Become More Volatile? An Empirical Exploration of Idiosyncratic Risk," Journal of Finance 56, 1-43.

Carhart, M.M., 1997, "On Persistence in Mutual Fund Performance," Journal of Finance 52, 57-82.

Chaney, P.K., M. Faccio, and D. Parsley, 2011, "The Quality of Accounting Information in Politically Connected Firms," Journal of Accounting and Economics 51, 58-76.

Chen, C.J.P., Y. Ding, and C. Kim, 2010, "High-Level Politically Connected Firms, Corruption, and Analyst Forecast Accuracy Around the World," Journal of International Business Studies 41, 1505-1524.

Chen, H., D.C. Parsley, and Y.-W. Yang, 2015, "Corporate Lobbying and Firm Performance," Journal of Business Finance and Accounting 42, 444-481.

Chun, H., J.-W. Kim, and R. Morck, 2011, "Varying Heterogeneity Among US Firms: Facts and Implications," Review of Economics and Statistics 93, 1034-1052.

Chun, H., J.-W. Kim, R. Morck, and B. Yeung, 2008, "Creative Destruction and Firm-Specific Performance Heterogeneity," Journal of Financial Economics 89, 109-135.

Cohen, L., J. Coval, and C. Malloy, 2011, "Do Powerful Politicians Cause Corporate Downsizing?" Journal of Political Economy 119, 1015-1060.

Cohen, L., K. Diether, and C. Malloy, 2013, "Legislating Stock Prices," Journal of Financial Economics 110, 574-595.

Cohen, L. and D. Lou, 2012, "Complicated Firms," Journal of Financial Economics 104, 383-400.

Conrad, J. and G. Kaul, 1988, "Time-variation in Expected Returns f Journal of Business 61, 409-425.

Cooper, M.J., H. Gulen, and A.V. Ovtchinnikov, 2010, "Corporate Political Contributions and Stock Returns," Journal of Finance 65, 687-724.

Croce, M.M., H. Kung, T.T. Nguyen, and L. Schmid, 2012, "Fiscal Policies and Asset Prices," Review of Financial Studies 25, 2635-2672.

Croci, E. C., Pantzalis, J.C. Park, and D. Petmezas, 2017, "The Role of Corporate Political Strategies in M&As," Journal of Corporate Finance 43, 260-287.

Drutman, L., 2015, The Business of America is Lobbying: How Corporations Become Politicized and Politics Became More Corporate, Oxford, UK, Oxford University Press.

Duffee, G.R., 1995, "Stock Returns and Volatility: A Firm-Level Analysis," Journal of Financial Economics 37, 399-420.

Faccio, M., 2006, "Politically Connected Firms," The American Economic Review 96, 369-386.

Fama, E.F. and J. MacBeth, 1973, "Risk, Return and Equilibrium: Empirical Tests," Journal of Political Economy 81, 607-636.

Ferguson, M.F. and H.D. Witte, 2006, "Congress and the Stock Market," University of Cincinnati Working Paper.

Francis, B.B., I. Hasan, and X. Sun, 2009, "Political Connections and the Process of Going Public: Evidence from China," Journal of International Money and Finance 28, 696-719.

Frank, M.Z. and T. Shen, 2016, "Investment and the Weighted Average Cost of Capital," Journal of Financial Economics 119, 300-315.

Fuss, R. and M.M. Bechtel, 2008, "Partisan Politics and Stock Market Performance: The Effect of Expected Government Partisanship on Stock Returns in the 2002 German Federal Election," Public Choice 135, 131-150.

Goldman, E., J. Rocholl, and J. So, 2009, "Do Politically Connected Boards Affect Firm Value?" Review of Financial Studies 22, 2331-2360.

Goldman, E., J. Rocholl, and J. So, 2013, "Politically Connected Boards of Directors and the Allocation of Procurement Contracts," Review of Finance 17, 1617-1648.

Grullon, G., E. Lyandres, and A. Zhdanov, 2012, "Real Options, Volatility, and Stock Returns," Journal of Finance 67, 1499-1537.

Hamada, R.S., 1972, "The Effect of the Firm's Capital Structure on the Systematic Risk of Common Stocks," Journal of Finance 27, 435-452.

Herron, M.C., 2000, "Estimating the Economic Impact of Political Party Competition in the 1992 British Election," American Journal of Political Science 44, 326-337.

Hill, M.D., G.W. Kelly, G.B. Lockhart, and R.A. Van Ness, 2013, "Determinants and Effects of Corporate Lobbying," Financial Management 42, 931-957.

Hong, G. and S. Sarkar, 2007, "Equity Systematic Risk (Beta) and Its Determinants," Contemporary Accounting Research 24, 423-466.

Houston, J.F., L. Jiang, C. Lin, and Y. Ma, 2014, "Political Connections and the Cost of Bank Loans," Journal of Accounting Research 52(1), 193-243.

Jayachandran, S., 2006, "The Jeffords Effect," Journal of Law and Economics 49, 397-425.

Kang, J.-K. and L. Zhang, 2015, "From Backroom to Boardroom: Role of Government Directors in US Public Firms and Their Impact on Performance," Nanyang Technological University Working Paper.

Kim, C., C. Pantzalis, and J.C. Park, 2012, "Political Geography and Stock Returns: The Value and Risk Implications of Proximity to Political Power," Journal of Financial Economics 106, 196-228.

Kim, C. and L. Zhang, 2016, "Corporate Political Connections and Tax Aggressiveness," Contemporary Accounting Research 33, 78-114.

Knight, B., 2006, "Are Policy Platforms Capitalized into Equity Prices? Evidence from the Bush/Gore 2000 Presidential Election," Journal of Public Economics 90, 751-773.

Leblang, D. and B. Mukherjee, 2005, "Government Partisanship, Elections, and the Stock Market: Examining American and British Stock Returns, 1930-2000," American Journal of Political Science 49, 780-802.

Leahy, J. and T. Whited, 1996, "The Effect of Uncertainty on Investment: Some Stylized Facts," Journal of Money, Credit and Banking 28, 64-83.

Lieberman, M.B. and D.B. Montgomery, 1988, "First-Mover Advantages," Strategic Management Journal 9, 41-58.

Malkiel, B.G., 1979, "The Capital Formation Problem in the United States," Journal of Finance 34, 291-306.

Mazzucato, M. and M. Tancioni, 2008, "Innovation and Idiosyncratic Risk: An Industry- and Firm-Level Analysis," Industrial and Corporate Change 17, 779-811.

McGillivray, E., 2003, "Redistributive Policies and Stock Price Dispersion," British Journal of Political Science 33, 367-396.

Ovtchinnikov, A.V., S.W. Reza, and Y. Wu, 2015, "Political Activism and Firm Innovation," HEC Paris Working Paper.

Pastor, L. and P. Veronesi, 2012, "Uncertainty About Government Policy and Stock Prices," Journal of Finance 67, 1219-1264.

Roberts, B.E., 1990, "A Dead Senator Tells No Lies: Seniority and the Distribution of Federal Benefits," American Journal of Political Science 34, 31-58.

Schumpeter, J.A., 1934, "Theorie der wirtschaftlichen Entwicklung, Leipzig, Dunker Und Humbolt," in R. Opie, Translated by The Theory of Economic Development: An Inquiry into Profits, Capital, Credit, Interest, and the Business Cycle, Cambridge, MA, Harvard University Press.

Shiller, R.J., 2000, Irrational Exuberance, Princeton, NJ, Princeton University Press.

Shon, J.J., 2010, "Do Stock Returns Vary with Campaign Contributions? The Bush versus Gore 2000 Presidential Elections," Economics and Politics 22, 257-281.

Subrahmanyam, M.G. and S.B. Thomadakis, 1980, "Systematic Risk and the Theory of the Firm," Quarterly Journal of Economics 94, 437-451.

Thompson, S.B., 2011, "Simple Formulas for Standard Errors that Cluster by Both Firm and Time," Journal of Financial Economics 99, 1-10.

Trigeorgis, L., 1993, "Real Options and Interactions with Financial Flexibility," Financial Management 22, 202-224.

Trigeorgis, L., 1996, Real Options: Managerial Flexibility and Strategy in Resource Allocation, Cambridge, MA, MIT Press.

Trigeorgis, L. and N. Lambertides, 2014, "The Role of Growth Options in Explaining Stock Returns," Journal of Financial and Quantitative Analysis 49, 749-771.

Wellman, L., 2017, "Mitigating Political Uncertainty," Review of Accounting Studies 22, 217-250.

Whited, H., 1980, "A Heteroskedasticity-consistent Covariance Matrix Estimator and a Direct Test for Heteroskedasticity," Econometrica 48, 817-838.

Chansog (Francis) Kim, Incheol Kim, Christos Pantzalis, and Jung Chul Park (*)

We are especially grateful to Rajkamal Iyer (Editor) and an anonymous reviewer for their many insightful and constructive suggestions. We thank Tao Shen for sharing his data used in Frank and Shen (2016). We also appreciate helpful comments from participants at the 2013 Financial Management Association Meeting.

(*) Chansog (Francis) Kim is an Associate Professor of Accounting in the College of Business at Stony Brook University in Stony Brook. NY. Incheol Kim is an Assistant Professor of Finance in the Robert C. Vackar College of Business and Entrepreneurship at The University of Texas Rio Grande Valley in Edinburg, TX. Christos Pantzalis is a Professor in the Finance Department in the Muma College of Business, at the University of South Florida in Tampa. FL. Jung Chul Park is an Assistant Professor in the Finance Department in the Muma College of Business, at the University of South Florida in Tampa, FL.

(1) Malkiel (1979) first argued that Congressional activity as a proxy for regulatory uncertainty can hamper economic performance. He hypothesized that investors viewed greater Congressional activity as increasing regulatory uncertainty and that this greater uncertainty would be reflected in higher return volatility and lower returns. Ferguson and Witte (2006) provide support for this hypothesis by demonstrating that stock returns are dramatically lower and volatility higher when Congress is in session. Moreover, they find that more than 90% of the capital gains over the life of the DJIA have come on days when Congress is out of session.

(2) This widely-held view that connected firms possess "inside" political information, while non-connected firms do not, is also supported by anecdotal evidence, such as the one from an article that appeared on Bloomberg Markets (http://www.bloomberg.com/news/articles/2011-11-29/how-henry-paulson-gave-hedge-funds-advance-word-of-2008-fannie-mae-rescue). The article describes that Hank Paulson, in a private meeting with big investors at Eton Park including several fellow Goldman Sachs alumni, revealed how he would nationalize Fannie and Freddie and wipe out shareholders, while at the same time telling the public, via the NY Times and Congress, that this was not going to happen.

(3) The results in Chun et al. (2008) support the notion that technological improvements can induce innovation across many industries wherein some firms can end up as winners, while others as losers depending upon how well they exploit opportunities. They view firm performance heterogeneity as a readily observable measure of ongoing creative destruction, a process that Schumpeter (1934) argues sustains economic growth.

(4) Asian and Grinstein (2011) find that politically connected chief executive officers (CEOs) receive higher compensation packages than their non-connected peers. Chaney et al. (2011), in their cross-country study, determine that the quality of earnings reported by politically connected firms is significantly poorer than that of similar non-connected companies. Kim and Zhang (2016) confirm that politically connected firms are more tax aggressive than non-connected firms suggesting more managerial rent diversion for connected firms. Chen et al. (2010) find that analyst forecasts are less accurate for politically connected firms than for non-connected firms implying more exacerbated information asymmetry problems for politically connected firms.

(5) Chun et al. (2011) find that firms in US industries that experienced a propagation of a new "general purpose technology" (GPT) would be characterized by higher levels of firm-specific heterogeneity, reflected in higher firm-specific volatility. They test and find support for the argument that high idiosyncratic volatility reflects a wave of IT-driven creative destruction (Schumpeter, 1934) that exacerbates the range between those who adopted IT optimally (winners) and those that did not (losers). Although corporate political strategies cannot be regarded as GPTs per se, their rate of implementation and their importance for firms' ability to grow has increased tremendously over the past 30 years (Drutman, 2015). Thus, it is conceivable that industries may be experiencing creative destruction-like effects emanating from corporate political strategies. In this context, the emergence of corporate political strategies should be accompanied by a rise in firm-specific heterogeneity with a wider gap between firms with investment in political capital and firms with no such investment in connectedness.

(6) The political connections allow firms to better assess whether the advantages outweigh the costs associated with being a "first mover" into a new market or with respect to a new opportunity that may arise from the resolution of political uncertainty. Lieberman and Montgomery (1988) identify first mover advantages (e.g., technological leadership, preemption of rivals, and the imposition of switching costs on buyers) and disadvantages (e.g., the ability of rivals to free ride on pioneers, the resolution in the market of technological or market uncertainty, and technological discontinuities that make early investments obsolete).

(7) Two notes in this regard. First, the use of stock return volatility as a proxy for underlying business volatility is in line with Leahy and Whited (1996) who argue that stock price returns capture the effects of any aspect of a firm's environment that investors deem important. It could be problematic to measure underlying business volatility directly in terms of identifying the most important sources of uncertainty and not least measuring such uncertainty. In addition, stock return volatility measures equity risk and not overall firm risk. However, a stock is an option on the firm's assets and, as such, its value is sensitive to the volatility of the underlying asset. This justifies the use of stock return volatility as a proxy for the volatility of the value of the firm (Bulan, 2005).

(8) The lists of politicians cover historical information on the US president, vice president, candidates, secretaries of departments (e.g., Secretaries of State, Treasury, and Defense, etc.), governors, senators and house representatives, attorney generals. White House executives, SEC commissioners, and ambassadors, as well as assistant and deputy secretaries of all departments. The references and sources that we used are as follows: for the US president (http://en.wikipedia.org/wiki/President_of_the_United_States); for the US House of Representatives (http://www.house.gov/); for the U.S. senators (http://www.senate.gov/); for the secretaries of departments (e.g., Secretary of Defense (http://en.wikipedia.org/wiki/United_States_Secretary_of_Defense) and Secretary of the Treasury (http://en.wikipedia.org/wiki/United_States_Secretary_of_the_Treasury), etc.

(9) A popular data source for directors' background information is BoardEx. However, we have some concerns about BoardEx data coverage. BoardEx began to cover only the S&P 1500 firms and their board members from 1999 to the early 2000s. Information on companies' senior level directors has been included in BoardEx data since 2006. In 2008, BoardEx has extended its coverage beyond the S&P 1500, but it does not backfill data.

(10) Since our measures of board connectedness require that the connected directors' political party be the same as the incumbent President's party, they are somewhat more narrowly-defined than those used in other studies (Faccio, 2006; Goldman et al., 2009; Goldman, Rocholl, and So, 2013; Kim and Zhang, 2016).

(11) We would like to thank an anonymous referee for suggesting the test on expected returns.

(12) This, in turn is, consistent with political intelligence playing the role of innovation wherein firms' ability to exploit opportunities hinges on how well it is employed. Thus, in a creative-destruction framework (Chun et al., 2008, 2011), the proliferation of political intelligence across each industry should be manifested in higher firm-specific heterogeneity (IV).
Table I. Descriptive Statistics

This table provides the descriptive statistics for the sample of 66,059
firm-year observations. Bills is the number of bills linked to the
industry in which a firm belongs and that are introduced by either the
home state Senators or House Representatives. N. of connected directors
is the number of board members who are politically connected. Board's
political experience is the average tenure of past political activities
of the boards of directors. Board's political freshness is the board's
political freshness based on the directors' elapse period. B-index is
the political strategy index that combines the yearly ranks of N. of
connected directors, Board's political experience, and Board's
political freshness. N. of supported candidates is the number of
supported candidates. [B-index.sup.R] is the residual value of B-index.
Strength of relationships is the strength of the relationships between
the candidates and the contributing firm. Supported candidates' ability
is the ability of the candidates to help the firm. Supported candidates
'power is the power of the candidates. P-index is the political
strategy index that combines the yearly ranks of N. of supported
candidates, Strength of relationships, Supported candidates' ability,
and Supported candidates' power. [P-index.sup.R] is the residual value
of P-index. Lobbying expenditures (thousand $) is the corporate total
lobbying expenditures. L-index is the political strategy index that
measures the yearly ranks of Lobbying expenditures. [L-index.sup.R] is
the residual value of L-index. PSI is the political strategy index that
combines B-index, P-index, and L-index. [PSI.sup.R] is the residual
value of PSI. Beta is systematic risk computed from the market model
using daily returns over the year. Size is the market value of equity
at the end of year y. BM is the ratio of the book-to-market value of
equity. Leverage is the ratio of debts to assets. ROA is net income
divided by assets. HHI is the Herfindahl index based on the sale of the
first three digits of SIC code. Tangibility is the ratio of properties,
plant, and equipment divided by assets. Firm age is the years since a
firm is listed in Compustat. Volume is monthly trading divided by
shares outstanding. [Ret.sub.(t-1,t-12)] is the cumulative returns from
Month -12 to Month -1. R&D is a firm's R&D expenditures divided by
assets. FreeCash is free cash flow normalized by assets. Foreign is a
dummy that is equal to one if a firm's foreign sales is greater than
zero in a given calendar year and zero otherwise. Union is the
percentage of union membership in a given four-digit SIC industry code.
Refer to Appendix C for detailed variable descriptions.

Variable Name              Obs.     Mean       Median     Std.     Min

Policv Risk
Bills                      66,059       5.68     0.00      11.99    0.00
Political Strategy
N. of connected directors  66,059       0.14     0.00       0.46    0.00
Board's political          64,100       0.43     0.00       2.07    0.00
experience
Board s political          63,935       2.90     0.00      10.27    0.00
freshness
B-index                    66,059       0.10     0.00       0.28    0.00
[B-index.sup.R]            66,059       0.00    -0.07       0.27   -0.39
N. of supported            66,059       9.63     0.00      43.32    0.00
candidates
Strength of relationships  66,059     561.72     0.00   6,448.42    0.00
Supported candidates'      66,059       3.93     0.00     104.35    0.00
ability
Supported candidates'      66,059       1.26     0.00      10.98    0.00
power
P-index                    66,059       0.10     0.00       0.29    0.00
[P-index.sup.R]            66,059       0.00    -0.05       0.26   -0.54
Lobexp (thousand $)        47,607     125.92     0.00     825.62    0.00
L-index                    47,607       0.15     0.00       0.34    0.00
[L-index.sup.R]            47,607       0.01    -0.08       0.30   -0.60
PSI                        66,059       0.11     0.00       0.23    0.00
[PSI.sup.R]                66,059       0.00    -0.04       0.19   -0.47
Firm Characteristics
Beta                       66,059       0.76     0.69       0.63   -0.54
IV                         66,059       0.19     0.09       0.60    0.00
[Cost.sub.equity]          40,129       0.15     0.14       0.09    0.00
Size (million $)           66,059   1,775.49   181.03    5610.53    2.53
BM                         66,059       0.63     0.51       0.59   -0.83
Leverage                   66,059       0.22     0.17       0.21    0.00
ROA                        66,059       0.00     0.05       0.23   -1.17
Tangibility                66,059       0.24     0.16       0.24    0.00
HHI                        66,059       0.13     0.10       0.12    0.02
Firm Age                   66,059      16.81    12.00      13.41    0.00
Volume                     66,059       0.67     0.57       0.47    0.05
[Ret.sub.(t-1,t-12)]       66,059       0.13     0.03       0.67   -0.88
R&D                        66,059       0.05     0.00       0.11    0.00
FreeCash                   60,644      -0.05     0.02       0.23   -1.24
Foreign                    66,059       0.31     0.00       0.46    0.00
Union                      66,059       0.06     0.02       0.09    0.00

Variable Name               Max

Policv Risk
Bills                           117.00
Political Strategy
N. of connected directors         7.00
Board's political                39.00
experience
Board s political                50.00
freshness
B-index                           1.00
[B-index.sup.R]                   1.05
N. of supported                 766.00
candidates
Strength of relationships   725,069.63
Supported candidates'        12,617.31
ability
Supported candidates'           532.24
power
P-index                           1.00
[P-index.sup.R]                   1.10
Lobexp (thousand $)          29,368.50
L-index                           1.00
[L-index.sup.R]                   1.09
PSI                               0.99
[PSI.sup.R]                       0.96
Firm Characteristics
Beta                              2.61
IV                               66.79
[Cost.sub.equity]                 0.57
Size (million $)             41,682.34
BM                                3.32
Leverage                          0.94
ROA                               0.35
Tangibility                       0.89
HHI                               0.72
Firm Age                         58.00
Volume                            2.12
[Ret.sub.(t-1,t-12)]              3.37
R&D                               0.65
FreeCash                          0.23
Foreign                           1.00
Union                             0.40

Table II. Political Strategy Index and Firm Risk

This table reports the estimated coefficients of the cross-sectional
regressions where dependent variable, Beta, is systematic risk computed
from the market model using daily returns over the year. Bills is the
number of bills linked to the industry in which a firm belongs and that
are introduced by either the home state Senators or House
Representatives. [PSI.sup.R] is the residual value from the regression
of PSI on firm size, where PSI is the annual corporate political
strategy index. Refer to Appendix C for detailed variable descriptions.
Year and industry dummies are included, but the coefficients are
omitted for brevity. Numbers in parentheses are t-statistics computed
using standard errors that are clustered at both the firm and year
level.

                       (1)               (2)
                       Dependent variable: [Beta.sub.t]

Ln([Bills.sub.t])           0.03 (***)
                           (2.97)
[PSI.sup.R.sub.t]                            -0.23 (***)
                                            (-5.72)
Ln([Bills.sub.t]) (*)
[PSI.sup.R.sub.t]
Ln[(Size).sub.t]            0.14 (***)           0.14 (***)
                          (22.97)           (22.28)
[BM.sub.t]                 -0.02             -0.02
                          (-1.49)           (-1.06)
[Leverage.sub.t]           -0.01              0.01
                          (-0.22)            (0.22)
[ROA.sub.t]                -0.30 (***)       -0.32 (***)
                          (-4.81) (***)     (-5.15)
[Tangibility.sub.t]        -0.15             -0.16 (***)
                          (-4.07)           (-4.41)
[HHI.sub.t],               -0.11 (*)         -0.11 (*)
                          (-1.86)           (-1.86)
Ln[(Firm Age).sub.i]       -0.13 (***)       -0.12 (***)
                         (-10.65)          (-10.53)
Constant                   -0.57 (***)       -0.59 (***)
                          (-4.63)           (-4.47)
Year fixed effects     YES               YES
Industry fixed         YES               YES
effects
Obs.                   66,076            66,076
Adj. [R.sup.2]              0.31              0.31

                       Dependent variable: [Beta.sub.t]
                         (3)

Ln([Bills.sub.t])           0.03 (***)
                           (2.81)
[PSI.sup.R.sub.t]          -0.17 (***)
                          (-4.28)
Ln([Bills.sub.t]) (*)      -0.05 (***)
[PSI.sup.R.sub.t]
                          (-2.77)
Ln[(Size).sub.t]            0.14 (***)
                          (22.10)
[BM.sub.t]                 -0.02
                          (-1.09)
[Leverage.sub.t]            0.01
                           (0.21)
[ROA.sub.t]                -0.32 (***)
                          (-5.10)
[Tangibility.sub.t]        -0.15 (***)
                          (-4.34)
[HHI.sub.t],               -0.11 (*)
                          (-1.89)
Ln[(Firm Age).sub.i]       -0.12 (***)
                         (-10.47)
Constant                   -0.62 (***)
                          (-4.81)
Year fixed effects       YES
Industry fixed           YES
effects
Obs.                   66,076
Adj. [R.sup.2]              0.31

(***) Significant at the 0.01 level.
(*) Significant at the 0.10 level.

Table III. Individual Political Strategy and Firm Risk

This table reports the estimated coefficients of the cross-sectional
regressions where the dependent variable, Beta, is systematic risk,
computed from the market model using daily returns over the year. Bills
is the number of bills linked to the industry in which a firm belongs
and that are introduced by either the home state Senators or House
Representatives. [B-index.sup.R] is the residual value of the political
strategy index that combines the yearly ranks of N. of connected
directors, Board's political experience, and Board's political
freshness. [P-index.sup.R] is the residual value of the political
strategy index that combines the yearly ranks of N. of supported
candidates, Strength of relationships, Supported candidates' ability,
and Supported candidates' power. [L-index.sup.R] is the residual value
of the political strategy index that measures the yearly ranks of
Lobbying expenditures. Refer to Appendix C for detailed variable
descriptions. Firm characteristic-related control variables, year, and
industry dummies are included, but the coefficients are omitted for
brevity. Numbers in parentheses are t-statistics computed using
standard errors that are clustered at both the firm and year level.

                         (1)              (2)
                           Dependent variable: [Beta.sub.t]

Ln[(Bills).sub.t]             0.02 (***)       0.02 (***)
                             (2.80)           (2.84)
[B-index.sup.R.sub.t]        -0.08 (***)      -0.05 (**)
                            (-3.95)          (-2.11)
Ln[(Bills).sub.t]                             -0.03 (***)
x [B-index.sup.R.sub.t]                      (-2.65)
[P-index.sup.R.sub.t]
Ln[(Bills).sub.t] x
[P-index.sup.R.sub.t]
[L-index.sup.R.sub.t]
Ln[(Bills).sub.t]
x [L-index.sup.R.sub.t]
Controls                    YES             YES
Year fixed effects          YES             YES
Industry fixed effects      YES             YES
Obs.                     66,059           66,059
Adj. [R.sup.2]                0.31             0.31

                         (3)             (4)
                           Dependent variable: [Beta.sub.t]

Ln[(Bills).sub.t]             0.02 (**)        0.02 (**)
                             (2.55)           (2.40)
[B-index.sup.R.sub.t]
Ln[(Bills).sub.t]
x [B-index.sup.R.sub.t]
[P-index.sup.R.sub.t]        -0.21 (***)      -0.15 (***)
                            (-7.55)          (-5.65)
Ln[(Bills).sub.t] x                           -0.07 (***)
[P-index.sup.R.sub.t]                        (-4.21)
[L-index.sup.R.sub.t]
Ln[(Bills).sub.t]
x [L-index.sup.R.sub.t]
Controls                    YES             YES
Year fixed effects          YES             YES
Industry fixed effects      YES             YES
Obs.                     66,059           66,059
Adj. [R.sup.2]                0.31             0.31

                         (5)              (6)
                           Dependent variable: [Beta.sub.t]

Ln[(Bills).sub.t]             0.02 (*)         0.02 (*)
                             (1.86)           (1.86)
[B-index.sup.R.sub.t]
Ln[(Bills).sub.t]
x [B-index.sup.R.sub.t]
[P-index.sup.R.sub.t]
Ln[(Bills).sub.t] x
[P-index.sup.R.sub.t]
[L-index.sup.R.sub.t]        -0.11 (***)      -0.07 (***)
                            (-5.56)          (-2.97)
Ln[(Bills).sub.t]                             -0.03 (**)
x [L-index.sup.R.sub.t]                      (-2.12)
Controls                   YES              YES
Year fixed effects         YES              YES
Industry fixed effects     YES              YES
Obs.                     47,607           47,607
Adj. [R.sup.2]                0.35             0.35

(***) Significant at the 0.01 level.
(**) Significant at the 0.05 level.
(*) Significant at the 0.10 level.

Table IV. Economic Effect of Political Strategy

Panel A reports the estimated intercept (alpha) coefficients in the
time-series tests of the four-factor models (Carhart, 1997). The sample
includes 180 monthly observations spanning from January 1994-December
2008. Policy risk is the total number of bills introduced by the home
state politicians in a given year. A firm is classified into the low
(high) policy risk group if the total number of bills in the firm's
state is lower (higher) than the median value in year y. A firm is
classified into the low (high) political strategy group if the firm's
political strategy index is lower (higher) than the median value in
year y. Panel B reports the OLS results where dependent variable is the
cost of equity (Frank and Shen, 2016). [PSI.sup.R], [B-index.sup.R],
[P-index.sup.R], and [L-index.sup.R] are the residual value of PSI,
B-index, P-index, and L-index, respectively. Refer to the appendix for
detailed variable descriptions.

                Panel A. Abnormal Returns (%) from the Four-Factor Model
                   High Political   Low Political   Low--High
                   Strategy         Strategy

High policy risk     0.03            1.29 (***)      1.26 (***)
                    (0.14)          (9.80)          (9.50)
Low policy risk     -0.13            0.82 (***)      0.95 (***)
                   (-0.78)          (7.17)          (6.08)
High--Low            0.15            0.46 (***)
                    (1.24)          (2.97)

                               Panel B. Cost of Equity
                         (1)                  (2)
                         Dependent variable: [Cost.sub.Equity]

Ln[(Bills).sub.t]             0.00 (***)          0.00 (***)
                             (3.13)              (3.04)
[PSI.sup.R.sub.t]                                -0.03 (***)
                                                (-5.65)
Ln([Bills.sub.t]) (*)
[PSI.sup.R.sub.t]
[B-index.sup.R.sub.t]
Ln[(Bills).sub.t]
x [B-index.sup.R.sub.t]
[P-index.sup.R.sub.t]
Ln[(Bills).sub.t] x
[P-index.sup.R.sub.t]
[L-index.sup.R.sub.t]
Ln[(Bills).sub.t] x
[[L-index.sup.R].sub.t]
Ln[(Size).sub.t]              0.02 (***)          0.02 (***)
                            (15.60)             (15.48)
[BM.sub.t]                   -0.01 (***)         -0.01 (***)
                            (-4.80)             (-4.48)
[Leverage.sub.t]             -0.01               -0.01
                            (-1.49)             (-0.96)
[ROA.sub.t]                  -0.09 (***)         -0.09 (***)
                            (-4.83)             (-5.05)
[Tangibility.sub.t]          -0.02 (***)         -0.02 (***)
                            (-2.98)             (-3.12)
[HHI.sub.t]                  -0.03 (***)         -0.03 (***)
                            (-3.40)             (-3.39)
Ln[(Firm Age).sub.t]         -0.02 (***)         -0.02 (***)
                           (-12.28)            (-12.40)
Constant                      0.02                0.01
                             (1.29)              (0.63)
Year fixed effects          YES                 YES
Industry fixed effects      YES                 YES
Observations             40,129              40,129
Adj. [R.sup.2]                0.32                0.32

                             Panel B. Cost of Equity
                            (3)               (4)
                            Dependent variable:
                            [Cost.sub.Equity]

Ln[(Bills).sub.t]             0.00 (***)        0.00 (***)
                             (2.99)            (3.13)
[PSI.sup.R.sub.t]            -0.02 (***)
                            (-3.58)
Ln([Bills.sub.t]) (*)        -0.01 (***)
[PSI.sup.R.sub.t]
                            (-4.49)
[B-index.sup.R.sub.t]                          -0.01 (***)
                                              (-4.49)
Ln[(Bills).sub.t]
x [B-index.sup.R.sub.t]
[P-index.sup.R.sub.t]
Ln[(Bills).sub.t] x
[P-index.sup.R.sub.t]
[L-index.sup.R.sub.t]
Ln[(Bills).sub.t] x
[[L-index.sup.R].sub.t]
Ln[(Size).sub.t]              0.02 (***)        0.02 (***)
                            (15.51)           (15.50)
[BM.sub.t]                   -0.01 (***)       -0.01 (***)
                            (-4.47)           (-4.72)
[Leverage.sub.t]             -0.01             -0.01
                            (-0.97)           (-1.38)
[ROA.sub.t]                  -0.09 (***)       -0.09 (***)
                            (-5.03)           (-4.92)
[Tangibility.sub.t]          -0.02 (***)       -0.02 (***)
                            (-3.19)           (-3.04)
[HHI.sub.t]                  -0.03 (***)       -0.03 (***)
                            (-3.34)           (-3.38)
Ln[(Firm Age).sub.t]         -0.02 (***)       -0.02 (***)
                           (-12.39)          (-12.44)
Constant                      0.0102            0.0188
                             (0.57)            (1.10)
Year fixed effects          YES               YES
Industry fixed effects      YES               YES
Observations             40,129            40,129
Adj. [R.sup.2]                0.32              0.32

                               Panel B. Cost of Equity
                               (5)            (6)
                            Dependent variable:
                            [Cost.sub.Equity]

Ln[(Bills).sub.t]             0.00 (***)        0.00 (***)
                             (3.10)            (3.02)
[PSI.sup.R.sub.t]
Ln([Bills.sub.t]) (*)
[PSI.sup.R.sub.t]
[B-index.sup.R.sub.t]        -0.01 (***)
                            (-2.83)
Ln[(Bills).sub.t]            -0.01 (***)
x [B-index.sup.R.sub.t]
                            (-2.79)
[P-index.sup.R.sub.t]                          -0.02 (***)
                                              (-5.51)
Ln[(Bills).sub.t] x
[P-index.sup.R.sub.t]
[L-index.sup.R.sub.t]
Ln[(Bills).sub.t] x
[[L-index.sup.R].sub.t]
Ln[(Size).sub.t]              0.02 (***)        0.02 (***)
                            (15.49)           (15.69)
[BM.sub.t]                   -0.01 (***)       -0.01 (***)
                            (-4.70)           (-4.54)
[Leverage.sub.t]             -0.01             -0.01
                            (-1.37)           (-1.09)
[ROA.sub.t]                  -0.09 (***)       -0.09 (***)
                            (-4.92)           (-4.95)
[Tangibility.sub.t]          -0.02 (***)       -0.02 (***)
                            (-3.06)           (-2.87)
[HHI.sub.t]                  -0.03 (***)       -0.03 (***)
                            (-3.38)           (-3.35)
Ln[(Firm Age).sub.t]         -0.02 (***)       -0.02 (***)
                           (-12.43)          (-12.27)
Constant                      0.02              0.01
                             (1.07)            (0.81)
Year fixed effects          YES               YES
Industry fixed effects      YES               YES
Observations             40,129            40,129
Adj. [R.sup.2]                0.32              0.32

                             Panel B. Cost of Equity
                             (7)             (8)
                             Dependent variable:
                             [Cost.sub.Equity]

Ln[(Bills).sub.t]             0.00 (***)       0.01 (***)
                             (2.89)           (2.65)
[PSI.sup.R.sub.t]
Ln([Bills.sub.t]) (*)
[PSI.sup.R.sub.t]
[B-index.sup.R.sub.t]
Ln[(Bills).sub.t]
x [B-index.sup.R.sub.t]
[P-index.sup.R.sub.t]        -0.01 (***)
                            (-3.01)
Ln[(Bills).sub.t] x          -0.01 (***)
[P-index.sup.R.sub.t]
                            (-4.41)
[L-index.sup.R.sub.t]                         -0.02 (***)
                                              -5.66)
Ln[(Bills).sub.t] x
[[L-index.sup.R].sub.t]
Ln[(Size).sub.t]              0.02 (***)       0.02 (***)
                            (15.75)          (12.82)
[BM.sub.t]                   -0.01 (***)      -0.01 (***)
                            (-4.54)          (-3.71)
[Leverage.sub.t]             -0.01            -0.01
                            (-1.10)          (-0.80)
[ROA.sub.t]                  -0.09 (***)      -0.11 (***)
                            (-4.93)          (-5.18)
[Tangibility.sub.t]          -0.02 (***)      -0.02 (***)
                            (-2.89)          (-1.94)
[HHI.sub.t]                  -0.03 (***)      -0.03 (***)
                            (-3.31)          (-3.09)
Ln[(Firm Age).sub.t]         -0.02 (***)      -0.02 (***)
                           (-12.28)          (-9.94)
Constant                      0.01            -0.03
                             (0.74)          (-1.55)
Year fixed effects          YES               YES
Industry fixed effects      YES               YES
Observations             40,129           28,259
Adj. [R.sup.2]                0.32             0.35

                              Panel B. Cost of Equity
                               (9)
                             Dependent variable:
                             [Cost.sub.Equity]

Ln[(Bills).sub.t]             0.00 (***)
                             (2.65)
[PSI.sup.R.sub.t]
Ln([Bills.sub.t]) (*)
[PSI.sup.R.sub.t]
[B-index.sup.R.sub.t]
Ln[(Bills).sub.t]
x [B-index.sup.R.sub.t]
[P-index.sup.R.sub.t]
Ln[(Bills).sub.t] x
[P-index.sup.R.sub.t]
[L-index.sup.R.sub.t]        -0.02 (***)
                            (-4.49)
Ln[(Bills).sub.t] x          -0.00 (*)
[[L-index.sup.R].sub.t]
                            (-1.76)
Ln[(Size).sub.t]              0.02 (***)
                            (12.79)
[BM.sub.t]                   -0.01 (***)
                            (-3.71)
[Leverage.sub.t]             -0.01
                            (-0.80)
[ROA.sub.t]                  -0.11 (***)
                            (-5.18)
[Tangibility.sub.t]          -0.02 (***)
                            (-1.96)
[HHI.sub.t]                  -0.03 (***)
                            (-3.07)
Ln[(Firm Age).sub.t]         -0.02 (***)
                            (-9.95)
Constant                     -0.03
                            (-1.56)
Year fixed effects           YES
Industry fixed effects       YES
Observations             28,259
Adj. [R.sup.2]                0.36

(***) Significant at the 0.01 level.
(**) Significant at the 0.05 level.
(*) Significant at the 0.10 level

Table V. Corporate Political Strategies and Firm-specific Performance
Heterogeneity

This table reports the estimated coefficients of the cross-sectional
regressions where the dependent variable (IV) is defined as the
variance of residuals, obtained from the Carhart (1997) four-factor
model, in which a firm's daily returns are regressed on the size,
value, and momentum factors along with market risk premium over the
year. Bills is the number of bills linked to the industry in which a
firm belongs and that are introduced by either the home state Senators
or House Representatives. Political [strategy.sup.R] is the residual
value from the regression of Political strategy on firm size, where
Political strategy = PSI, B-index, P-index, or L-index. PSI is the
annual corporate political strategy index. B-index is the political
strategy index that combines the yearly ranks of N. of connected
directors (the number of board members who are politically connected),
Board s political experience (the average tenure of past political
activities of the boards of directors) and Board's political freshness
(= board's political freshness based on the directors' elapsed
period). P-index = the political strategy index that combines the
yearly ranks of N. of supported candidates (the number of supported
candidates), Strength of relationships (the strength of the
relationships between candidates and the contributing firm), Supported
candidates' ability (the ability of the candidates to help the firm),
and Supported candidates' power (the power of the candidates). L-index
is the political strategy index that measures the yearly rank of
Lobbying expenditures. Refer to Appendix C for detailed variable
descriptions. Firm characteristic-related control variables, year, and
industry dummies are included, but the coefficients are omitted for
brevity. Numbers in parentheses are t-statistics computed using
standard errors that are clustered at both the firm and year level.

                            (1)              (2)               (3)
                                             PSI

Ln ([Bills.sub.t])           0.02 (***)       0.02 (***)      0.02 (***)
                            (2.85)           (3.03)          (2.63)
Political                                     0.16 (***)      0.14 (***)
[Strategy.sup.R.sub.t]
                                             (6.51)          (7.04)
Ln([Bills.sub.t]) x                                           0.02 (**)
Political
[Strategy.sup.R.sub.t]
                                                              (1.96)
Controls                    YES              YES              YES
Year fixed effects          YES              YES              YES
Industry fixed              YES              YES              YES
effects
Obs.                    66,059           66,059          66,059
Adj. [R.sup.2]               0.11             0.11            0.11

                             (4)                     (5)
                                   B-index
                                  Dependent variable:
                                  [IV.sub.t]

Ln ([Bills.sub.t])             0.02 (***)             0.02 (***)
                              (2.86)                 (2.87)
Political                      0.05 (***)             0.03 (***)
[Strategy.sup.R.sub.t]
                              (4.69)                 (4.88)
Ln([Bills.sub.t]) x                                   0.02 (*)
Political
[Strategy.sup.R.sub.t]
                                                     (1.74)
Controls                     YES                     YES
Year fixed effects           YES                     YES
Industry fixed               YES                     YES
effects
Obs.                      66,059                 66,059
Adj. [R.sup.2]                 0.11                   0.11

                            (6)             (7)
                                   P-index
                             Dependent variable:
                             [IV.sub.t]

Ln ([Bills.sub.t])           0.02 (***)       0.02 (***)
                            (3.10)           (3.12)
Political                    0.11 (***)       0.09 (***)
[Strategy.sup.R.sub.t]
                            (7.40)           (6.11)
Ln([Bills.sub.t]) x                           0.02 (***)
Political
[Strategy.sup.R.sub.t]
                                             (3.34)
Controls                    YES              YES
Year fixed effects          YES              YES
Industry fixed              YES              YES
effects
Obs.                    66,059           66,059
Adj. [R.sup.2]               0.11             0.11

                           (8)               (9)
                                  L-index
                                 Dependent variable:
                                  [IV.sub.t]

Ln ([Bills.sub.t])           0.02 (***)        0.02 (***)
                            (3.68)            (3.73)
Political                    0.08 (***)        0.06 (***)
[Strategy.sup.R.sub.t]
                            (3.36)            (5.90)
Ln([Bills.sub.t]) x                            0.02
Political
[Strategy.sup.R.sub.t]
                                              (1.23)
Controls                    YES               YES
Year fixed effects          YES               YES
Industry fixed              YES               YES
effects
Obs.                    47,607            47,607
Adj. [R.sup.2]               0.14              0.14

(***) Significant at the 0.01 level.
(**) Significant at the 0.05 level.
(*) Significant at the 0.10 level.

Table VI. Real Options and Political Strategies

This table reports the estimated coefficients of the Fama-MacBeth
(1973) regressions where the dependent variable (excess return) is a
firm's monthly stock return minus the risk-free rate. [Vol.sub.t]
measures a firm's return volatility using daily returns at month t.
[DELTA] [Vol.sub.t] is a monthly change in a firm's return volatility
(Vol) from t-l to t. Political [strategy.sup.R] is the residual value
from the regression of Political strategy on firm size, where
Political strategy = PSI, B-index, P-index, or L-index. PSI is the
annual corporate political strategy index. B-index is the political
strategy index that combines the yearly ranks of N. of connected
directors (the number of board members who are politically connected),
Board's political experience (the average tenure of past political
activities of the boards of directors) and Board's political freshness
(the board's political freshness based on the directors' elapsed
period). P-index is the political strategy index that combines the
yearly ranks of N. of supported candidates (the number of supported
candidates), Strength of relationships (the strength of the
relationships between candidates and the contributing firm), Supported
candidates' ability (the ability of the candidates to help the firm),
and Supported candidates' power (the power of the candidates). L-index
is the political strategy index that measures the yearly rank of
Lobbying expenditures. Ln(Size) is a firm's market value of equity. BM
is a ratio of the book value to the market value of equity.
[Ret.sub.(t-1,t-12)] is a firm's cumulative returns for past 12
months. [Volume.sub.t-1] is a firm's monthly trading volume normalized
by the shares outstanding. [Beta.sub.t-1] is a firm's systematic risk
measured as its past 60 month returns and the market returns. R&D is
the R&D expenditures divided by assets. FreeCash is the free cash flow
divided by assets. Foreign is a dummy that is equal to one if a firm's
foreign sales are greater than zero in a given calendar year and zero
otherwise. Union is a percentage of labor union coverage in a given
four-digit SIC industry code. Numbers in parentheses are t-statistics
computed using Newey-West autocorrelation standard errors up to six
lags.

                          (1)           (2)           (3)
                                        PSI           PSI
                           Dependent variable: Excess [Return.sub.t]

[DELTA] [Vol.sub.t]         1.22 (***)    1.22 (***)
                          (15.25)       (15.26)       (14.56)
Political                                 0.01 (***)    0.01 (***)
[Strategy.sup.R.sub.t-1]
                                         (5.30)        (3.31)
[DELTA] [Vol.sub.t]                                     1.01 (***)
x Political
[Strategy.sup.R.sub.t-1]
                                                       (4.75)
Ln[(Size).sub.t-1]         -0.00 (***)   -0.00 (***)   -0.00 (***)
                          (-3.97)       (-3.98)       (-3.91)
[BM.sub.t-1]                0.01 (***)    0.01 (***)    0.01 (***)
                           (2.95)        (2.76)        (2.84)
[Ret.sub.(t-1,t-12)]       -0.01 (***)   -0.01 (***)   -0.01 (***)
                          (-3.55)       (-3.46)       (-3.44)
[Volume.sub.t-1]            0.02 (***)    0.02 (***)    0.02 (***)
                           (4.96)        (5.03)        (4.95)
[Beta.sub.t-1]             -0.00         -0.00         -0.00
                          (-1.38)       (-1.26)       (-1.25)
[R&D.sub.t-1]               0.03 (***)    0.03 (***)    0.03 (***)
                           (4.26)        (4.30)        (4.35)
[FreeCash.sub.t-1]          0.02 (***)    0.02 (***)    0.02 (***)
                           (5.94)        (6.19)        (6.28)
[Foreign.sub.t-1]           0.00 (***)    0.00 (***)    0.00 (***)
                           (3.34)        (3.25)        (3.25)
[Union.sub.t-1]             0.01 (***)    0.01 (**)     0.01 (**)
                           (3.32)        (2.37)        (2.25)
Constant                    0.04 (***)    0.04 (***)    0.04 (***)
                           (3.16)        (3.17)        (3.14)
Number of months          180           180           180
Avg. [R.sup.2]              0.02          0.02          0.03

                          (4)            (5)           (6)
                          B-lndex        B-lndex       P-lndex
                          Dependent variable: Excess [Return.sub.t]

[DELTA] [Vol.sub.t]         1.22 (***)    1.22 (***)    1.22 (***)
                          (15.25)       (15.09)       (15.26)
Political                   0.00 (***)    0.00 (**)     0.01 (***)
[Strategy.sup.R.sub.t-1]
                           (4.02)        (2.51)        (5.42)
[DELTA] [Vol.sub.t]                       0.24 (*)
x Political
[Strategy.sup.R.sub.t-1]
                                         (1.88)
Ln[(Size).sub.t-1]         -0.00 (***)   -0.00 (***)   -0.00 (***)
                          (-3.97)       (-3.96)       (-3.96)
[BM.sub.t-1]                0.01 (***)    0.01 (***)    0.01 (***)
                           (2.90)        (2.93)        (2.76)
[Ret.sub.(t-1,t-12)]       -0.01 (***)   -0.01 (***)   -0.01 (***)
                          (-3.53)       (-3.51)       (-3.49)
[Volume.sub.t-1]            0.02 (***)    0.02 (***)    0.02 (***)
                           (4.98)        (4.97)        (5.03)
[Beta.sub.t-1]             -0.00         -0.00         -0.00
                          (-1.35)       (-1.35)       (-1.27)
[R&D.sub.t-1]               0.03 (***)    0.03 (***)    0.03 (***)
                           (4.26)        (4.24)        (4.34)
[FreeCash.sub.t-1]          0.02 (***)    0.02 (***)    0.02 (***)
                           (6.02)        (6.11)        (6.12)
[Foreign.sub.t-1]           0.00 (***)    0.00 (***)    0.00 (***)
                           (3.32)        (3.36)        (3.32)
[Union.sub.t-1]             0.01 (***)    0.01 (***)    0.01 (**)
                           (3.09)        (3.11)        (2.55)
Constant                    0.04 (***)    0.04 (***)    0.04 (***)
                           (3.17)        (3.16)        (3.16)
Number of months          180           180           180
Avg. [R.sup.2]              0.02          0.02          0.02

                          (7)           (8)           (9)
                          P-lndex       L-lndex       L-lndex
                          Dependent variable: Excess [Return.sub.t]

[DELTA] [Vol.sub.t]         1.21 (***)    1.32 (***)    1.28 (***)
                          (14.74)       (13.53)       (13.11)
Political                   0.00 (***)    0.01 (***)    0.00 (**)
[Strategy.sup.R.sub.t-1]
                           (3.30)        (3.59)        (2.18)
[DELTA] [Vol.sub.t]         1.11 (***)                  0.74 (***)
x Political
[Strategy.sup.R.sub.t-1]
                           (5.29)                      (5.32)
Ln[(Size).sub.t-1]         -0.00 (***)   -0.00 (***)   -0.00 (***)
                          (-3.88)       (-3.79)       (-3.71)
[BM.sub.t-1]                0.01 (***)    0.01 (*)      0.01 (*)
                           (2.78)        (1.77)        (1.87)
[Ret.sub.(t-1,t-12)]       -0.01 (***)   -0.01 (***)   -0.01 (***)
                          (-3.46)       (-3.43)       (-3.43)
[Volume.sub.t-1]            0.02 (***)    0.02 (***)    0.02 (***)
                           (4.92)        (4.71)        (4.67)
[Beta.sub.t-1]             -0.00         -0.00         -0.00
                          (-1.22)       (-1.21)       (-1.20)
[R&D.sub.t-1]               0.03 (***)    0.03 (***)    0.03 (***)
                           (4.31)        (3.15)        (3.17)
[FreeCash.sub.t-1]          0.02 (***)    0.01 (***)    0.01 (***)
                           (6.14)        (4.34)        (4.34)
[Foreign.sub.t-1]           0.00 (***)    0.00 (***)    0.00 (***)
                           (3.23)        (3.00)        (2.99)
[Union.sub.t-1]             0.01 (**)     0.01 (*)      0.01 (*)
                           (2.41)        (1.89)        (1.88)
Constant                    0.04 (***)    0.04 (***)    0.04 (***)
                           (3.12)        (3.04)        (2.97)
Number of months          180           132           132
Avg. [R.sup.2]              0.03          0.03          0.03

(***) Significant at the 0.01 level.
(**) Significant at the 0.05 level.
(*) Significant at the 0.05 level.

Table VII. Real Options Test for High and Low Policy Uncertainty
Subsamples

This table reports the estimated coefficients of the Fama-MacBeth
(1973) regressions where the dependent variable (excess return) is a
firm's monthly stock return minus THE risk-free rate. [Vol.sub.t]
measures a firm's return volatility using daily returns at month t.
[DELTA] [Vol.sub.t] is a monthly change in a firm's return volatility
(Vol) from t-1 to t. Political [strategy.sup.R] IS the residual value
from the regression of Political strategy on firm size, where
Political strategy = PSI, B-index, P-index, or L-index. PSI is the
annual corporate political strategy index. B-index is the political
strategy index that combines the yearly ranks of N. of connected
directors (the number of board members who are politically connected),
Board's political experience (the average tenure of past political
activities of boards of directors) and Board's political freshness
(the board's political freshness based on the directors' elapsed
period). P-index is the political strategy index that combines the
yearly ranks of N. of supported candidates (the number of supported
candidates), Strength of relationships (the strength of the
relationships between candidates and the contributing firm), Supported
candidates' ability (the ability of the candidates to help the firm),
and Supported candidates' power (the power of the candidates). L-index
is the political strategy index that measures the yearly rank of
Lobbying expenditures. High (Low) indicates a group in the top
(bottom) tercile of policy risk. A tercile group of policy risk is
based on Bills defined as the number of bills linked to the industry
in which a firm belongs and that are introduced by either the home
state Senators or House Representatives. PSI, B-index, P-index, and
L-index is annual corporate political strategy index. Numbers in
parentheses are t-statistics computed using the Newey-West
autocorrelation standard errors up to six lags.

                           (1)                   (2)
                                       PSI
                            Dependent variable: Excess [Return.sub.t]
                            High                  Low

[DELTA] [Vol.sub.t]          1.34 (***)           1.09 (***)
                            14.15)              (11.14)
Political                    0.01 (***)           0.01 (**)
[Strategy.sup.R.sub.t-1]
                            (3.48)               (2.12)
[DELTA] [Vol.sub.t]          1.68 (***)           0.69 (**)
x Political
[Strategy.sup.R.sub.t-1]
                            (4.72)               (2.46)
                              Ho: [beta](1) = [beta](2)
                              (p-value = 0.03)
Controls                   YES                   YES
Number of months           180                  180
Avg. [R.sup.2]               0.02                 0.03

                             (3)               (4)
                                         B-lndex
                             Dependent variable: Excess [Return.sub.t]
                              High             Low

[DELTA] [Vol.sub.t]           1.39 (***)       1.14 (***)
                            (14.27)          (11.48)
Political                     0.00 (**)        0.00 (**)
[Strategy.sup.R.sub.t-1]
                             (2.19)           (2.03)
[DELTA] [Vol.sub.t]           0.33             0.08
x Political
[Strategy.sup.R.sub.t-1]
                             (1.53)           (0.49)
                               Ho: [beta](3) = [beta](4)
                                (p-value = 0.34)
Controls                    YES              YES
Number of months            180              180
Avg. [R.sup.2]                0.02             0.02

                            (5)                  (6)
                                     P-lndex
                           Dependent variable: Excess [Return.sub.t]
                              High               Low

[DELTA] [Vol.sub.t]           1.35 (***)         1.11 (***)
                            (14.02)            (11.36)
Political                     0.01 (***)         0.01 (**)
[Strategy.sup.R.sub.t-1]
                             (2.63)             (2.56)
[DELTA] [Vol.sub.t]           1.88 (***)         0.92 (***)
x Political
[Strategy.sup.R.sub.t-1]
                             (5.13)             (3.32)
                                Ho: [beta](5) = [beta](6)
                                 (p-value = 0.04)
Controls                    YES                YES
Number of months            180                180
Avg. [R.sup.2]                0.02               0.03

                           (7)                (8)
                                    L-lndex
                            Dependent variable: Excess [Return.sub.t]
                            High                  Low

[DELTA] [Vol.sub.t]          1.41 (***)           1.17 (***)
                            12.64)              (10.13)
Political                    0.01 (***)           0.00
[Strategy.sup.R.sub.t-1]
                            (2.86)               (1.05)
[DELTA] [Vol.sub.t]          1.04 (***)           0.73 (***)
x Political
[Strategy.sup.R.sub.t-1]
                            (4.29)               (3.73)
                              Ho: [beta](7) = [beta](8)
                               (p-value = 0.32)
Controls                   YES                 YES
Number of months           132                  132
Avg. [R.sup.2]               0.03                 0.03

(***) Significant at the 0.01 level.
(**) Significant at the 0.05 level.

Table VIII. Politician's Sudden Death and the Effect of Political
Strategy: A Causality Test

Panel A reports the estimated coefficients of the cross-sectional
regressions. Beta is systematic risk computed from the market model
using daily returns over the year. IV is idiosyncratic volatility
computed as the variance of residuals obtained from the Carhart (1997)
four-factor model in which a firm's daily returns are regressed on the
size, value, and momentum factors along with the market risk premium
over the year. Bills is the number of bills linked to the industry in
which a firm belongs and that are introduced by either the home state
Senators or House Representatives. [B-index.sup.R] is the residual
value of the political strategy index that combines the yearly ranks
of N. of connected directors, Board's political experience, and
Board's political freshness. N. of connected directors is the number
of board members who are politically connected. Board's political
experience is the average tenure of past political activities of the
boards of directors. Board's political freshness is the board's
political freshness based on the directors' elapsed period. Postdeath
is one for the years after an ex-politician on the board suddenly dies
and zero for the years prior to death. Firm characteristic-related
control variables, year, and industry dummies are included, but the
coefficients are omitted for brevity. Numbers in parentheses are
t-statistics computed using standard errors that are clustered at both
the firm and year level. Panel B reports the estimated coefficients of
the Fama-MacBeth (1973) regressions where the dependent variable
(excess return) is a firm's monthly stock return minus the risk-free
rate. [Vol.sub.t] measures a firm's return volatility using daily
returns at month t. [DELTA] [Vol.sub.t] is a monthly change in a
firm's return volatility (Vol) from t-1 to t. High (Low) policy risk
indicates a group in the top (bottom) tercile of policy risk. A
tercile group of policy risk is based on Bills defined as the number
of bills linked to the industry in which a firm belongs and that are
introduced by either the home state Senators or House Representatives.
Numbers in parentheses are t-statistics computed using Newey-West
autocorrelation standard errors up to six lags. Refer to Appendix C
for detailed variable descriptions.

                  Panel A. Political Strategy Index and Firm Risk
                                        (1)              (2)
                                        [Beta.sub.t]     [IV.sub.t]

Ln[(Bills).sub.t]                           0.04 (***)      -0.03 (***)
                                           (5.59)          (-3.09)
[B-Index.sup.R.sub.t]                       0.12 (***)      -0.06
                                           (4.26)          (-1.45)
[PostDeath.sub.t]                           0.12 (***)      -0.21 (**)
                                           (4.42)          (-2.41)
Ln(Bilts)t x [B-Index.sup.R.sub.t]         -0.05 (***)       0.05
                                          (-2.76)           (1.45) (***)
PosDeath x [B-Index.sup.R.sub.t]           -0.12 (***)       0.16
                                          (-4.68)           (7.17)
Ln(Bills)t x [PostDeath.sub.t]             -0.05 (***)       0.12 (***)
                                          (-3.64)           (4.33)
Ln(Bills)t x [B-Index.sup.R.sub.t] x        0.03 (*)        -0.03 (**)
[PostDeath.sub.t]
                                           (1.68)          (-2.02)
Controls                                YES              YES
Year fixed effects                      YES              YES
Industry fixed effects                  YES              YES
Obs.                                    3,403            3,403
Adj. [R.sup.2]                              0.41             0.48

       Panel B. Policy Risk, Political Strategy, and Stock Return
                           (1)                   (2)
                           Dependent Variable: Excess return

LLow [DELTA] [Vol.sub.t]     1.40 (***)          1.57 (***)
                           (10.56)             (13.31)
[B-index.sup.R]              0.00 (**)           0.00 (*)
                            (2.32)              (1.73)
Postdeath                                        0.16
                                                (1.48)
[DELTA] [Vol.sub.t] x        0.69 (***)          0.81 (***)
[B-index.sup.R]
                            (3.88)              (5.96)
[B-index.sup.R] x                              -11.42
Postdeath
                                               (-0.45)
[DELTA] [Vol.sub.t]                             -0.17
x Postdeath                                    (-1.39)
[DELTA] [Vol.sub.t] x                            4.45
[B-index.sup.R] x
Postdeath
                                                (0.48)
Controls                   YES                   YES
Number of months           108                 108
Avg. [R.sup.2]               0.02                0.02

       Panel B. Policy Risk, Political Strategy, and Stock Return

                           (3)                         (4)
                           Dependent Variable: Excess return
                           High Policy Risk            Low Policy
                                                       Risk

LLow [DELTA] [Vol.sub.t]      1.81 (***)                  1.39 (***)
                            (13.15)                     (11.75)
[B-index.sup.R]               0.00 (*)                    0.00
                             (1.72)                      (1.44)
Postdeath                     0.00                       -0.00
                             (0.63)                     (-0.66)
[DELTA] [Vol.sub.t] x         0.96 (***)                  0.52 (***)
[B-index.sup.R]
                             (3.99)                      (3.19)
[B-index.sup.R] x             0.04                      -48.84
Postdeath
                             (0.08)                     (-0.94)
[DELTA] [Vol.sub.t]           0.00                        0.01 (*)
x Postdeath                  (0.16)                      (1.65)
[DELTA] [Vol.sub.t] x        -0.95 (*)                    7.67
[B-index.sup.R] x
Postdeath
                            (-1.69)                      (1.01)
                           Ho: [beta](3) = [beta](4)
                           (p-value = 0.31)
Controls                    YES                         YES
Number of months            108                         108
Avg. [R.sup.2]                0.02                        0.02

(***) Significant at the 0.01 level.
(**) Significant at the 0.05 level.
(*) Significant at the 0.10 level.
COPYRIGHT 2019 Financial Management Association
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2019 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Kim, Chansog "Francis"; Kim, Incheol; Pantzalis, Christos; Park, Jung Chul
Publication:Financial Management
Article Type:Report
Geographic Code:1USA
Date:Jun 22, 2019
Words:18364
Previous Article:Are specialist funds "special"?
Next Article:Expected versus Ex Post Profitability in the Cross-Section of Industry Returns.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters