Strategic deviation and investment inefficiency

Dinithi Ranasinghe; Ahsan Habib

doi:10.1177/03128962231152764

Strategic deviation and investment inefficiency

Ranasinghe, Dinithi; Habib, Ahsan 2023-01-01 00:00:00 We examine the association between strategic deviation and investment inefficiency. We conceptualize strategic deviation as the extent to which the pattern of a firm’s resource allocation deviates from its industry peers. We posit that firms pursuing deviant strategies are prone to increased information asymmetry and hence, are able to engage in self-serving behaviour as manifested in inefficient investments. Our results suggest that deviant firms have sub-optimal investments. A battery of robustness tests validates our findings. We further provide evidence to suggest that weaker monitoring, high product market competition and a low-quality information environment moderate the relation between strategic deviation and investment inefficiency. JEL Classification: M41, G41 Keywords Information asymmetry, investments, product market competition, strategic deviation 1. Introduction In this study, we examine whether strategic deviation, conceptualized as resource allocation that deviates from industry peers, is associated with investment inefficiency. Managers can avoid competition by pursuing strategic choices that are different from their peers (Porter, 1996), thereby making it difficult for shareholders to evaluate the managerial performance of such firms (Carpenter, 2000). Accordingly, prior research finds that firms that deviate from common industry Corresponding author: Dinithi Ranasinghe, Department of Accountancy and Finance, University of Otago, P.O. Box 56, Dunedin 9054, New Zealand. Email: dinithi.ranasinghe@otago.ac.nz Final transcript accepted on 3 January 2023 by Tom Smith (AE Finance) 2 Australian Journal of Management 00(0) strategies hold cash for opportunistic reasons (Dong et al., 2021), have less synchronous stock returns (Ye et al., 2021) and exhibit extreme performance compared to their industry peers (Tang et al., 2011). We extend the literature on the consequences of strategic deviation by examining its association with firms’ investment decisions. Prior research in this space have primarily captured strategy based on strategy typologies, such as Miles and Snow’s (1978) prospector, analyzer and defender strategies. For example, Navissi et al. (2017) find that prospector (defender)-type firms are more likely to over (under)-invest. They argue that prospector-type firms enjoy greater managerial discretion and less stringent mon- itoring, thereby enabling them to over-invest for self-serving behaviour, including, but not limited to, reducing their career-related risks. Defender-type firms, on the other hand, are subject to a relatively higher level of managerial monitoring and less managerial discretion that results in under-investments. We differentiate our study form Navissi et al. (2017) by examining the consequences of strategic deviation, which is conceptually different from strategy typology. While strategy typology focuses on resource allocation within firms, strategic deviation focuses on an inter-firm strategy perspective and captures the differences in the focal firm’s resource allocation decisions when compared to industry peers (Finkelstein and Hambrick, 1990; Geletkanycz and Hambrick, 1997). While strategy typologies, such as prospector and defender strategies, are entrenched in a firm’s competitive envi- ronment (Porter, 1980), strategic deviation captures a firm’s competitive positioning relative to its peers. Furthermore, researchers arbitrarily choose cut-off scores to categorize firms into prospector, defender and analyzer groups (Bentley et al., 2013; Dong et al., 2021). In contrast, strategic devia- tion focuses on the resource allocation of a firm relative to the industry which reflects how managers pursue strategies in comparison to the commonly adopted industry practices. We complement the strategy-based explanation for corporate investment policies by providing evidence that the adop- tion of a deviant strategy increases investment inefficiencies. Business strategy determines the resource allocations for investments (Miles and Snow, 1978). A strategy that is in line with the industry norms, that is, conforming strategy, is likely to provide benchmarks for evaluation (Porter, 1996). Deviant strategy, that is, a strategy that is different from industry norms, on the other hand, fails to provide appropriate benchmarks for comparisons (Carpenter, 2000). The lack of benchmarks creates information risk and uncertainties for capital market participants. Furthermore, managers of firms pursuing deviant strategies are required to invest in new technology and other projects to attract new customers, segments and markets: investments that reduce the benefit of economies of scale and create higher risks. While some researchers argue that being different has first-entry benefits (Porter, 1980, 1986), we take the view that deviant firms face high risk and uncertainty in cash flow generation and escalate information asymmetry and an opaque information environment (Carpenter, 2000; Deephouse, 1999; Litov et al., 2012). Taken together, the lack of benchmark comparison, information asymmetry and esca- lated needs for capital expenditures among deviant firms incentivize managers to pursue sub- optimal investments. We empirically investigate this proposition. We then explore the settings under which the expected relation between strategic deviation and investment inefficiency might vary cross-sectionally. First, we examine whether internal and exter- nal governance mechanisms moderate this association. Effective governance mechanisms exert effective monitoring (Fama and Jensen, 1983). Shleifer and Vishny (1997) argue that independent directors are effective in monitoring managerial actions. In addition, monitoring by institutional investors reduces over-investments (Ferreira and Matos, 2008). Strategic deviation creates infor- mation asymmetry, and thus constrains strong monitoring compared to other industry peers. Therefore, investment inefficiency for firms with high strategic deviation will be exacerbated by poor monitoring. Ranasinghe and Habib 3 Then, we test whether product market competition moderates the association between strategic deviation and investment inefficiency. A stream of research supports the disciplinary view of prod- uct market competition (Babar and Habib, 2021). However, intense competition increases informa- tion risk because the managers operating in highly competitive industries are reluctant to release proprietary information: an action that hinders obtaining product profitability information ex ante by peers (Stoughton et al., 2017), thus enabling managers of deviant firms to engage in sub-optimal investments. We then investigate the moderating role of information asymmetry on the association between strategic deviation and investment inefficiency. Ye et al. (2021) suggest that deviant firms are dif- ferent from industry peers and, therefore, have high information processing costs and increased information asymmetry. Prior research argues that increased information asymmetries arising from information opaqueness increase investment inefficiency (Chen et al., 2017; Lin et al., 2021). We, therefore, expect that the positive association between strategic deviation and investment ineffi- ciency will be more pronounced in a high-information-asymmetry setting. Finally, we consider the role of financial statement comparability. More comparable financial statements increase the abil- ity of capital market participants to benchmark managerial and firm performance and, as a result, reduces information acquisition and processing costs (Choi et al., 2019; De Franco et al., 2011; Imhof et al., 2022; Kim et al., 2016). As argued before, since strategic deviation enables managers to avoid strict monitoring, the incentives for producing more comparable financial statements are likely lower for such firms. We, therefore, argue that the positive association between strategic deviation and investment inefficiency will be more pronounced when financial statement compa- rability is low. A majority of research on investment efficiency has modelled in(efficient) investment in two- stages with residuals obtained in the first stage being used as the dependent variable in second stage as proxy for investment efficiency (Biddle et al., 2009). However, several studies have questioned the appropriateness of such a research design highlighting biased coefficients and t-statistics, which are not in the expected direction (Chen et al., 2018, 2022; Christodoulou et al., 2018; Jackson, 2022). Jackson (2022) suggests a one-stage regression procedure with a set of indicator variables for industries, years, the independent variable from the first-step regression and interaction terms between each of industries, years and the independent variable from the first-step regression. We follow this procedure in testing our hypothesis. Using a US sample of 56,133 firm-year observations from 1987 to 2020, we document a positive and significant rela- tionship between strategic deviation and investment inefficiency. Fixed effect estimates, two- stage least square (2SLS) estimates, the entropy balancing test and two-step system GMM (generalized method of moments) to allay endogeneity concerns validate our original findings. Furthermore, weaker monitoring, increased information asymmetry and low financial statement comparability exacerbate the positive association between strategic deviation and investment inefficiency. In an additional test, we find that investments by the deviant firms are discounted by the capital market. Our study is motivated based on the calls for investigations into the repercussions of adopting a conforming or a deviant strategy (Deephouse, 1996), as it is important to understand the conse- quences of deviating from industry strategy norms (Chen and Hambrick, 1995). While there are some studies that examine the consequences of strategy typology, only a few have looked into the strategy deviation aspect (Dong et al., 2021; Provaty et al., 2022; Tang et al., 2011; Ye et al., 2021), and We also differ from strategy differentiation, which identifies resource allocation between seg- ments (Dong et al., 2021), and instead focus on inter-firm differences of resource allocation across functions, such as production, marketing, innovation and finance (Finkelstein and Hambrick, 1990). Thus, we capture a firm’s strategy positioning in a competitive market. Furthermore, we 4 Australian Journal of Management 00(0) respond to calls for further research on examining the first-order determinants of investment in(efficiency) (Biddle et al., 2009) by providing evidence from a strategic deviation perspective. We contribute to the strategic management literature by answering the call (Deephouse, 1999) for the implications, investment inefficiency in our case, of a firm being different from its peers. We further extend corporate investment literature by identifying that being different from industry peers increases investment inefficiency. Thereby, we extend strategy and investments literature, such as Navissi et al. (2017). The paper is organized as follows. This introduction is followed by literature and hypotheses development in section 2. Section 3 presents the methodology, while section 4 discusses the empir- ical results and robustness tests. Section 5 concludes the paper. 2. Literature and hypotheses development According to Modigliani and Miller (1958), firms are likely to pursue optimal investment strate- gies in a perfect market, but agency frictions and financial constraints make markets imperfect and hence, affect optimal investment levels (Jensen, 1986; Myers, 1977; Shleifer and Vishny, 1989). Under-investments occur when managers pass on positive NPVs to protect their career concerns, while over-investments occur when they invest in negative NPV projects for empire- building incentives (Biddle et al., 2009). These agency problems manifest through managerial empire building, career motives, herding behaviour and managerial myopia (Bebchuk and Stole, 1993; Holmström, 1999; Jensen, 1986; Malmendier and Tate, 2005). Moral hazard–based agency frictions escalate under-investments, while adverse selection-induced agency frictions increase over-investments. Shleifer and Vishny (1997) suggest that weak corporate governance aggravates agency frictions and, consequently, accentuates investment inefficiencies. Agency theory, there- fore, posits that managers subject to weak monitoring engage in over-investments with the intent of empire building (Jensen, 1986). Strategies determine resource allocation in an entity. Therefore, it is important to understand the type of strategies that firms need to follow to pursue effective resource allocation for efficient investments. Corporate strategy can be defined as a pattern that echoes a series of decisions, which determines product markets, technology deployment, organizational structures and business mod- els (Mintzberg, 1978). Miles and Snow’s (1978) strategy typology suggests that firms adopt strate- gies to remain competitive in the market. According to Miles and Snow (1978), three types of business strategies can exist: prospectors, defenders and analyzers. These strategies vary depend- ing on product markets, processes, organizational structures and technology. Specifically, prospec- tor strategy focuses on innovation and market leadership, while the defender strategy focuses on competition based on price, service or quality. The analyzer strategy falls in between the prospector and defender strategy continuum. Bentley et al. (2013) find that prospector firms are more likely to have financial irregularities, despite high audit efforts, due to the inherent business risk. Prospector firms have high risk and uncertainty relative to defender firms, and, as a result, increase the incremental information acqui- sition costs for the analysts (Bentley-Goode et al., 2019). However, prospector firm managers have incentives to reduce information asymmetry to gain access to financial markets for pursuing their innovative strategies. In line with this argument, Bentley-Goode et al. (2019) find that prospector firms make more management earnings forecasts to attract more analyst coverage compared to defender firm managers: actions that result in lower information asymmetry. As the prospector- type firms seek new innovations, they are more likely to engage in over-investments, while defender-type firms suffer from under-investments (Navissi et al., 2017): sub-optimal investments that is associated with poor future performance. Ranasinghe and Habib 5 Strategy deviation differs from strategy typology in the sense that strategy deviation is based on firms’ preference to conform to industry peers. Based on the institutional theory, managers tend to adopt practices that conform to their peers, which drives inter-organizational homogeneity (DiMaggio and Powell, 1983). As a result, when a strategy deviates from the industry norms, it induces costs to investors. While a deviant strategy can assist firms to explore new markets, build unique customer and supplier relationships and achieve competitive advantage (Porter, 1980, 1986), the non-conformity eliminates the benchmarks for comparison (Carpenter, 2000). This cre- ates higher information processing costs for investors to evaluate firm and managerial performance (Litov et al., 2012). As a result, strategic deviation increases agency costs and hinders firms with high strategic deviation to access external resources at a cheaper cost (Deephouse, 1999), creates information asymmetry and increases cash holdings (Dong et al., 2021). Strategic deviation, therefore, allows managers to pursue self-serving behaviour at the expense of shareholders’ inter- est (Jensen and Meckling, 1976; Jensen, 1986). Furthermore, the lack of benchmarks emanating from high strategic deviation reduces competition and escalates agency problems (Shleifer and Vishny, 1997). Therefore, managers of firms with high strategic deviation are likely to over-invest to capture customers and retain suppliers who otherwise may stick with peers who follow conform- ing strategies. Similarly, deviant firm managers may increase capital expenditures and research and development (R&D) expenditures to seek new customers and technology: investments that are risky and may adversely affect future performance Based on the preceding arguments, we develop the following directional hypothesis: H1: There is a positive association between strategic deviation and investment inefficiency. It is important to understand what factors exacerbate the strategic deviation and investment inef- ficiency association. We argue that corporate governance, product market competition and the quality of the information environment moderate the association between strategic deviation and investment inefficiency. 2.1. Corporate governance Independent monitoring is important to alleviate managerial opportunistic behaviour (Fama and Jensen, 1983). Weakly monitored CEOs may engage in empire building by investing in unprofitable projects, leading to over-investments (Jensen, 1986). Similarly, CEOs are likely to subsidize poorly performing divisions to gain personal benefits (Jensen and Meckling, 1976), which also results in investment inefficiencies. At the same time, excessive monitoring distorts efficient decision-making. Therefore, it is important to implement effective monitoring mechanisms to encourage efficient investments. Independent board members act in the best interest of shareholders by effectively moni- toring managerial investment behaviour, among other activities (Fama and Jensen, 1983). Rajkovic (2020) finds that board independence increases investment efficiency, but board independence com- promised by CEO and director social ties decreases investment efficiency (Kang et al., 2021). Institutional ownership, an external governance mechanism, can also play a role in determining investment efficiencies through stronger monitoring (Biddle et al., 2009). Ferreira and Matos (2008) find lower capital expenditure in firms with high institutional ownership, which suggests that institu- tional ownership reduces over-investments. Therefore, we argue that the positive association in H1 above is likely to be stronger for deviant firms lacking effective governance mechanisms. H2: The positive association between strategic deviation and investment inefficiency is stronger for poorly governed firms. 6 Australian Journal of Management 00(0) 2.2. Product market competition Two competing arguments exist on the association between product market competition and invest- ment efficiency (Babar and Habib, 2021). Product market competition acts as an external govern- ance mechanism, and thus disciplines managerial behaviour. In line with this disciplinary argument, a stream of studies find that product market competition increases capital expenditure and R&D but curbs over-investments (Jiang et al., 2015; Laksmana and Yang, 2015) and increases cash flow–enhancing investments (Abdoh and Varela, 2017). A competing view based on the signal precision perspective posits a negative association between product market competition and invest- ment efficiency. Firms operating in a highly competitive market are reluctant to obtain precise signals about their rivals’ actions due to the marginal effect of a single firm’s signal. As information is costly, managers tend to gather information only when it generates higher profits, ex ante. The impact of one firm’s signal is marginal when there are a lot of firms in the industry. Therefore, there is less incentive for information gathering in a competitive market. Investment efficiency is weaker in such a setting. Stoughton et al. (2017) find support for this prediction. Furthermore, prior research provides evidence on the risk-increasing effect of competition (Irvine and Pontiff, 2008). Given the inconclusive evidence on the association between product market competition and cor- porate investment efficiency (Babar and Habib, 2021), we develop the following hypothesis: H3: Product market competition moderates the association between strategic deviation and investment inefficiency. 2.3. Information environment We examine how the information environment proxied by analyst following and financial state- ment comparability moderates the association between strategic deviation and investment ineffi- ciency. Low analyst following reduces the quality of the information environment because of less transparency and high information processing costs for investors. High analyst following, on the other hand, reduces mispricing and information asymmetry (Brennan and Subrahmanyam, 1995; Chung et al., 1995; Roulstone, 2003). As strategic deviation augments information asymmetry, capital market participants tend to discount firms with unique strategies (Litov et al., 2012). Litov et al. (2012) argue that unique strategies create substantial cost for analysts to understand a firm’s resource allocation behaviour, and hence, such firms are followed by fewer analysts. Weaker monitoring courtesy of a low analyst following for firms with deviant strategies provides incen- tives for managers of such firms to engage in sub-optimal investments. We, therefore, hypothesize the following: H4a: The positive association between strategic deviation and investment inefficiency is stronger for firms with high information asymmetry. Financial statement comparability (Comparability) is the extent of similarity in accounting choices between two or more entities under similar economic conditions (Financial Accounting Standards Board (FASB, 2010). Comparability enables investors to properly identify similarities and differences in firm performance within an industry (Francis et al., 2014). Higher comparability helps outsiders to identify peer firms’ risk of under- and over-investments (i.e. investment ineffi- ciencies) due to the availability of benchmark information. As a result, high comparability reduces agency frictions of empire building and moral hazard behaviour. Al-Hadi et al. (2021) find that comparability increases investment efficiency. On the other hand, low comparability increases information asymmetry and reduces the quality of the information environment (De Franco et al., Ranasinghe and Habib 7 2011). Firms that follow deviant strategies have extremely poor performance (Tang et al., 2011), and thus provide incentives for managers to produce less comparable financial statements to con- ceal this poor performance. As sub-optimal investments by firms following deviant strategies may be associated with poor future performance, managers of such firms are likely to produce less comparable financial statements to conceal such sub-optimal behaviour. We, therefore, hypothe- size as follows: H4b: The positive association between strategic deviation and investment inefficiency is stronger for firms with low financial statement comparability. 3. Methodology 3.1. Sample The sample covers data ranging from 1987 to 2020. We begin with 1987, as the data on cash flow from operations are available on the Compustat dataset from 1987. We obtain governance data from Thomson Reuters and analyst following data from I/B/E/S. Since we require five years of data for calculating strategic deviation and also require five years of lagged values for cash flows to calculate the control variables (e.g. cash flow volatility), our actual regression spans the period 1992 to 2020. We winsorize all the continuous variables at the top and bottom one percentiles to reduce the impact of outliers. We also exclude financial firms (SIC codes between 6000 and 6999). We further exclude observations with missing values for the measurement of key dependent, inde- pendent and control variables. Our final sample consists of 56,133 firm-year observations. The number of observations in any given regression varies depending on the model-specific data requirements. Panel A in Table 1 presents the sample selection procedure, while Panel B provides details of the industry distribution of the sample based on the two-digit SIC industry classification. About 33% observations come from machinery, electrical, computer equipment, followed by busi- ness services (16%) and chemical, petroleum and rubber and allied products (12%). 3.2. Variable measurements 3.2.1. Independent variable: Strategic deviation (STRAT_DEV). We follow previous research (Dong et al., 2021; Ye et al., 2021) and measure strategic deviation (_ STRAT DEV ) using six indicators, namely, (a) Advertising intensity, measured as advertising expenditure (XAD) over sales (REVT); (b) R&D intensity, measured as R&D expenditure scaled by sales; (c) Plant and equipment new- ness, measured as net property, plant and equipment (PPENT) scaled by gross property, plant and equipment (PPEGT); (d) Non-production overhead, measured using selling, general and adminis- trative (SG&A) expenses scaled by sales; (e) Inventory level, measured using inventories (INVT) scaled by sales; and (f) Financial leverage, measured as debt (DLTT) over equity (CEQ). For each fiscal year, we standardize each of the measures by industry (based on two-digit SIC codes) and calculate the absolute difference between a firm’s score and its industry average. STRAT _ DEV is the sum of the six standardized difference scores. A higher value of STRAT _ DEV implies a larger deviation of a firm’s strategy from industry norms. 3.2.2. Dependent variable: Investment ( Following prior research (Biddle et al., 2009; Rajkovic, INV ). 2020), we first use the following regression. INV =+ αβ ΔSales + ε , (1) it 01 it−1 it , 8 Australian Journal of Management 00(0) Table 1. Sample selection and distribution. Panel A: sample selection. Initial sample for the period 1992–2020 238,691 Less: observations in the financial institutions (#60 – #69) (67,490) Less: observations lost for strategy deviation calculation (40,081) Less: observations lost for investment efficiency and control variable calculation (74,987) Final sample 56,133 Panel B: sample distribution by industry. 2-Digit SIC Industry Observations % 01-14 Agriculture & mining 2,276 4.05 15-17 Building construction 444 0.79 20-21 Food & kindred products 1,553 2.77 22-23 Textile mill products & apparels 1,111 1.98 24-27 Lumber, Furniture, Paper, and Printing 2,088 3.72 28-30 Chemical, Petroleum, and Rubber & Allied Products 6,769 12.06 31-34 Metal 2,547 4.54 35-39 Machinery, electrical, computer equipment 18,416 32.81 40-49 Railroad, Communications and Other Transportation 2,541 4.53 50-51 Wholesale goods, building materials 2,783 4.96 53-59 Store merchandise, auto dealers, home furniture stores 3,372 6.01 70-79 Business services 9,085 16.18 80-99 Others 3,148 5.60 Total 56,133 100.00 Note: This table presents the sample selection procedure (Panel A) and sample distribution by industry (Panel B). SIC = Standard Industrial Classification. where INV is total investments, calculated as the sum of R&D expenditure, capital expenditure (CAPEX) and acquisition expenditure (AQC), less cash receipts from sale of property, plant and equipment (SPPE), and scaled by the lagged total assets (AT). ΔSales is the sales change from it−1 the last year (for the remainder of the paper we denote this as SALESG). An overwhelming body of empirical research uses two-step regression to estimate and then examine the key determinant(s) of investment in(efficiency). More specifically, the residuals from Equation (1) above is consid- ered by researchers to represent in(efficient) investment and is used as the dependent variable in the second stage regression. Due to the concerns regarding the use of two-step regression models (see Chen et al., 2018, 2022; Jackson, 2022), we use a one-step regression model which includes the first step control (i.e. SALESG), years and interaction terms between each of industries, years and the independent variable from the first-step regression to alleviate the biased coefficient and other econometrics concerns (Chen et al., 2022; Jackson, 2022). 3.3. Empirical model We test the association between STRAT _ DEV and INV using the following OLS regression model: Ranasinghe and Habib 9 INVS =+ ββ TRAT __ DEVM ++ ββ TB SIZE + β CFLVOL it ,, 01 it 23 it ,, it 4 it , + β SALESV __ OL ++ ββ INVVOL TANG + β CFOSALES 5 56 it ,, it 78 it ,, it + β SLACK KL ++ ββ OSSAGE ++ ββ ZSCORE DIV 9 it ,, 10 it 11 it ,, 12 it 13 it , (2) + β OPCYCLLE ++ ββ || DACC SALESG + λ INDUSTRY 14 it ,, 15 it 16 it ,, jj t + λ YEAR ++ λλ INDUSTRYS ** ALESGYEAR SALESG + ε kk,,t jj,, tj t kk,, ti ti,t ∑ ∑∑ k j k where INV is the investment and STRAT _ DEV is the strategic deviation measure as mentioned above. β is the first-stage variable which is also interacted with INDUSTRY and YEAR varia- bles. A positive and significant coefficient on β would support our H1. Detailed variable defini- tions of the control variables are in Appendix 1. 4. Empirical results 4.1. Descriptive statistics We present descriptive statistics of the variables in Table 2. Results show that the mean and median INV is 0.160 and 0.103 ._ STRAT DEV has a mean and a median of 2.237 and 1.817. The mean (median) of MTB is 3.384 (2.015), CFLV _ OL is 0.579 (0.015), SALESV _ OL is 0.211 (0.147) INVV _ OL and is 0.151 (0.058). Of the sample observations, 36% have reported a loss and about 28% have paid dividends. The average absolute discretionary accruals ( || DACC ) are 10% of lagged total assets. Table 3 shows the correlation coefficients between the variables used in the empirical models. INV and STRAT _ DEV are positively significantly correlated (0.004). The correlation between STRAT _ DEV and all of the independent variables is significant at the 1% level, except INV_ VOL. As per the unreported variance inflation factors (VIFs), the maximum vif is 1.85 with a mean of 1.32, suggesting that multicollinearity is not a concern. 4.2. Regression results Table 4 presents the baseline regression results whereby we regress INV on STRAT _ DEV and a set of control variables with standard errors clustered at the firm level. As elaborated in Section 3.2.2, we use one-step regression model. In Column (1), we present pooled OLS regression results where INV is regressed on STRAT _ DEV , a set of control variables, the first stage control vari- able (SALESG), industry and year indicators, SLAESG*industry effects and SALESG*year effects. The coefficient on STRAT _ DEV is positive and significant (β = 0.003, p < 0.01), sug- gesting a positive relationship between strategic deviation and investment inefficiency, thereby supporting H1. The reported coefficient implies that a one-standard-deviation increase in STRAT _ DEV increases INV by 3.23% relative to mean INV ((0.003*1.725 (S.D. of STRAT_ DEV)/0.16 (mean INV)). In Column 2, we present robust regression results. The coefficient on STRAT _ DEV continues to be positive and significant (β = 0.001, p < 0.01). Among the control variables, firm size (SIZE) is positively and significantly associated with INV. Sales volatility (SALES_VOL) is positively and significantly associated with INV in the pooled model but insig- nificant in the robust model. Other controls, such as, SLACK, ZSCORE and OPCYCLE, are 10 Australian Journal of Management 00(0) Table 2. Descriptive statistics. Variables Obs Mean SD 0.25 0.50 0.75 INV 56,133 0.160 0.194 0.045 0.103 0.202 STRAT_DEV 56,133 2.237 1.725 1.259 1.817 2.594 MTB 56,133 3.384 5.261 1.193 2.015 3.570 SIZE 56,133 5.210 2.214 3.622 5.150 6.732 CFL_VOL 56,133 0.579 4.218 0.002 0.015 0.108 SALES_VOL 56,133 0.211 0.206 0.081 0.147 0.264 INV_VOL 56,133 0.151 1.623 0.026 0.058 0.129 TANG 56,133 0.240 0.214 0.079 0.170 0.336 CFOSALES 56,133 0.005 0.214 –0.003 0.060 0.117 SLACK 56,133 4.206 13.451 0.125 0.623 2.757 LOSS 56,133 0.357 0.479 0.000 0.000 1.000 AGE 56,133 2.913 0.534 2.565 3.045 3.367 ZSCORE 56,133 1.143 1.679 0.637 1.345 1.999 DIV 56,133 0.275 0.447 0.000 0.000 1.000 OPCYCLE 56,133 4.719 0.793 4.322 4.804 5.208 |DACC| 56,133 0.103 0.105 0.031 0.070 0.137 SALESG 56,133 0.230 0.788 –0.033 0.078 0.247 BIND 18,686 0.745 0.124 0.667 0.750 0.833 INSTOWN 15,168 0.367 0.482 0.000 0.000 1.000 COMPETITION 53,204 6.817 3.668 4.160 6.015 8.646 ANALYST 27,766 2.125 0.667 1.609 2.079 2.639 FCOMP 33,313 –3.133 1.686 –3.820 –2.800 –2.070 Note: This table shows summary statistics of the variables used in the regression models. negatively and significantly associated with INV. Discretionary accruals (|DACC|) is positively related to INV. These are consistent with prior research (Biddle et al., 2009; Rajkovic, 2020). 4.3. Addressing endogeneity Although the preceding analyses control for a variety of firm characteristics that might explain the association between STRAT _ DEV and investment inefficiency, endogeneity could be a concern that may bias the findings. Endogeneity could arise from omitted variable concern, reverse causa- tion problem (i.e. firms that are less investment efficient could deviate more from their peers), and design choices. We use several endogeneity tests and report the results in Table 5. First, we esti- mate fixed effect models. As shown in Panel A, our original results hold when we use fixed effects estimates (β = 0.004, p < 0.01). Next, we use the entropy balancing method to address any design choices. McMullin and Schonberger (2020) document that entropy balancing noticeably improves covariate balance when compared with propensity-score approaches. In addition, entropy balanc- ing improves the balance quality without removing any observations from either treatment or con- trol group (Hainmueller, 2012). To conduct entropy balancing estimates, we divide our sample into two groups based on the mean STRAT_DEV. We consider firm-year observations with above (below) mean STRAT_DEV as the treated (control) group. In Panel B, we first report means, vari- ances and skewness of all covariates before and after balancing, and the results suggest a reason- able covariate balance. Using the entropy balanced sample, we then run the baseline regression and find the coefficient on STRAT _ DEV positive and significant (β = 0.004, p < 0.01). Ranasinghe and Habib 11 Table 3. Correlation analysis. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1. INV 1.000 2. STRAT_DEV 0.004 1.000 3. MTB 0.182 0.159 1.000 4. SIZE 0.088 –0.145 0.191 1.000 5. CFL_VOL 0.085 0.079 0.104 –0.114 1.000 6. SALES_VOL –0.021 0.042 0.012 –0.245 0.095 1.000 7. INV_VOL 0.054 –0.001 0.024 –0.006 0.096 0.041 1.000 8. TANG 0.011 0.020 –0.083 –0.002 –0.041 –0.146 0.002 1.000 9. CFOSALES –0.187 –0.195 –0.174 0.294 –0.187 –0.063 –0.037 0.175 1.000 10. SLACK 0.100 0.047 0.077 –0.023 0.141 0.015 0.016 –0.281 –0.247 1.000 11. LOSS 0.106 0.126 0.098 –0.317 0.115 0.075 0.033 –0.060 –0.657 0.134 1.000 12. AGE –0.100 –0.071 –0.043 0.260 –0.085 –0.151 –0.030 0.017 0.204 –0.067 –0.223 1.000 13. ZSCORE –0.289 –0.137 –0.218 0.143 –0.180 0.134 –0.045 0.024 0.675 –0.214 –0.535 0.185 1.000 14. DIV –0.146 –0.038 –0.041 0.339 –0.081 –0.167 –0.027 0.131 0.260 –0.117 –0.311 0.291 0.264 1.000 15. OPCYCLE 0.013 0.050 –0.021 –0.067 0.025 –0.140 –0.000 –0.301 –0.059 –0.015 0.023 0.044 –0.099 –0.036 1.000 16. |DACC| 0.185 0.060 0.133 –0.164 0.100 0.111 0.025 –0.117 –0.281 0.109 0.206 –0.122 –0.232 –0.165 0.042 1.000 Note: This table presents the correlation coefficient between variables used in the base models. Bold-faced correlations are significant at p < 0.01. Variables are defined in Appendix 1. 12 Australian Journal of Management 00(0) Table 4. Baseline regression: Strategic deviation and investment inefficiency. (1) (2) Pooled Robust STRAT_DEV 0.003*** 0.001** (0.001) (0.000) MTB 0.000*** 0.000*** (0.000) (0.000) SIZE 0.016*** 0.009*** (0.001) (0.000) CFL_VOL 0.000 –0.000 (0.000) (0.000) SALES_VOL 0.032*** 0.001 (0.006) (0.001) INV_VOL 0.003*** 0.001*** (0.001) (0.000) TANG 0.077*** 0.089*** (0.008) (0.003) CFOSALES –0.028*** –0.000*** (0.010) (0.000) SLACK –0.000** –0.000*** (0.000) (0.000) LOSS –0.016*** –0.004*** (0.002) (0.001) AGE –0.014*** –0.008*** (0.002) (0.001) ZSCORE –0.020*** –0.006*** (0.001) (0.000) DIV –0.023*** –0.011*** (0.002) (0.001) OPCYCLE –0.007*** –0.002*** (0.002) (0.001) |DACC| 0.162*** 0.045*** (0.011) (0.003) First step control SALESG 0.060** 0.112*** (0.029) (0.024) Industry effects Yes Yes Year effects Yes Yes SALESG*Industry effects Yes Yes SALESG*Year effects Yes Yes Constant 0.097*** 0.1094*** (0.025) (0.024) Observations 56,133 56,133 Adjusted R 0.290 0.570 Note: This table shows the regression estimates of the association between strategic deviation (STRAT_DEV) and investment inefficiency (INV) for the pooled model (column 1) and for the robust model (column 2). Both models include the first step control variable, SALESG, and the interactions between SALESG and Fyear and SALESG and Industry (Chen et al., 2022; Jackson, 2022). Robust standard errors clustered at firm-level are in parentheses. **, *** denote a two-tailed p-value of less than 0.10, 0.05 and 0.01, respectively. Variables are defined in Appendix 1. Ranasinghe and Habib 13 Table 5. Endogeneity tests. Panel A: Fixed effect model. INV STRAT_DEV 0.004*** (0.001) Other control variables Yes First step control SALESG 0.078*** (0.001) Constant –0.117*** (0.015) Industry effects No Year Effects Yes SALESG*Year effects Yes Firm effect Yes Observations 56,133 Adjusted R 0.08 Panel B: Entropy balancing models. Covariate matching table. Before balancing After balancing Treat Control Control Mean Variance Skewness Mean Variance Skewness Mean Variance Skewness MTB 3.534 39.270 5.826 3.235 16.910 6.185 3.534 39.270 5.826 SIZE 4.921 5.048 0.213 5.588 4.811 0.135 4.921 5.048 0.213 CFL_VOL 0.676 19.830 13.810 0.452 14.320 17.700 0.676 19.830 13.810 SALES_VOL 0.227 0.050 2.270 0.196 0.036 2.502 0.227 0.050 2.270 INV_VOL 0.144 2.614 86.320 0.173 4.871 75.040 0.144 2.614 86.320 TANG 0.258 0.050 1.074 0.231 0.043 1.487 0.258 0.050 1.074 CFOSALES –0.012 0.048 –1.688 0.025 0.042 –1.850 –0.012 0.048 –1.688 SLACK 4.527 235.600 6.585 3.538 108.900 8.303 4.527 235.600 6.585 LOSS 0.381 0.236 0.493 0.334 0.223 0.703 0.381 0.236 0.493 AGE 2.900 0.285 –0.647 2.933 0.280 –0.744 2.900 0.285 –0.647 ZSCORE 1.133 3.501 –2.193 1.143 2.092 –2.252 1.133 3.501 –2.193 DIV 0.284 0.203 0.958 0.282 0.202 0.971 0.284 0.203 0.958 OPCYCLE 4.658 0.771 –0.563 4.754 0.499 –0.799 4.658 0.771 –0.563 |DACC| 0.105 0.011 1.993 0.100 0.011 2.070 0.105 0.011 1.993 SALESG 0.160 0.404 5.630 0.179 0.319 6.217 0.160 0.404 5.630 Entropy balancing regression Variable INV STRAT_DEV 0.004*** (0.0007) Controls Yes (Continued) 14 Australian Journal of Management 00(0) Table 5. (Continued) Entropy balancing regression Variable INV First step control SALESG 0.074** (0.033) Constant 0.085*** (0.0198) Industry effects Yes Year Effects Yes SALESG*Industry effects Yes SALESG*Year effects Yes Observations 56,133 Adjusted R 0.301 Panel C: 2SLS. 1st stage 2nd stage 1st stage 2nd stage Variables regression regression regression regression DV = SRAT_DEV DV = INV DV = SRAT_DEV DV = INV Geographical Proximity index Tightness/Looseness index PROXIMITY 0.001*** (0.000) LOOSENESS 0.057*** (0.013) SRAT_DEV_PRED 0.015* 0.202*** (0.009) (0.055) MTB 0.000*** 0.000 0.001*** –0.000*** (0.000) (0.000) (0.000) (0.000) SIZE –0.048*** 0.020*** –0.061*** 0.031*** (0.004) (0.001) (0.004) (0.003) CFL_VOL 0.002 0.001*** 0.011** –0.001** (0.002) (0.000) (0.002) (0.001) SALES_VOL 0.179*** 0.0324*** 0.202*** –0.006 (0.037) (0.005) (0.034) (0.014) INV_VOL 0.000 0.003*** –0.009** 0.005*** (0.003) (0.000) (0.004) (0.001) TANG –0.631*** 0.040*** –0.571*** 0.169*** (0.052) (0.008) (0.045) (0.034) CFOSALES –0.838*** –0.020* –0.976*** 0.156*** (0.052) (0.011) (0.056) (0.055) SLACK 0.008*** –0.000 0.006*** –0.001*** (0.001) (0.000) (0.001) (0.000) LOSS –0.018 –0.021*** –0.027 –0.020*** (0.019) (0.003) (0.018) (0.004) AGE –0.024* –0.022*** –0.010 –0.017*** (0.015) (0.002) (0.013) (0.003) (Continued) Ranasinghe and Habib 15 Table 5. (Continued) Panel C: 2SLS. 1st stage 2nd stage 1st stage 2nd stage Variables regression regression regression regression DV = SRAT_DEV DV = INV DV = SRAT_DEV DV = INV Geographical Proximity index Tightness/Looseness index ZSCORE –0.141*** –0.024*** –0.129*** 0.002 (0.006) (0.001) (0.006) (0.007) DIV –0.109*** –0.044*** –0.089*** –0.027*** (0.018) (0.003) (0.017) (0.007) OPCYCLE 0.295*** –0.013*** 0.319*** –0.072*** (0.011) (0.003) (0.010) (0.018) |DACC| 0.485*** 0.207*** 0.444*** 0.122*** (0.066) (0.010) (0.064) (0.029) Industry effects Yes Yes Yes Yes Year effects Yes Yes Yes Yes Constant 0.025 –0.341** (0.196) (0.142) Observations 40,951 40,951 55,172 55,172 Adj R 0.176 – 0.310 Kleibergen-Paap rk LM 0.000 0.000 statistic: p-value Weak identification test: Cragg-Donald Wald F-stat. 203.44 19.14 Stock-Yogo critical value 16.38 16.38 Panel D: Two-step system generalized method of moments (GMM). Variables (1) STRAT_DEV 0.012*** (0.003) MTB 0.0010 (0.001) SIZE 0.0446*** (0.005) CFL_VOL 0.0013 (0.001) SALES_VOL 0.0358** (0.017) INV_VOL 0.0048*** (0.001) TANG 0.1193*** (0.039) CFOSALES 0.1192** (0.059) SLACK –0.0011** (0.000) (Continued) 16 Australian Journal of Management 00(0) Table 5. (Continued) Panel D: Two-step system generalized method of moments (GMM). Variables (1) LOSS –0.0832*** (0.012) AGE –0.0395 (0.028) ZSCORE –0.0400*** (0.006) DIV –0.0265** (0.012) OPCYCLE –0.0594*** (0.014) |DACC| 0.2827*** (0.080) First stage control SALESG 0.0724 (0.137) AR(1) (p value) 0.000 AR(2) (p value) 0.445 Industry effects Yes Year effects Yes SALESG*Year effects Yes Observations 56,133 Note: This table shows the endogeneity tests. Panel A presents the fixed effect estimates. Panel B shows the results from the entropy balanced test. Panel C shows the 2SLS models. Panel D presents two-step system GMM results. Robust standard errors clustered at firm-level are in parentheses. *, **, *** denote a two-tailed p-value of less than 0.10, 0.05 and 0.01, respectively. Variables are defined in Appendix 1. GMM = generalized method of moments. Then we perform 2SLS estimates and present the results in Panel C of Table 5. First, we use geographical proximity (PROXIMITY) to local largest strategic-deviant firm as an instrumental variable. Prior research suggests that decision making strategies are similar among companies in the local community (Pool et al., 2015). Therefore, it is more likely that the focal firm pursues a deviant strategy if that firm is geographically closer to the largest strategic-deviant firm in the city. The largest deviant firm in the city is known as the local-largest deviant firm. Following prior research (Dong et al., 2021; Pool et al., 2015), we first find the latitudes and longitudes for all the observations. We then determine the latitudes and longitudes of the largest strategic-deviant firm at the city level. Then we use the Vincenty’s formula to calculate the geographical proximity between the focal firm and the largest strategic-deviant firm based on longitudes and latitudes of the focal firm and the local largest strategic-deviant firm. Column 1 reveals that the instrument PROXIMITY is positively and significantly associated with STRAT_DEV (β = 0.001, p < 0.01), supporting our assertion that geographical proximity to the local largest strategic-deviant firm increases focal firm’s strategic deviation. As shown in the second stage, the STRAT_DEV_PRED is positively and significantly associated with INV (β = 0.015, p < 0.10). The LM statistic reveals that the excluded instruments are ‘relevant’. The Cragg–Donald F-statistic of 203.44 is higher than the Stock and Yogo (2005) critical value of 16.38, implying that our instrument does not suffer from weak identification. Ranasinghe and Habib 17 Second, following Harrington and Gelfand (2014), we use the US state-level tightness-loose- ness index (LOOSENESS) as the instrumental variable. Harrington and Gelfand (2014) argue that the extent of enforcement of laws and regulations among the states are different and hence, the tolerance level of deviation from law implementation differs between the states. Provaty et al. (2022) argue that the state-level looseness (i.e. if the state’s tolerance for lack of law enforcement is high) has a positive association with strategy deviation. We, therefore, use LOOSENESS index as our second instrumental variable. We take the view that when the LOOSENESS score is high, there is more tolerance towards strategic deviation, hence, we expect a positive association between LOOSENESS and STRAT_DEV. As shown in Column 3, we indeed find a positive association between LOOSENESS and STRAT_DEV (β = 0.057, p < 0.01). In Column 4, we find the coeffi- cient on STRAT_DEV_PRED positive and significant (β = 0.202, p < 0.01). The Cragg–Donald F-statistic, however, is much lower than the first instrument. Finally, we use the two-step system GMM (generalized method of moments) approach adopted by Arellano and Bover (1995) and Blundell and Bond (1998) to validate our results documented in Table 4. This should also alleviate any concerns with reverse causality concern. The coefficient on STRAT_DEV is positive and significant (coefficient 0.012, p < 0.01). Given that errors in levels are serially uncorrelated, we expect significant first-order serial correlation, but insignificant sec- ond-order correlation in the first-differenced residuals. Test results reported at the bottom of Panel D confirm the desirable statistically significant AR(1), and statistically insignificant AR(2). Based on these endogeneity tests above, we conclude that our results are not affected by endo- geneity concerns. 4.4. Strategic deviation and investment inefficiency: Cross-sectional tests Table 6 presents the cross-sectional test results for the association between strategic deviation and investment inefficiency. As discussed in Section 2, we consider corporate governance (board inde- pendence and institutional ownership), product market competition, and the quality of the informa- tion environment proxied by analyst following and financial statement comparability as the three cross-sectional settings that are likely to moderate the relation between strategic deviation and investment inefficiency. Regression results are based on a sub-sampling procedure which allows the coefficients on all the control variables to vary between the groups. Columns 1 and 2 present regression estimates of the association between STRAT_DEV and INV for the board independence ( BIND ) context. The low- (high-) BIND sub-sample consists of firm-year observations that have board independence (i.e. proportion of independent directors on board) lower (higher) than the industry-year median, respectively. We do not find empirical evidence to support our assertion for board independence. Columns 3 and 4 present regression results for the low- and high-institutional ownership ( INSTOWN ) sub-sample. The low- (high-) INSTOWN sub-sample includes observations where the INSTOWN is lower (higher) than the industry-year median. As shown in Column 3, we find a positive and marginally significant (β = 0.004, p < 0.10) association between STRAT _ DEV and INV for the low- INSTOWN sub- sample, suggesting that strategic deviation increases investment inefficiencies in the presence of weaker external monitoring (i.e. low INSTOWN ). The coefficient on STRAT _ DEV for the high- INSTOWN sub-sample is insignificant. The results of board independence and institutional ownership, therefore, provides partial support for H2. Then, we examine the moderating effect of product market competition on the association between STRAT _ DEV and INV . We proxy product market competition using the fluidity meas- ure (FLUID) developed by Hoberg et al. (2014). The data are sourced from the Hoberg-Phillips Data Library. The low- (high-) product-market-competition sub-sample includes observations that 18 Australian Journal of Management 00(0) Table 6. Strategic deviation and investment efficiency: Cross-sectional tests. 1 2 3 4 5 6 7 8 9 10 BIND INSTOWN COMPETITION ANALYST FCOMP Low High Low High Low High Low High Low High STRAT_DEV 0.001 0.003 0.004* –0.000 0.001 0.004*** 0.003* 0.001 0.004*** –0.001 (0.001) (0.002) (0.002) (0.002) (0.001) (0.001) (0.002) (0.001) (0.001) (0.001) MTB 0.000*** 0.000* 0.000 0.000* 0.000*** 0.000 –0.000** 0.000*** 0.000 0.000 (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) SIZE 0.017*** 0.010*** 0.010*** 0.008*** 0.012*** 0.014*** 0.010*** 0.001 0.017*** 0.011*** (0.001) (0.002) (0.002) (0.002) (0.001) (0.001) (0.002) (0.002) (0.001) (0.001) CFL_VOL 0.001 0.001 0.003* 0.005* 0.001 –0.000 0.002* –0.001 0.001 0.003*** (0.001) (0.001) (0.001) (0.003) (0.001) (0.000) (0.001) (0.001) (0.001) (0.001) SALES_VOL 0.045*** 0.017 0.033** 0.040** 0.044*** 0.027*** 0.037*** –0.006 0.047*** 0.002 (0.013) (0.018) (0.015) (0.019) (0.007) (0.008) (0.011) (0.013) (0.011) (0.010) INV_VOL 0.004*** –0.005*** 0.002 –0.026 0.003*** 0.002 0.002*** –0.030*** –0.001 0.003*** (0.001) (0.001) (0.001) (0.016) (0.000) (0.002) (0.000) (0.010) (0.001) (0.001) TANG 0.018 0.033* 0.011 0.022 0.092*** 0.076*** 0.040** 0.062*** 0.035** 0.060*** (0.016) (0.018) (0.023) (0.022) (0.011) (0.013) (0.016) (0.016) (0.015) (0.014) CFOSALES –0.024 0.001 –0.028 –0.019 0.023* –0.037*** 0.013 –0.050** –0.018 –0.085*** (0.018) (0.023) (0.025) (0.023) (0.013) (0.013) (0.017) (0.020) (0.016) (0.017) SLACK –0.000* –0.000 –0.000* –0.001*** –0.000** –0.000*** –0.001** 0.000 –0.000 –0.001*** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) LOSS –0.008 0.007 0.000 0.001 –0.015*** –0.014** –0.019*** 0.006 –0.008* –0.013*** (0.006) (0.007) (0.007) (0.007) (0.003) (0.004) (0.005) (0.007) (0.004) (0.005) AGE –0.023*** –0.029*** –0.015** –0.027*** –0.009*** –0.014*** –0.012*** –0.006 –0.014*** –0.014*** (0.006) (0.007) (0.008) (0.007) (0.003) (0.003) (0.004) (0.004) (0.004) (0.004) ZSCORE –0.026*** –0.028*** –0.022*** –0.030*** –0.022*** –0.018*** –0.038*** –0.024*** –0.024*** –0.035*** (0.003) (0.003) (0.003) (0.004) (0.002) (0.002) (0.003) (0.004) (0.002) (0.003) DIV –0.011** –0.029*** –0.013** –0.003 –0.015*** –0.023*** –0.013*** –0.041*** –0.021*** –0.032*** (0.005) (0.005) (0.006) (0.006) (0.003) (0.004) (0.004) (0.005) (0.004) (0.004) OPCYCLE –0.013*** –0.010* –0.023*** –0.005 0.001 –0.010*** –0.012*** –0.003 –0.015*** –0.006* (0.004) (0.005) (0.007) (0.005) (0.003) (0.003) (0.004) (0.004) (0.004) (0.004) |DACC| 0.171*** 0.164*** 0.120*** 0.195*** 0.115*** 0.190*** 0.203*** 0.249*** 0.148*** 0.196*** (0.027) (0.032) (0.037) (0.033) (0.015) (0.017) (0.021) (0.027) (0.020) (0.019) (Continued) Ranasinghe and Habib 19 Table 6. (Continued) 1 2 3 4 5 6 7 8 9 10 BIND INSTOWN COMPETITION ANALYST FCOMP Low High Low High Low High Low High Low High First step control SALESG 0.408*** 0.108 –0.185** 0.686*** 0.048 0.096* 0.109 0.272 –0.029 0.191** (0.068) (0.091) (0.087) (0.049) (0.038) (0.050) (0.087) (0.181) (0.058) (0.096) Industry effects Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Year effects Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes SALESG*Industry Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes effects SALESG*Year effects Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Constant 0.127*** 0.085 0.164** 0.175*** 0.072** 0.064* 0.221*** 0.218*** 0.066* 0.332*** (0.032) (0.060) (0.071) (0.046) (0.032) (0.033) (0.045) (0.057) (0.039) (0.045) Observations 10,460 8,226 7,078 8,090 24,594 28,610 15,477 12,289 17,195 16,118 Adjusted R 0.272 0.260 0.316 0.333 0.301 0.262 0.323 0.268 0.321 0.259 Note: This table presents the cross-sectional test results. Columns 1 and 2 shows the association between STRAT_DEV and INV for Low and High board independence (BIND) sub-samples. Columns 3 and 4 shows the association between STRAT_DEV and INV for Low and High institutional ownership (INSTOWN) sub-samples. Columns 5 and 6 show regression estimates for the association between STRAT_DEV and INV for the Low and High competition (COMPETITION) sub-sample, when the competition is measured using product market fluidity. Columns 7 and 8 show the association between STRAT_DEV and INV for the Low and High sub-samples of number of analysists following (ANALYST). Columns 9 and 10 show the regression estimates of the association between STRAT_DEV and INV for the Low and High sub-samples of financial statement comparability (FCOMP). Robust standard errors clustered at firm-level are in parentheses *, **, *** denote a two-tailed p-value of less than 0.10, 0.05 and 0.01, respectively. Variables are defined in Appendix 1. 20 Australian Journal of Management 00(0) have a FLUID value lower (higher) than the industry-year median FLUID. The coefficient on STRAT _ DEV is positive and significant for the high-product-market-competition sub-sample, (β = 0.004, p < 0.01) (Column 6), while this is insignificant for the low-product-market-competi- tion sub-sample (Column 5). This finding suggests that the escalated risks and agency problems stemming from high competition are associated with sub-optimal investments by deviant firms and supports signal precision view of H3. Finally, to ascertain the moderating effects of the information environment on the association between STRAT _ DEV and INV , first we divide the sample into low and high analyst following. The low- (high-) analyst-following sub-sample includes observations with analyst following ( LNANALYST ) lower (higher) than the industry median for each year. Results show that the coef- ficient on STRAT _ DEV is positive and significant for the low-analyst-following sub-sample (β = 0.003, p < 0.10) (Column 7). However, the coefficient is insignificant for the high-analyst- following sub-sample (Column 8). This supports the view that strategic deviation increases infor- mation opaqueness and, hence, managerial opportunism and consequently sub-optimal investments. The result, therefore, supports H4a. We also investigate whether the association between STRAT _ DEV and INV is stronger when the financial statement comparability () FCOMP is lower. We follow De Franco et al. (2011) and measure financial statement comparability using time-series regression. We use INDC _ OMP , the median financial statement comparability for all firms j in the same industry as firm during year t, as our measure of financial statement com- parability. The low- (high-) FCOMP sub-sample includes observations that have an INDC _ OMP value lower (higher) than the industry-year median. As shown in Column 9, the coefficient on STRAT _ DEV is positive and significant for the low-FCOMP sub-sample only (β = 0.004, p < 0.01), thereby supporting H4b. 4.5. Robustness tests In our first robustness test, we develop an alternative investment inefficiency measure (_ ALTINV _) INEFF following Eisdorfer et al. (2013). First, we determine the actual invest- ments for each firm-year by scaling the gross CAPEX by the book value of total assets at the begin- ning of the year. We then determine expected investment where the expected investment is calculated as the median investment in the four-digit SIC industries. Then we estimate the abnor- mal investments by subtracting actual investments from expected investments. We use the absolute value of abnormal investment and denote this as ALTI __ NV INEFF. Table 7 shows the coeffi- cient on ALTI __ NV INEFF positive and significant (β = 0.313, p < 0.01) (Column 1). Second, we use an alternative measure of strategic deviation ( ALTS __ TRAT DEV ) and re- estimate our baseline model and report the result in Column 2 in Table 7. We follow prior research (Dong et al., 2021; Finkelstein and Hambrick, 1990) and construct ALTS __ TRAT DEV , exclud- ing R&D intensity and advertising intensity from the STRAT _ DEV measure. We regress INV on ALTS __ TRAT DEV , a set of control variables, the first stage control variable (SALESG), industry and year indicators, SLAESG*industry effects and SALESG*year effects. We find sup- port for the assertion that STRAT _ DEV and INV are positively associated using this alternative proxy of strategic deviation (β = 0.002, p < 0.05). 4.6. Strategic deviation, investment and firm value Finally, we examine whether the investment inefficiencies that arise due to strategic deviation have negative consequences for firm value. We use the following regression to test this proposition: Ranasinghe and Habib 21 Table 7. Strategic deviation and investment efficiency: robustness tests. (1) (2) ALT_INV_INEFF INV STRAT_DEV 0.313*** (0.044) ALT_STRAT_DEV 0.002** (0.001) MTB 0.000 0.000*** (0.000) (0.000) SIZE 0.205*** 0.015*** (0.039) (0.001) CFL_VOL 0.074*** 0.000 (0.027) (0.000) SALES_VOL 2.493*** 0.030*** (0.306) (0.005) INV_VOL –0.0155 0.003*** (0.016) (0.001) TANG 13.235*** 0.077*** (0.754) (0.008) CFOSALES –1.063 –0.029*** (0.943) (0.009) SLACK 0.004* –0.002* (0.002) (0.000) LOSS –0.813*** –0.016*** (0.181) (0.002) AGE –0.673*** –0.014*** (0.176) (0.002) ZSCORE 0.258*** –0.021*** (0.082) (0.001) DIV –1.351*** –0.023*** (0.108) (0.002) OPCYCLE 0.715*** –0.006*** (0.240) (0.002) |DACC| 5.887*** 0.160*** (0.710) (0.011) First step control SALESG – 0.057** (0.029) Industry effects Yes Yes Year effects Yes Yes SALESG*Industry effects – Yes SALESG*Year effects – Yes Constant –2.737 0.099*** (2.242) (0.025) Observations 64,510 59,423 Adjusted R 0.069 0.278 Note: This table shows the pooled regression estimates of the association between strategic deviation and investment using alternative measures. Column 1 shows the association between STRAT_DEV and ALT_INV_INEFF. Column 2 shows the association between ALT_STRAT_DEV and INV. Robust standard errors clustered at firm-level are in parentheses. *, **, *** denote a two-tailed p-value of less than 0.10, 0.05 and 0.01, respectively. Variables are defined in Appendix 1. 22 Australian Journal of Management 00(0) VALUE =+ ββ STRAT __ DEVI ++ ββ NV STRAT DEVI * NV it ,, +10 12 it it ,, 3 it (3) + λλ CONTROLS ++ INDUSTRYY λε EAR + ii,, t jj t kk,, ti t ∑ ∑∑ ∑ i j k Where VALUE is firm value proxied by Peters and Taylor (2017) investment-Q (QTOT) and TOBINQ measures and other variables are defined as before. We expect the coefficient β to be negative and significant if the capital market discounts the sub-optimal investments undertaken by the strategically deviant firms. We present the regression results in Table 8. Columns 1 and 2 report the result for QTOT and TOBINQ measures respectively. We find the coefficient on the interactive variable STRAT_DEV*INV negative and significant for both QTOT (coefficient -0.043, p < 0.01) and TOBINQ (β = -0.345, p < 0.01) suggesting that the capital market discounts the investments undertaken by the deviant firms. 5. Conclusion We explore the association between strategic deviation and investment inefficiency. It is important to understand the repercussions of deviating from the industry-level strategy. At the same time, it is of vital importance to know the direct antecedents of investment inefficiencies. We argue that firms deviating from industry norms are prone to increased information asymmetry and hence, are able to engage in self-serving behaviour as manifested in inefficient investments. Using a sample of U.S. data, we find support for our prediction. Then we examine the moderating role of weak monitoring, high competition and a low-quality information environment. Our findings suggest that the positive association between strategic deviation and investment inefficiency is pronounced for firms exposed to weak monitoring and with a low-quality information environment and for firms operating in a highly competitive environment. Our results remain robust to possible endo- geneity concerns. By examining the association between strategic deviation and investment inef- ficiency, we contribute to both the strategic management and the investment literature. Although there are a number of studies that examine the determinants of investment inefficiencies, there is no evidence on the association between deviant strategies and investment inefficiencies. This study fills this gap in the literature. Business strategy is a choice that managers have to make, and, hence, it is important to under- stand the managerial incentives behind choosing a particular business strategy. As elaborated above, deviating from industry peers results in negative consequences, such as high information asymmetry, high risk and uncertainty. On the other hand, conforming to industry peers bring about homogeneity and is preferred by capital market participants due to low information processing costs and low uncertainty (DiMaggio and Powell, 1983). Litov et al. (2012) argue that pursuing unique strategies brings about economic rents which are associated with firm value in the long term. In contrast, Navissi et al. (2017) find that sub-optimal investments resulting from different strategic choices adversely affect future performance. Similarly, Dong et al. (2021) find that the capital market discounts the value of cash holdings of firms pursuing non-conforming strategies. Given these mixed implications, managers face a paradox in selecting a strategy (Litov et al., 2012). Despite the negative implications, managers might pursue deviant strategies when the costs of being different is lower than the benefits of conforming to industry standards. In other words, such firms may confront information asymmetry and thereby negative capital market consequences in the short term, but may enjoy first-mover benefits in the long term. Alternatively, the impact of strategic deviance on firm-level outcomes may be stronger in firms with peculiar characteristics. Ranasinghe and Habib 23 Table 8. Strategic deviation, investment and firm value. (1) (2) QTOT TOBINQ t + 1 t + 1 STRAT_DEV 0.396*** 3.392*** (0.050) (0.258) INV 0.055*** 0.188*** (0.006) (0.028) STRAT_DEV*INV –0.043*** –0.345*** (0.014) (0.072) MTB 0.000* 0.001*** (0.000) (0.000) SIZE 0.232*** 0.872*** (0.006) (0.029) CFL_VOL 0.020*** 0.096*** (0.002) (0.011) SALES_VOL 0.235*** 1.093*** (0.043) (0.209) INV_VOL –0.002 –0.013 (0.006) (0.025) TANG –0.431*** –8.865*** (0.066) (0.298) CFOSALES 0.197*** 0.877** (0.076) (0.350) SLACK 0.004*** 0.064*** (0.001) (0.006) LOSS –0.124*** –0.339*** (0.015) (0.076) AGE 0.049** –0.177* (0.023) (0.102) ZSCORE –0.022** –0.458*** (0.009) (0.053) DIV –0.256*** –1.207*** (0.027) (0.117) OPCYCLE –0.009 –0.367*** (0.015) (0.078) |DACC| 0.682*** 2.263*** (0.061) (0.291) Industry effects Yes Yes Year effects Yes Yes Constant 0.187 4.540*** (0.312) (0.824) Observations 50,470 50,263 Adjusted R 0.244 0.387 Note: This table shows the association between investment and firm value when strategic deviance exists. Column 1 shows results when investment-Q model is used, following Peters and Taylor (2017). Column 2 shows the results when firm value is measured using TOBINQ. Robust standard errors clustered at firm-level are in parentheses. *, **, *** denote a two-tailed p-value of less than 0.10, 0.05 and 0.01, respectively. Variables are defined in Appendix 1. 24 Australian Journal of Management 00(0) For example, Tang et al. (2011) find that firms with dominant CEOs pursue deviant strategies and encounter extreme performance. In this study, we focus on whether strategically deviant firms are associated with sub-optimal investments as a first step to understanding the implications of strate- gic choice. Acknowledgement We acknowledge the Associate Editor, Professor Tom Smith, Deputy Editor (Finance), Associate Professor Chelsea Liu and the anonymous reviewer for their valuable comments. Funding The author(s) received no financial support for the research, authorship, and/or publication of this article. ORCID iDs Dinithi Ranasinghe https://orcid.org/0000-0002-2102-9481 Ahsan Habib https://orcid.org/0000-0003-2433-3961 Notes 1. As we use scaled investment as our dependent variable, control for the independent variable from the first stage regression and include several interaction terms as suggested by Jackson (2022) for perform- ing the one-stage regression model, we continue to use the term investment in(efficiency) for the remain- der of the paper but use INV instead of INV_INEFF. 2. However, Ye et al. (2021) find that firms pursuing a deviant strategy have more firm-specific information impounded into their stock prices and hence such firms have less synchronous stock price movement. This occurs because strategically deviant firms issue more managerial earnings forecasts and have a higher level of block ownership than the nondeviant firms. 3. We replace missing R&D and advertising expenses with zero consistent with prior research. 4. We do not use first step controls and the industry and year interactions in the 2SLS model because we use IVREG2 in STATA to estimate the 2SLS model, which supposedly makes standard error corrections and accounts for estimation issues (see Chen et al., 2022). 5. Firm location data (latitudes and longitudes) come from Professor Bill McDonald’s website (https:// www3.nd.edu/~mcdonald/). 2 2 (( COSa22 SINb −+ bC 1)) ( OSaS 12 INaS − INa1 12 COSa COSb () 21 − b ) 6. , we Distancea = 3963.* 19 rctan SINa SINa COSa COSa COSb () − b 12 12 21 multiply the value by negative one to convert into proximity. Distance 7. See http://hobergphillips.tuck.dartmouth.edu/. 8. See Appendix 2 for a detailed explanation of financial statement comparability calculation. 9. See the Appendix 1 for the detailed variable calculation. References Abdoh H and Varela O (2017) Product market competition, idiosyncratic and systematic volatility. Journal of Corporate Finance 43: 500–513. Al-Hadi A, Habib A, Taylor G, et al. (2021) Financial statement comparability and corporate investment efficiency. Meditari Accountancy Research 29: 1283–1313. Arellano M and Bover O (1995) Another look at the instrumental variable estimation of error-components models. Journal of Econometrics 68: 29–51. Babar M and Habib A (2021) Product market competition in accounting, finance, and corporate governance: A review of the literature. International Review of Financial Analysis 73: 101607. Ranasinghe and Habib 25 Bebchuk LA and Stole LA (1993) Do short-term objectives lead to under-or overinvestment in long-term projects? The Journal of Finance 48: 719–729. Bentley KA, Omer TC and Sharp NY (2013) Business strategy, financial reporting irregularities, and audit effort. Contemporary Accounting Research 30: 780–817. Bentley-Goode KA, Omer TC and Twedt BJ (2019) Does business strategy impact a firm’s information envi- ronment? Journal of Accounting, Auditing & Finance 34: 563–587. Biddle GC, Hilary G and Verdi RS (2009) How does financial reporting quality relate to investment effi- ciency? Journal of Accounting and Economics 48: 112–131. Blundell R and Bond S (1998) Initial conditions and moment restrictions in dynamic panel data models. Journal of Econometrics 87: 115–143. Brennan MJ and Subrahmanyam A (1995) Investment analysis and price formation in securities markets. Journal of Financial Economics 38: 361–381. Carpenter MA (2000) The price of change: The role of CEO compensation in strategic variation and deviation from industry strategy norms. Journal of Management 26: 1179–1198. Chen M-J and Hambrick DC (1995) Speed, stealth, and selective attack: How small firms differ from large firms in competitive behavior. Academy of Management Journal 38: 453–482. Chen T, Xie L and Zhang Y (2017) How does analysts’ forecast quality relate to corporate investment effi- ciency? Journal of Corporate Finance 43: 217–240. Chen W, Hribar P and Melessa S (2018) Incorrect inferences when using residuals as dependent variables. Journal of Accounting Research 56: 751–796. Chen W, Hribar P and Melessa S (2022) On the use of residuals as dependent variables. Journal of Financial Reporting 7: 69–83. Choi JH, Choi S, Myers LA, et al. (2019) Financial statement comparability and the informativeness of stock prices about future earnings. Contemporary Accounting Research 36: 389–417. Christodoulou D, Ma L and Vasnev A (2018) Inference-in-residuals as an estimation method for earnings management. Abacus 54: 154–180. Chung KH, McInish TH, Wood RA, et al. (1995) Production of information, information asymmetry, and the bid-ask spread: Empirical evidence from analysts’ forecasts. Journal of Banking & Finance 19: 1025–1046. Deephouse DL (1996) Does isomorphism legitimate? Academy of Management Journal 39: 1024–1039. Deephouse DL (1999) To be different, or to be the same? It’s a question (and theory) of strategic balance. Strategic Management Journal 20: 147–166. De Franco G, Kothari SP and Verdi RS (2011) The benefits of financial statement comparability. Journal of Accounting Research 49: 895–931. DiMaggio PJ and Powell WW (1983) The iron cage revisited: Institutional isomorphism and collective ration- ality in organizational fields. American Sociological Review48: 147–160. Dong X, Chan KC, Cui Y, et al. (2021) Strategic deviance and cash holdings. Journal of Business Finance & Accounting 48: 742–782. Eisdorfer A, Giaccotto C and White R (2013) Capital structure, executive compensation, and investment efficiency. Journal of Banking & Finance 37: 549–562. Fama EF and Jensen MC (1983) Separation of ownership and control. The Journal of Law and Economics 26: 301–325. Ferreira MA and Matos P (2008) The colors of investors’ money: The role of institutional investors around the world. Journal of Financial Economics 88: 499–533. Financial Accounting Standards Board (FASB) (2010) Statement of financial accounting concepts No. 8. Conceptual framework for financial reporting. Norwalk, CT: FASB. Finkelstein S and Hambrick DC (1990) Top-management-team tenure and organizational outcomes: The moderating role of managerial discretion. Administrative Science Quarterly 35: 484–503. Francis JR, Pinnuck ML and Watanabe O (2014) Auditor style and financial statement comparability. The Accounting Review 89: 605–633. Geletkanycz MA and Hambrick DC (1997) The external ties of top executives: Implications for strategic choice and performance. Administrative Science Quarterly 42: 654–681. 26 Australian Journal of Management 00(0) Hainmueller J (2012) Entropy balancing for causal effects: A multivariate reweighting method to produce balanced samples in observational studies. Political Analysis 20: 25–46. Harrington JR and Gelfand MJ (2014.) Tightness–looseness across the 50 united states. Proceedings of the National Academy of Sciences 111: 7990–7995. Hoberg G, Phillips G and Prabhala N (2014) Product market threats, payouts, and financial flexibility. The Journal of Finance 69: 293–324. Holmström B (1999) Managerial incentive problems: A dynamic perspective. The Review of Economic Studies 66: 169–182. Imhof MJ, Seavey SE and Watanabe OV (2022) Competition, proprietary costs of financial reporting, and financial statement comparability. Journal of Accounting, Auditing & Finance 37: 114–142. Irvine PJ and Pontiff J (2008) Idiosyncratic return volatility, cash flows, and product market competition. The Review of Financial Studies 22: 1149–1177. Jackson A (2022) Residuals from two-step research designs. Accounting & Finance 62: 4345–4358. Jensen MC (1986) Agency costs of free cash flow, corporate finance, and takeovers. The American Economic Review 76: 323–329. Jensen MC and Meckling WH (1976) The theory of the firm: managerial behavior, agency costs, and owner- ship structure. Journal of Financial Economics 3(4): 305–360. Jiang F, Kim KA, Nofsinger JR, et al. (2015) Product market competition and corporate investment: Evidence from China. Journal of Corporate Finance 35: 196–210. Kang YJ, Lee H-Y, Park H-Y, et al. (2021) Social ties, managerial overconfidence, and investment efficiency. Finance Research Letters 46: 102300. Kim J-B, Li L, Lu LY, et al. (2016) Financial statement comparability and expected crash risk. Journal of Accounting and Economics 61: 294–312. Laksmana I and Yang Y-w (2015) Product market competition and corporate investment decisions. Review of Accounting and Finance 14: 128–148. Lin Y-E, Li Y-W, Cheng TY, et al. (2021) Corporate social responsibility and investment efficiency: Does business strategy matter? International Review of Financial Analysis 73: 101585. Litov LP, Moreton P and Zenger TR (2012) Corporate strategy, analyst coverage, and the uniqueness para- dox. Management Science 58: 1797–1815. McMullin JL and Schonberger B (2020) Entropy-balanced accruals. Review of Accounting Studies 25: 84– Malmendier U and Tate G (2005) CEO overconfidence and corporate investment. The Journal of Finance 60: 2661–2700. Miles RE and Snow CC (1978) Organizational Strategy, Structure, and Process. New York: McGraw Hill. Mintzberg H (1978) Patterns in strategy formation. Management Science 24: 934–948. Modigliani F and Miller MH (1958) The cost of capital, corporation finance and the theory of investment. The American Economic Review 48: 261–297. Myers SC (1977) Determinants of corporate borrowing. Journal of Financial Economics 5: 147–175. Navissi F, Sridharan V, Khedmati M, et al. (2017) Business strategy, over-(under-) investment, and manage- rial compensation. Journal of Management Accounting Research 29: 63–86. Peters RH and Taylor LA (2017) Intangible capital and the investment-q relation. Journal of Financial Economics 123: 251–272. Pool VK, Stoffman N and Yonker SE (2015) The people in your neighborhood: Social interactions and mutual fund portfolios. The Journal of Finance 70: 2679–2732. Porter ME (1980) Competitive Strategy: Techniques for Analyzing Industries and Competitors. New York: The Free Press. Porter ME (1986) Changing patterns of international competition. California Management Review 28: 9–40. Porter ME (1996) What is strategy? Harvard Business Review, November–December, 37–55. Provaty S, Amin S and Hasan MM (2022) Strategic deviation and debt maturity structure. Finance Research Letters 50: 103317. Ranasinghe and Habib 27 Rajkovic T (2020) Lead independent directors and investment efficiency. Journal of Corporate Finance 64: Roulstone DT (2003) The relation between insider-trading restrictions and executive compensation. Journal of Accounting Research 41: 525–551. Shleifer A and Vishny RW (1989) Management entrenchment: The case of manager-specific investments. Journal of Financial Economics 25: 123–139. Shleifer A and Vishny RW (1997) A Survey of Corporate Governance. The Journal of Finance 52(2): 737–783. Stock J and Yogo M (2005) Asymptotic distributions of instrumental variables statistics with many instruments. In: Andrews DWK, Stock JH and Rothenberg TJ (eds) Identification and Inference for Econometric Models: Essays in Honor of Thomas Rothenberg, pp. 109–120. Cambridge: Cambridge University Press. Stoughton NM, Wong KP and Yi L (2017) Investment efficiency and product market competition. Journal of Financial and Quantitative Analysis 52: 2611–2642. Tang J, Crossan M and Rowe WG (2011) Dominant CEO, deviant strategy, and extreme performance: The moderating role of a powerful board. Journal of Management Studies 48: 1479–1503. Ye K, Juan JX and Zhang B (2021) Strategic deviation and stock return synchronicity. Journal of Accounting, Auditing & Finance 36: 172–194. Appendix 1 Variable definitions. Variable Definition Dependent variable INV INV is total investments, calculated as the sum of research and development (R&D) expenditure, capital expenditure (CAPEX) and acquisition expenditure (AQC), less cash receipts from sale of property, plant and equipment (SPPE), and scaled by the lagged total assets (AT). Explanatory variable STRAT_DEV Strategic deviation, measured as the extent to which a firm’s strategy deviates from industry average. See Section 3.2.2. for detailed estimation procedure. Control variables MTB Market-to-book ratio, measured as market value of equity (PRCC_F*CSHO) scaled by book value of equity (CEQ). SIZE The natural logarithm of the market value of equity. CFO_VOL Standard deviation of the cash flow from operations (OANCF) deflated by average total assets for the period t-5 to t-1. SALES_VOL Standard deviation of the sales deflated by average total assets for the period t-5 to t-1. INV_VOL Standard deviation of the investment defined in Section 3.2 for the period t-5 to t-1. TANG The ratio of net property, plant and equipment to total assets. CFOSALES The ratio of the cash flow from operations to sales. SLACK The ratio of cash and short-term investments (CHE) to net property, plant and equipment (PPENT). LOSS An indicator variable coded 1 for firms incurring losses, and zero otherwise. AGE The natural logarithm of firm age estimated as the number of years since the firm’s initial appearance in Compustat annual file. (Continued) 28 Australian Journal of Management 00(0) Appendix 1. (Continued) Variable Definition ZSCORE (3.3*Pretax Income/Total Assets + Sales/Total Asset + 0.25*Retained Earnings/ Total Assets + 0.5*Working Capital/Total Assets) (Rajkovic, 2020). DIV An indicator variable that equals one for a dividend paying firm, and zero otherwise. OPCYCLE The natural log of receivables (RECT) to sales plus inventory (INVENT) to cost of goods sold (COGS) multiplied by 360. |DACC| Absolute value of performance adjusted discretionary accruals, estimated using the below equation for all firms in the same industry with at least 15 observations for each industry-year       ACC 1 ΔΔ SALESR − ECEIV PPE it , it ,, it it , = γγ + + γ 0 1 2       TA TA TA TA (4) it ,, −− 1 it 1 it , − −− 1 it , 1       + γε ROA +… 31 () it ,, − it where ACC is total accruals calculated as earnings before extraordinary items and discontinued operations minus operating cash flows; TA is total assets in year t-1; ΔSALES is change in sales from year t-1 to year t; ∆RECEIV is change in accounts receivable from year t-1 to year t; PPE is gross property plant & equipment; ROA is return on assets measured as earnings before extraordinary items and discontinued operations for the preceding year divided by total assets for the same year. The coefficient estimates from Equation (4) are used to estimate the non-discretionary component of total accruals (NDAC) for our sample firms. The discretionary accruals are then the residuals from equation (4). First step controls SALESG Sales change from the last year. Instrumental variables LOOSENSS US state-level tightness–looseness index following Harrington and Gelfand (2014). The authors used nine items to create a composite index. Four items reflect strength of punishment: (i) the legality of corporal punishment in schools, (ii) the percentage of students hit/punished in schools, (iii) the rate of executions from 1976 to 2011, and (iv) the severity of punishment for violating laws (i.e., selling, using, or possessing marijuana). Two items reflect latitude/permissiveness: (i) access to alcohol (i.e., ratio of dry to total counties per state) and (ii) the legality of same-sex civil unions. Institutions that reinforce moral order and constrain behaviour were assessed with two items: (i) state-level religiosity and (ii) percentage of individuals claiming no religious affiliation. The final indicator was the percentage of total population that is foreign (p. 7991). PROXIMITY Geographical proximity to the local largest strategic deviant firm (see footnote 4 in-text). Moderating variables BIND Board independence is the proportion of independent directors on board, i.e. the number of independent directors on the board divided by the total number of directors on the board. INSTOWN Institutional ownership is the proportion of top five institutional ownership relative to total ownership. COMPETITION Product market threat proxied by the Fluidity score obtained from Hoberg- Phillips Data Library (Hoberg et al. (2014). ANALYST The natural log of number of analysts following. (Continued) Ranasinghe and Habib 29 Appendix 1. (Continued) Variable Definition FCOMP Firm-year level accounting comparability, which is the industry median of comparability combinations for firm i and other firms in the same two-digit SIC in a given year (see Appendix 2 for detailed estimation). Firm value Measured by scaling firm value by the sum of physical and intangible capital using measures Peters and Taylor (2017) methodology as follows: QTOT (5) Qtot =+ VK /( phy Kint ) it ,, it it ,, it where Qtot is measured by scaling firm value by the sum of the physical and intangible capital. V is firm’s market value defined as market value of equity (PRCC_F*CSHO) plus book value of debt (DLTT + DLC), minus current assets (ACT). Kphy is the replacement value of physical capital (PPEGT). Kint is the replacement cost of intangible capital, which is the sum of a firm’s externally purchased (INTAN) and internally created intangible capital. If INTAN is missing, we set the value to zero. Internally created intangible capital is the sum of knowledge capital (G) and organizational capital (O). (6) GG =− () 1 ∂+ RD & it ,& RD it ,, −1 it (7) OO =− () 1 ∂+ SG & A it ,& SG Ai ,, ti −1 t G is the end-of-period stock of knowledge capital, ∂ is depreciation rate, i, t R&D and R&D is the real research and development expenditure (XRD) for the i,t year. We replace missing XRD with zero following Peters and Taylor (2017). Following Peters and Taylor (2017) we use a 15% depreciation rate for ∂ . R&D Based on the suggestion Peters and Taylor (2017), we set G = 0. O is the i,0 i, t organizational capital, ∂ is depreciation rate, and SG&A is the selling, general SG&A i,t and administrative expenses (XSGA minus XRD minus RDIP) for the year. If XRD is higher than XSGA but is less than COGS, or XGA missing, we measure SG&A i,t as XSGA with no additional adjustments. We set XSGA, XRD and RDIP to zero if missing, following Peters and Taylor (2017). Following Peters and Taylor (2017) we use a 20% depreciation rate for ∂ . We further set O = 0 following Peters SG&A i,0 and Taylor (2017). TOBINQ Market value of equity plus book value of total assets minus book value of equity, divided by book value of total assets at the beginning of fiscal year. Appendix 2 Computation of financial statement comparability We follow De Franco et al. (2011) in measuring financial statement comparability. As presented in equation (8) using firm i ’s 16 previous quarters of earnings and stock returns: EARNINGS =+ αβ RETURN +∈ (8) it ,, ii it it , where EARNINGS is the quarterly net income before extraordinary items scaled by the beginning- of-period market value of equity, and RETURN is the raw stock return during quarter t. and β denote the mapping of firm i ’s economic events (i.e. returns) onto its accounting numbers (i.e. earnings). Firm from the same two-digit industry as firm i has and β , which reflects its ∝ 30 Australian Journal of Management 00(0) economic events to accounting numbers. Next, predicted earnings of firm and are calculated using their respective accounting functions with firm i’s economic events, � � E() EARNINGS =∝ + β RETURN (9) iit ,, i it , � � (10) E() EARNINGS =∝ + β RETURN ij ,,t j it , where E() EARNINGS is firm i’s predicted earnings based on firm i’s accounting function and iit ,, firm i’s return in period t, and E() EARNINGS is firm j’s predicted earnings based on firm j’s ij ,,t accounting function and firm i’s return in period t. The negative value of the average absolute dif- ference between the predicted earnings using firm i’s and firm j’s accounting functions is the pair- wise comparability between firms i and j, FCOMP . ij ,,t (11) FCOMPE =− () EARNINGS − E() EARNINGS ij ,,t ∑ iit ,, ij ,,t t−15 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Australian Journal of Management SAGE http://www.deepdyve.com/lp/sage/strategic-deviation-and-investment-inefficiency-tl4xl0Infv

Loading next page...

References (55)

Jong-Hag Choi, Sunhwa Choi, Linda Myers, D. Ziebart (2017)
Financial Statement Comparability and the Informativeness of Stock Prices About Future Earnings
Econometric Modeling: Capital Markets - Asset Pricing eJournal
David Deephouse (1996)
Does Isomorphism Legitimate?
ERN: Other Organizations & Markets: Policies & Processes (Topic)
G De Franco, SP Kothari, RS Verdi (2011)
The benefits of financial statement comparability
Journal of Accounting Research, 49
Sagira Provaty, S. Amin, M. Hasan (2022)
Strategic deviation and debt maturity structure
Finance Research Letters
Paul Irvine, Jeffrey Pontiff (2005)
Idiosyncratic Return Volatility, Cash Flows, and Product Market Competition
Financial Accounting eJournal
Gary Biddle, G. Hilary, Rodrigo Verdi (2009)
How Does Financial Reporting Quality Relate to Investment Efficiency?
Finance Educator: Courses
Hussein Abdoh, O. Varela (2017)
Product market competition, idiosyncratic and systematic volatility
Journal of Corporate Finance, 43
Y. Kang, H. Lee, Hyun-Young Park, J. Park (2021)
Social ties, managerial overconfidence, and investment efficiency
Finance Research Letters
E. Fama, M. Jensen (1983)
Separation of Ownership and Control
The Journal of Law and Economics, 26
M Arellano, O Bover (1995)
Another look at the instrumental variable estimation of error-components models
Journal of Econometrics, 68
A Al-Hadi, A Habib, G Taylor (2021)
Financial statement comparability and corporate investment efficiency
Meditari Accountancy Research, 29
F. Navissi, V. Sridharan, Mehdi Khedmati, E. Lim, Egor Evdokimov (2017)
Business strategy, over- (under-) investment and managerial compensation
Journal of Management Accounting Research, 29
Tijana Rajkovic (2020)
Lead Independent Directors and Investment Efficiency
S&P Global Market Intelligence Research Paper Series
KH Chung, TH McInish, RA Wood (1995)
Production of information, information asymmetry, and the bid-ask spread: Empirical evidence from analysts’ forecasts
Journal of Banking & Finance, 19
Andrei Shleifer, Robert Vishny (1996)
A Survey of Corporate Governance
Organizations & Markets eJournal
Yu-En Lin, Yi-Wen Li, T. Cheng, K. Lam (2021)
Corporate social responsibility and investment efficiency: Does business strategy matter?
International Review of Financial Analysis, 73
Jeong‐Bon Kim, Leye Li, L. Lu, Yangxin Yu (2016)
Financial Statement Comparability and Expected Crash Risk
Risk Management eJournal
Demetris Christodoulou, Le Ma, A. Vasnev (2018)
Inference‐In‐Residuals as an Estimation Method for Earnings Management
Financial Accounting eJournal
Andrew Jackson (2022)
Residuals from two‐step research designs
Accounting & Finance
R Blundell, S Bond (1998)
Initial conditions and moment restrictions in dynamic panel data models
Journal of Econometrics, 87
DL Deephouse (1999)
To be different, or to be the same? It’s a question (and theory) of strategic balance
Strategic Management Journal, 20
Ming-Jer Chen, D. Hambrick (1995)
Speed, Stealth, and Selective Attack: How Small Firms Differ From Large Firms in Competitive Behavior
Academy of Management Journal, 38
Jianyun Tang, M. Crossan, W. Rowe (2011)
Dominant CEO, Deviant Strategy, and Extreme Performance: The Moderating Role of a Powerful Board
SPGMI: Compustat Fundamentals (Topic)
Jens Hainmueller (2012)
Entropy Balancing for Causal Effects: A Multivariate Reweighting Method to Produce Balanced Samples in Observational Studies
Political Analysis, 20
Marta Geletkanycz, D. Hambrick (1997)
The external ties of top executives: Implications for strategic choice and performance.
Administrative Science Quarterly, 42
Lubomir Litov, P. Moreton, Todd Zenger (2011)
Corporate Strategy, Analyst Coverage, and the Uniqueness Paradox
Corporate Finance: Valuation
Jere Francis, Matt Pinnuck, Olena Watanabe (2013)
Auditor Style and Financial Statement Comparability
SPGMI: Compustat Fundamentals (Topic)
M Babar, A Habib (2021)
Product market competition in accounting, finance, and corporate governance: A review of the literature
International Review of Financial Analysis, 73
PJ DiMaggio, WW Powell (1983)
The iron cage revisited: Institutional isomorphism and collective rationality in organizational fields
American Sociological Review, 48
Miguel Ferreira, Pedro Matos (2008)
The colors of investors’ money: The role of institutional investors around the world
Journal of Financial Economics, 88
Md. Babar, Ahsan Habib (2020)
Product Market Competition in Accounting, Finance, and Corporate Governance: A Review of the Literature
Emerging Markets Economics: Firm Behavior & Microeconomic Issues eJournal
Jeff McMullin, B. Schonberger (2020)
Entropy-balanced accruals
Review of Accounting Studies, 25
Fuxiu Jiang, Kenneth Kim, John Nofsinger, B. Zhu (2015)
Product market competition and corporate investment: Evidence from China
Journal of Corporate Finance, 35
Paul DiMaggio, W. Powell (1983)
THE IRON CAGE REVISITED:
The New Economic Sociology
X Dong, KC Chan, Y Cui (2021)
Strategic deviance and cash holdings
Journal of Business Finance & Accounting, 48
Assaf Eisdorfer, Carmelo Giaccotto, Reilly White (2013)
Capital structure, executive compensation, and investment efficiency
Journal of Banking and Finance, 37
L. Bebchuk, Lars Stole (1993)
Do Short‐Term Objectives Lead to Under‐ or Overinvestment in Long‐Term Projects?
Journal of Finance, 48
W Chen, P Hribar, S Melessa (2022)
On the use of residuals as dependent variables
Journal of Financial Reporting, 7
V. Pool, Noah Stoffman, Scott Yonker (2015)
The People in Your Neighborhood: Social Interactions and Mutual Fund Portfolios: Social Interactions and Mutual Fund Portfolios
Journal of Finance, 70
Xueyan Dong, Kam Chan, Yujia Cui, Jenny Guan (2020)
Strategic Deviance and Cash Holdings
Corporate Finance: Valuation
Mason Carpenter (2000)
The Price of Change: The Role of CEO Compensation in Strategic Variation and Deviation From Industry Strategy Norms
Journal of Management, 26
J. Harrington, M. Gelfand (2014)
Tightness–looseness across the 50 united states
Proceedings of the National Academy of Sciences, 111
JH Choi, S Choi, LA Myers (2019)
Financial statement comparability and the informativeness of stock prices about future earnings
Contemporary Accounting Research, 36
S. Finkelstein, D. Hambrick (1990)
Top-management-team tenure and organizational outcomes: The moderating role of managerial discretion.
Administrative Science Quarterly, 35
KA Bentley-Goode, TC Omer, BJ Twedt (2019)
Does business strategy impact a firm’s information environment?
Journal of Accounting, Auditing & Finance, 34
Kathleen Bentley, Thomas Omer, Nathan Sharp (2013)
Business Strategy, Financial Reporting Irregularities, and Audit Effort
Contemporary Accounting Research, 30
W Chen, P Hribar, S Melessa (2018)
Incorrect inferences when using residuals as dependent variables
Journal of Accounting Research, 56
Ulrike Malmendier, Geoffrey Tate (2002)
CEO Overconfidence and Corporate Investment
AFA 2003 Washington
Gerard Hoberg, G. Phillips, N. Prabhala (2014)
Product Market Threats, Payouts, and Financial Flexibility: Product Market Threats, Payouts, and Financial Flexibility
Journal of Finance, 69
Henry Mintzberg (1978)
Patterns in Strategy Formation
International Studies of Management & Organization, 9
GC Biddle, G Hilary, RS Verdi (2009)
How does financial reporting quality relate to investment efficiency?
Journal of Accounting and Economics, 48
M. Porter (1986)
Changing Patterns of International Competition
California Management Review, 28
R. Peters, Lucian Taylor (2016)
Intangible Capital and the Investment-q Relation
Econometric Modeling: Macroeconomics eJournal
Tao Chen, Lingmin Xie, Yuanyuan Zhang (2017)
How does analysts' forecast quality relate to corporate investment efficiency?
Journal of Corporate Finance, 43
MJ Brennan, A Subrahmanyam (1995)
Investment analysis and price formation in securities markets
Journal of Financial Economics, 38

Publisher: SAGE
Copyright: © The Author(s) 2023
ISSN: 0312-8962
eISSN: 1327-2020
DOI: 10.1177/03128962231152764
Publisher site: See Article on Publisher Site

Abstract

We examine the association between strategic deviation and investment inefficiency. We conceptualize strategic deviation as the extent to which the pattern of a firm’s resource allocation deviates from its industry peers. We posit that firms pursuing deviant strategies are prone to increased information asymmetry and hence, are able to engage in self-serving behaviour as manifested in inefficient investments. Our results suggest that deviant firms have sub-optimal investments. A battery of robustness tests validates our findings. We further provide evidence to suggest that weaker monitoring, high product market competition and a low-quality information environment moderate the relation between strategic deviation and investment inefficiency. JEL Classification: M41, G41 Keywords Information asymmetry, investments, product market competition, strategic deviation 1. Introduction In this study, we examine whether strategic deviation, conceptualized as resource allocation that deviates from industry peers, is associated with investment inefficiency. Managers can avoid competition by pursuing strategic choices that are different from their peers (Porter, 1996), thereby making it difficult for shareholders to evaluate the managerial performance of such firms (Carpenter, 2000). Accordingly, prior research finds that firms that deviate from common industry Corresponding author: Dinithi Ranasinghe, Department of Accountancy and Finance, University of Otago, P.O. Box 56, Dunedin 9054, New Zealand. Email: dinithi.ranasinghe@otago.ac.nz Final transcript accepted on 3 January 2023 by Tom Smith (AE Finance) 2 Australian Journal of Management 00(0) strategies hold cash for opportunistic reasons (Dong et al., 2021), have less synchronous stock returns (Ye et al., 2021) and exhibit extreme performance compared to their industry peers (Tang et al., 2011). We extend the literature on the consequences of strategic deviation by examining its association with firms’ investment decisions. Prior research in this space have primarily captured strategy based on strategy typologies, such as Miles and Snow’s (1978) prospector, analyzer and defender strategies. For example, Navissi et al. (2017) find that prospector (defender)-type firms are more likely to over (under)-invest. They argue that prospector-type firms enjoy greater managerial discretion and less stringent mon- itoring, thereby enabling them to over-invest for self-serving behaviour, including, but not limited to, reducing their career-related risks. Defender-type firms, on the other hand, are subject to a relatively higher level of managerial monitoring and less managerial discretion that results in under-investments. We differentiate our study form Navissi et al. (2017) by examining the consequences of strategic deviation, which is conceptually different from strategy typology. While strategy typology focuses on resource allocation within firms, strategic deviation focuses on an inter-firm strategy perspective and captures the differences in the focal firm’s resource allocation decisions when compared to industry peers (Finkelstein and Hambrick, 1990; Geletkanycz and Hambrick, 1997). While strategy typologies, such as prospector and defender strategies, are entrenched in a firm’s competitive envi- ronment (Porter, 1980), strategic deviation captures a firm’s competitive positioning relative to its peers. Furthermore, researchers arbitrarily choose cut-off scores to categorize firms into prospector, defender and analyzer groups (Bentley et al., 2013; Dong et al., 2021). In contrast, strategic devia- tion focuses on the resource allocation of a firm relative to the industry which reflects how managers pursue strategies in comparison to the commonly adopted industry practices. We complement the strategy-based explanation for corporate investment policies by providing evidence that the adop- tion of a deviant strategy increases investment inefficiencies. Business strategy determines the resource allocations for investments (Miles and Snow, 1978). A strategy that is in line with the industry norms, that is, conforming strategy, is likely to provide benchmarks for evaluation (Porter, 1996). Deviant strategy, that is, a strategy that is different from industry norms, on the other hand, fails to provide appropriate benchmarks for comparisons (Carpenter, 2000). The lack of benchmarks creates information risk and uncertainties for capital market participants. Furthermore, managers of firms pursuing deviant strategies are required to invest in new technology and other projects to attract new customers, segments and markets: investments that reduce the benefit of economies of scale and create higher risks. While some researchers argue that being different has first-entry benefits (Porter, 1980, 1986), we take the view that deviant firms face high risk and uncertainty in cash flow generation and escalate information asymmetry and an opaque information environment (Carpenter, 2000; Deephouse, 1999; Litov et al., 2012). Taken together, the lack of benchmark comparison, information asymmetry and esca- lated needs for capital expenditures among deviant firms incentivize managers to pursue sub- optimal investments. We empirically investigate this proposition. We then explore the settings under which the expected relation between strategic deviation and investment inefficiency might vary cross-sectionally. First, we examine whether internal and exter- nal governance mechanisms moderate this association. Effective governance mechanisms exert effective monitoring (Fama and Jensen, 1983). Shleifer and Vishny (1997) argue that independent directors are effective in monitoring managerial actions. In addition, monitoring by institutional investors reduces over-investments (Ferreira and Matos, 2008). Strategic deviation creates infor- mation asymmetry, and thus constrains strong monitoring compared to other industry peers. Therefore, investment inefficiency for firms with high strategic deviation will be exacerbated by poor monitoring. Ranasinghe and Habib 3 Then, we test whether product market competition moderates the association between strategic deviation and investment inefficiency. A stream of research supports the disciplinary view of prod- uct market competition (Babar and Habib, 2021). However, intense competition increases informa- tion risk because the managers operating in highly competitive industries are reluctant to release proprietary information: an action that hinders obtaining product profitability information ex ante by peers (Stoughton et al., 2017), thus enabling managers of deviant firms to engage in sub-optimal investments. We then investigate the moderating role of information asymmetry on the association between strategic deviation and investment inefficiency. Ye et al. (2021) suggest that deviant firms are dif- ferent from industry peers and, therefore, have high information processing costs and increased information asymmetry. Prior research argues that increased information asymmetries arising from information opaqueness increase investment inefficiency (Chen et al., 2017; Lin et al., 2021). We, therefore, expect that the positive association between strategic deviation and investment ineffi- ciency will be more pronounced in a high-information-asymmetry setting. Finally, we consider the role of financial statement comparability. More comparable financial statements increase the abil- ity of capital market participants to benchmark managerial and firm performance and, as a result, reduces information acquisition and processing costs (Choi et al., 2019; De Franco et al., 2011; Imhof et al., 2022; Kim et al., 2016). As argued before, since strategic deviation enables managers to avoid strict monitoring, the incentives for producing more comparable financial statements are likely lower for such firms. We, therefore, argue that the positive association between strategic deviation and investment inefficiency will be more pronounced when financial statement compa- rability is low. A majority of research on investment efficiency has modelled in(efficient) investment in two- stages with residuals obtained in the first stage being used as the dependent variable in second stage as proxy for investment efficiency (Biddle et al., 2009). However, several studies have questioned the appropriateness of such a research design highlighting biased coefficients and t-statistics, which are not in the expected direction (Chen et al., 2018, 2022; Christodoulou et al., 2018; Jackson, 2022). Jackson (2022) suggests a one-stage regression procedure with a set of indicator variables for industries, years, the independent variable from the first-step regression and interaction terms between each of industries, years and the independent variable from the first-step regression. We follow this procedure in testing our hypothesis. Using a US sample of 56,133 firm-year observations from 1987 to 2020, we document a positive and significant rela- tionship between strategic deviation and investment inefficiency. Fixed effect estimates, two- stage least square (2SLS) estimates, the entropy balancing test and two-step system GMM (generalized method of moments) to allay endogeneity concerns validate our original findings. Furthermore, weaker monitoring, increased information asymmetry and low financial statement comparability exacerbate the positive association between strategic deviation and investment inefficiency. In an additional test, we find that investments by the deviant firms are discounted by the capital market. Our study is motivated based on the calls for investigations into the repercussions of adopting a conforming or a deviant strategy (Deephouse, 1996), as it is important to understand the conse- quences of deviating from industry strategy norms (Chen and Hambrick, 1995). While there are some studies that examine the consequences of strategy typology, only a few have looked into the strategy deviation aspect (Dong et al., 2021; Provaty et al., 2022; Tang et al., 2011; Ye et al., 2021), and We also differ from strategy differentiation, which identifies resource allocation between seg- ments (Dong et al., 2021), and instead focus on inter-firm differences of resource allocation across functions, such as production, marketing, innovation and finance (Finkelstein and Hambrick, 1990). Thus, we capture a firm’s strategy positioning in a competitive market. Furthermore, we 4 Australian Journal of Management 00(0) respond to calls for further research on examining the first-order determinants of investment in(efficiency) (Biddle et al., 2009) by providing evidence from a strategic deviation perspective. We contribute to the strategic management literature by answering the call (Deephouse, 1999) for the implications, investment inefficiency in our case, of a firm being different from its peers. We further extend corporate investment literature by identifying that being different from industry peers increases investment inefficiency. Thereby, we extend strategy and investments literature, such as Navissi et al. (2017). The paper is organized as follows. This introduction is followed by literature and hypotheses development in section 2. Section 3 presents the methodology, while section 4 discusses the empir- ical results and robustness tests. Section 5 concludes the paper. 2. Literature and hypotheses development According to Modigliani and Miller (1958), firms are likely to pursue optimal investment strate- gies in a perfect market, but agency frictions and financial constraints make markets imperfect and hence, affect optimal investment levels (Jensen, 1986; Myers, 1977; Shleifer and Vishny, 1989). Under-investments occur when managers pass on positive NPVs to protect their career concerns, while over-investments occur when they invest in negative NPV projects for empire- building incentives (Biddle et al., 2009). These agency problems manifest through managerial empire building, career motives, herding behaviour and managerial myopia (Bebchuk and Stole, 1993; Holmström, 1999; Jensen, 1986; Malmendier and Tate, 2005). Moral hazard–based agency frictions escalate under-investments, while adverse selection-induced agency frictions increase over-investments. Shleifer and Vishny (1997) suggest that weak corporate governance aggravates agency frictions and, consequently, accentuates investment inefficiencies. Agency theory, there- fore, posits that managers subject to weak monitoring engage in over-investments with the intent of empire building (Jensen, 1986). Strategies determine resource allocation in an entity. Therefore, it is important to understand the type of strategies that firms need to follow to pursue effective resource allocation for efficient investments. Corporate strategy can be defined as a pattern that echoes a series of decisions, which determines product markets, technology deployment, organizational structures and business mod- els (Mintzberg, 1978). Miles and Snow’s (1978) strategy typology suggests that firms adopt strate- gies to remain competitive in the market. According to Miles and Snow (1978), three types of business strategies can exist: prospectors, defenders and analyzers. These strategies vary depend- ing on product markets, processes, organizational structures and technology. Specifically, prospec- tor strategy focuses on innovation and market leadership, while the defender strategy focuses on competition based on price, service or quality. The analyzer strategy falls in between the prospector and defender strategy continuum. Bentley et al. (2013) find that prospector firms are more likely to have financial irregularities, despite high audit efforts, due to the inherent business risk. Prospector firms have high risk and uncertainty relative to defender firms, and, as a result, increase the incremental information acqui- sition costs for the analysts (Bentley-Goode et al., 2019). However, prospector firm managers have incentives to reduce information asymmetry to gain access to financial markets for pursuing their innovative strategies. In line with this argument, Bentley-Goode et al. (2019) find that prospector firms make more management earnings forecasts to attract more analyst coverage compared to defender firm managers: actions that result in lower information asymmetry. As the prospector- type firms seek new innovations, they are more likely to engage in over-investments, while defender-type firms suffer from under-investments (Navissi et al., 2017): sub-optimal investments that is associated with poor future performance. Ranasinghe and Habib 5 Strategy deviation differs from strategy typology in the sense that strategy deviation is based on firms’ preference to conform to industry peers. Based on the institutional theory, managers tend to adopt practices that conform to their peers, which drives inter-organizational homogeneity (DiMaggio and Powell, 1983). As a result, when a strategy deviates from the industry norms, it induces costs to investors. While a deviant strategy can assist firms to explore new markets, build unique customer and supplier relationships and achieve competitive advantage (Porter, 1980, 1986), the non-conformity eliminates the benchmarks for comparison (Carpenter, 2000). This cre- ates higher information processing costs for investors to evaluate firm and managerial performance (Litov et al., 2012). As a result, strategic deviation increases agency costs and hinders firms with high strategic deviation to access external resources at a cheaper cost (Deephouse, 1999), creates information asymmetry and increases cash holdings (Dong et al., 2021). Strategic deviation, therefore, allows managers to pursue self-serving behaviour at the expense of shareholders’ inter- est (Jensen and Meckling, 1976; Jensen, 1986). Furthermore, the lack of benchmarks emanating from high strategic deviation reduces competition and escalates agency problems (Shleifer and Vishny, 1997). Therefore, managers of firms with high strategic deviation are likely to over-invest to capture customers and retain suppliers who otherwise may stick with peers who follow conform- ing strategies. Similarly, deviant firm managers may increase capital expenditures and research and development (R&D) expenditures to seek new customers and technology: investments that are risky and may adversely affect future performance Based on the preceding arguments, we develop the following directional hypothesis: H1: There is a positive association between strategic deviation and investment inefficiency. It is important to understand what factors exacerbate the strategic deviation and investment inef- ficiency association. We argue that corporate governance, product market competition and the quality of the information environment moderate the association between strategic deviation and investment inefficiency. 2.1. Corporate governance Independent monitoring is important to alleviate managerial opportunistic behaviour (Fama and Jensen, 1983). Weakly monitored CEOs may engage in empire building by investing in unprofitable projects, leading to over-investments (Jensen, 1986). Similarly, CEOs are likely to subsidize poorly performing divisions to gain personal benefits (Jensen and Meckling, 1976), which also results in investment inefficiencies. At the same time, excessive monitoring distorts efficient decision-making. Therefore, it is important to implement effective monitoring mechanisms to encourage efficient investments. Independent board members act in the best interest of shareholders by effectively moni- toring managerial investment behaviour, among other activities (Fama and Jensen, 1983). Rajkovic (2020) finds that board independence increases investment efficiency, but board independence com- promised by CEO and director social ties decreases investment efficiency (Kang et al., 2021). Institutional ownership, an external governance mechanism, can also play a role in determining investment efficiencies through stronger monitoring (Biddle et al., 2009). Ferreira and Matos (2008) find lower capital expenditure in firms with high institutional ownership, which suggests that institu- tional ownership reduces over-investments. Therefore, we argue that the positive association in H1 above is likely to be stronger for deviant firms lacking effective governance mechanisms. H2: The positive association between strategic deviation and investment inefficiency is stronger for poorly governed firms. 6 Australian Journal of Management 00(0) 2.2. Product market competition Two competing arguments exist on the association between product market competition and invest- ment efficiency (Babar and Habib, 2021). Product market competition acts as an external govern- ance mechanism, and thus disciplines managerial behaviour. In line with this disciplinary argument, a stream of studies find that product market competition increases capital expenditure and R&D but curbs over-investments (Jiang et al., 2015; Laksmana and Yang, 2015) and increases cash flow–enhancing investments (Abdoh and Varela, 2017). A competing view based on the signal precision perspective posits a negative association between product market competition and invest- ment efficiency. Firms operating in a highly competitive market are reluctant to obtain precise signals about their rivals’ actions due to the marginal effect of a single firm’s signal. As information is costly, managers tend to gather information only when it generates higher profits, ex ante. The impact of one firm’s signal is marginal when there are a lot of firms in the industry. Therefore, there is less incentive for information gathering in a competitive market. Investment efficiency is weaker in such a setting. Stoughton et al. (2017) find support for this prediction. Furthermore, prior research provides evidence on the risk-increasing effect of competition (Irvine and Pontiff, 2008). Given the inconclusive evidence on the association between product market competition and cor- porate investment efficiency (Babar and Habib, 2021), we develop the following hypothesis: H3: Product market competition moderates the association between strategic deviation and investment inefficiency. 2.3. Information environment We examine how the information environment proxied by analyst following and financial state- ment comparability moderates the association between strategic deviation and investment ineffi- ciency. Low analyst following reduces the quality of the information environment because of less transparency and high information processing costs for investors. High analyst following, on the other hand, reduces mispricing and information asymmetry (Brennan and Subrahmanyam, 1995; Chung et al., 1995; Roulstone, 2003). As strategic deviation augments information asymmetry, capital market participants tend to discount firms with unique strategies (Litov et al., 2012). Litov et al. (2012) argue that unique strategies create substantial cost for analysts to understand a firm’s resource allocation behaviour, and hence, such firms are followed by fewer analysts. Weaker monitoring courtesy of a low analyst following for firms with deviant strategies provides incen- tives for managers of such firms to engage in sub-optimal investments. We, therefore, hypothesize the following: H4a: The positive association between strategic deviation and investment inefficiency is stronger for firms with high information asymmetry. Financial statement comparability (Comparability) is the extent of similarity in accounting choices between two or more entities under similar economic conditions (Financial Accounting Standards Board (FASB, 2010). Comparability enables investors to properly identify similarities and differences in firm performance within an industry (Francis et al., 2014). Higher comparability helps outsiders to identify peer firms’ risk of under- and over-investments (i.e. investment ineffi- ciencies) due to the availability of benchmark information. As a result, high comparability reduces agency frictions of empire building and moral hazard behaviour. Al-Hadi et al. (2021) find that comparability increases investment efficiency. On the other hand, low comparability increases information asymmetry and reduces the quality of the information environment (De Franco et al., Ranasinghe and Habib 7 2011). Firms that follow deviant strategies have extremely poor performance (Tang et al., 2011), and thus provide incentives for managers to produce less comparable financial statements to con- ceal this poor performance. As sub-optimal investments by firms following deviant strategies may be associated with poor future performance, managers of such firms are likely to produce less comparable financial statements to conceal such sub-optimal behaviour. We, therefore, hypothe- size as follows: H4b: The positive association between strategic deviation and investment inefficiency is stronger for firms with low financial statement comparability. 3. Methodology 3.1. Sample The sample covers data ranging from 1987 to 2020. We begin with 1987, as the data on cash flow from operations are available on the Compustat dataset from 1987. We obtain governance data from Thomson Reuters and analyst following data from I/B/E/S. Since we require five years of data for calculating strategic deviation and also require five years of lagged values for cash flows to calculate the control variables (e.g. cash flow volatility), our actual regression spans the period 1992 to 2020. We winsorize all the continuous variables at the top and bottom one percentiles to reduce the impact of outliers. We also exclude financial firms (SIC codes between 6000 and 6999). We further exclude observations with missing values for the measurement of key dependent, inde- pendent and control variables. Our final sample consists of 56,133 firm-year observations. The number of observations in any given regression varies depending on the model-specific data requirements. Panel A in Table 1 presents the sample selection procedure, while Panel B provides details of the industry distribution of the sample based on the two-digit SIC industry classification. About 33% observations come from machinery, electrical, computer equipment, followed by busi- ness services (16%) and chemical, petroleum and rubber and allied products (12%). 3.2. Variable measurements 3.2.1. Independent variable: Strategic deviation (STRAT_DEV). We follow previous research (Dong et al., 2021; Ye et al., 2021) and measure strategic deviation (_ STRAT DEV ) using six indicators, namely, (a) Advertising intensity, measured as advertising expenditure (XAD) over sales (REVT); (b) R&D intensity, measured as R&D expenditure scaled by sales; (c) Plant and equipment new- ness, measured as net property, plant and equipment (PPENT) scaled by gross property, plant and equipment (PPEGT); (d) Non-production overhead, measured using selling, general and adminis- trative (SG&A) expenses scaled by sales; (e) Inventory level, measured using inventories (INVT) scaled by sales; and (f) Financial leverage, measured as debt (DLTT) over equity (CEQ). For each fiscal year, we standardize each of the measures by industry (based on two-digit SIC codes) and calculate the absolute difference between a firm’s score and its industry average. STRAT _ DEV is the sum of the six standardized difference scores. A higher value of STRAT _ DEV implies a larger deviation of a firm’s strategy from industry norms. 3.2.2. Dependent variable: Investment ( Following prior research (Biddle et al., 2009; Rajkovic, INV ). 2020), we first use the following regression. INV =+ αβ ΔSales + ε , (1) it 01 it−1 it , 8 Australian Journal of Management 00(0) Table 1. Sample selection and distribution. Panel A: sample selection. Initial sample for the period 1992–2020 238,691 Less: observations in the financial institutions (#60 – #69) (67,490) Less: observations lost for strategy deviation calculation (40,081) Less: observations lost for investment efficiency and control variable calculation (74,987) Final sample 56,133 Panel B: sample distribution by industry. 2-Digit SIC Industry Observations % 01-14 Agriculture & mining 2,276 4.05 15-17 Building construction 444 0.79 20-21 Food & kindred products 1,553 2.77 22-23 Textile mill products & apparels 1,111 1.98 24-27 Lumber, Furniture, Paper, and Printing 2,088 3.72 28-30 Chemical, Petroleum, and Rubber & Allied Products 6,769 12.06 31-34 Metal 2,547 4.54 35-39 Machinery, electrical, computer equipment 18,416 32.81 40-49 Railroad, Communications and Other Transportation 2,541 4.53 50-51 Wholesale goods, building materials 2,783 4.96 53-59 Store merchandise, auto dealers, home furniture stores 3,372 6.01 70-79 Business services 9,085 16.18 80-99 Others 3,148 5.60 Total 56,133 100.00 Note: This table presents the sample selection procedure (Panel A) and sample distribution by industry (Panel B). SIC = Standard Industrial Classification. where INV is total investments, calculated as the sum of R&D expenditure, capital expenditure (CAPEX) and acquisition expenditure (AQC), less cash receipts from sale of property, plant and equipment (SPPE), and scaled by the lagged total assets (AT). ΔSales is the sales change from it−1 the last year (for the remainder of the paper we denote this as SALESG). An overwhelming body of empirical research uses two-step regression to estimate and then examine the key determinant(s) of investment in(efficiency). More specifically, the residuals from Equation (1) above is consid- ered by researchers to represent in(efficient) investment and is used as the dependent variable in the second stage regression. Due to the concerns regarding the use of two-step regression models (see Chen et al., 2018, 2022; Jackson, 2022), we use a one-step regression model which includes the first step control (i.e. SALESG), years and interaction terms between each of industries, years and the independent variable from the first-step regression to alleviate the biased coefficient and other econometrics concerns (Chen et al., 2022; Jackson, 2022). 3.3. Empirical model We test the association between STRAT _ DEV and INV using the following OLS regression model: Ranasinghe and Habib 9 INVS =+ ββ TRAT __ DEVM ++ ββ TB SIZE + β CFLVOL it ,, 01 it 23 it ,, it 4 it , + β SALESV __ OL ++ ββ INVVOL TANG + β CFOSALES 5 56 it ,, it 78 it ,, it + β SLACK KL ++ ββ OSSAGE ++ ββ ZSCORE DIV 9 it ,, 10 it 11 it ,, 12 it 13 it , (2) + β OPCYCLLE ++ ββ || DACC SALESG + λ INDUSTRY 14 it ,, 15 it 16 it ,, jj t + λ YEAR ++ λλ INDUSTRYS ** ALESGYEAR SALESG + ε kk,,t jj,, tj t kk,, ti ti,t ∑ ∑∑ k j k where INV is the investment and STRAT _ DEV is the strategic deviation measure as mentioned above. β is the first-stage variable which is also interacted with INDUSTRY and YEAR varia- bles. A positive and significant coefficient on β would support our H1. Detailed variable defini- tions of the control variables are in Appendix 1. 4. Empirical results 4.1. Descriptive statistics We present descriptive statistics of the variables in Table 2. Results show that the mean and median INV is 0.160 and 0.103 ._ STRAT DEV has a mean and a median of 2.237 and 1.817. The mean (median) of MTB is 3.384 (2.015), CFLV _ OL is 0.579 (0.015), SALESV _ OL is 0.211 (0.147) INVV _ OL and is 0.151 (0.058). Of the sample observations, 36% have reported a loss and about 28% have paid dividends. The average absolute discretionary accruals ( || DACC ) are 10% of lagged total assets. Table 3 shows the correlation coefficients between the variables used in the empirical models. INV and STRAT _ DEV are positively significantly correlated (0.004). The correlation between STRAT _ DEV and all of the independent variables is significant at the 1% level, except INV_ VOL. As per the unreported variance inflation factors (VIFs), the maximum vif is 1.85 with a mean of 1.32, suggesting that multicollinearity is not a concern. 4.2. Regression results Table 4 presents the baseline regression results whereby we regress INV on STRAT _ DEV and a set of control variables with standard errors clustered at the firm level. As elaborated in Section 3.2.2, we use one-step regression model. In Column (1), we present pooled OLS regression results where INV is regressed on STRAT _ DEV , a set of control variables, the first stage control vari- able (SALESG), industry and year indicators, SLAESG*industry effects and SALESG*year effects. The coefficient on STRAT _ DEV is positive and significant (β = 0.003, p < 0.01), sug- gesting a positive relationship between strategic deviation and investment inefficiency, thereby supporting H1. The reported coefficient implies that a one-standard-deviation increase in STRAT _ DEV increases INV by 3.23% relative to mean INV ((0.003*1.725 (S.D. of STRAT_ DEV)/0.16 (mean INV)). In Column 2, we present robust regression results. The coefficient on STRAT _ DEV continues to be positive and significant (β = 0.001, p < 0.01). Among the control variables, firm size (SIZE) is positively and significantly associated with INV. Sales volatility (SALES_VOL) is positively and significantly associated with INV in the pooled model but insig- nificant in the robust model. Other controls, such as, SLACK, ZSCORE and OPCYCLE, are 10 Australian Journal of Management 00(0) Table 2. Descriptive statistics. Variables Obs Mean SD 0.25 0.50 0.75 INV 56,133 0.160 0.194 0.045 0.103 0.202 STRAT_DEV 56,133 2.237 1.725 1.259 1.817 2.594 MTB 56,133 3.384 5.261 1.193 2.015 3.570 SIZE 56,133 5.210 2.214 3.622 5.150 6.732 CFL_VOL 56,133 0.579 4.218 0.002 0.015 0.108 SALES_VOL 56,133 0.211 0.206 0.081 0.147 0.264 INV_VOL 56,133 0.151 1.623 0.026 0.058 0.129 TANG 56,133 0.240 0.214 0.079 0.170 0.336 CFOSALES 56,133 0.005 0.214 –0.003 0.060 0.117 SLACK 56,133 4.206 13.451 0.125 0.623 2.757 LOSS 56,133 0.357 0.479 0.000 0.000 1.000 AGE 56,133 2.913 0.534 2.565 3.045 3.367 ZSCORE 56,133 1.143 1.679 0.637 1.345 1.999 DIV 56,133 0.275 0.447 0.000 0.000 1.000 OPCYCLE 56,133 4.719 0.793 4.322 4.804 5.208 |DACC| 56,133 0.103 0.105 0.031 0.070 0.137 SALESG 56,133 0.230 0.788 –0.033 0.078 0.247 BIND 18,686 0.745 0.124 0.667 0.750 0.833 INSTOWN 15,168 0.367 0.482 0.000 0.000 1.000 COMPETITION 53,204 6.817 3.668 4.160 6.015 8.646 ANALYST 27,766 2.125 0.667 1.609 2.079 2.639 FCOMP 33,313 –3.133 1.686 –3.820 –2.800 –2.070 Note: This table shows summary statistics of the variables used in the regression models. negatively and significantly associated with INV. Discretionary accruals (|DACC|) is positively related to INV. These are consistent with prior research (Biddle et al., 2009; Rajkovic, 2020). 4.3. Addressing endogeneity Although the preceding analyses control for a variety of firm characteristics that might explain the association between STRAT _ DEV and investment inefficiency, endogeneity could be a concern that may bias the findings. Endogeneity could arise from omitted variable concern, reverse causa- tion problem (i.e. firms that are less investment efficient could deviate more from their peers), and design choices. We use several endogeneity tests and report the results in Table 5. First, we esti- mate fixed effect models. As shown in Panel A, our original results hold when we use fixed effects estimates (β = 0.004, p < 0.01). Next, we use the entropy balancing method to address any design choices. McMullin and Schonberger (2020) document that entropy balancing noticeably improves covariate balance when compared with propensity-score approaches. In addition, entropy balanc- ing improves the balance quality without removing any observations from either treatment or con- trol group (Hainmueller, 2012). To conduct entropy balancing estimates, we divide our sample into two groups based on the mean STRAT_DEV. We consider firm-year observations with above (below) mean STRAT_DEV as the treated (control) group. In Panel B, we first report means, vari- ances and skewness of all covariates before and after balancing, and the results suggest a reason- able covariate balance. Using the entropy balanced sample, we then run the baseline regression and find the coefficient on STRAT _ DEV positive and significant (β = 0.004, p < 0.01). Ranasinghe and Habib 11 Table 3. Correlation analysis. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1. INV 1.000 2. STRAT_DEV 0.004 1.000 3. MTB 0.182 0.159 1.000 4. SIZE 0.088 –0.145 0.191 1.000 5. CFL_VOL 0.085 0.079 0.104 –0.114 1.000 6. SALES_VOL –0.021 0.042 0.012 –0.245 0.095 1.000 7. INV_VOL 0.054 –0.001 0.024 –0.006 0.096 0.041 1.000 8. TANG 0.011 0.020 –0.083 –0.002 –0.041 –0.146 0.002 1.000 9. CFOSALES –0.187 –0.195 –0.174 0.294 –0.187 –0.063 –0.037 0.175 1.000 10. SLACK 0.100 0.047 0.077 –0.023 0.141 0.015 0.016 –0.281 –0.247 1.000 11. LOSS 0.106 0.126 0.098 –0.317 0.115 0.075 0.033 –0.060 –0.657 0.134 1.000 12. AGE –0.100 –0.071 –0.043 0.260 –0.085 –0.151 –0.030 0.017 0.204 –0.067 –0.223 1.000 13. ZSCORE –0.289 –0.137 –0.218 0.143 –0.180 0.134 –0.045 0.024 0.675 –0.214 –0.535 0.185 1.000 14. DIV –0.146 –0.038 –0.041 0.339 –0.081 –0.167 –0.027 0.131 0.260 –0.117 –0.311 0.291 0.264 1.000 15. OPCYCLE 0.013 0.050 –0.021 –0.067 0.025 –0.140 –0.000 –0.301 –0.059 –0.015 0.023 0.044 –0.099 –0.036 1.000 16. |DACC| 0.185 0.060 0.133 –0.164 0.100 0.111 0.025 –0.117 –0.281 0.109 0.206 –0.122 –0.232 –0.165 0.042 1.000 Note: This table presents the correlation coefficient between variables used in the base models. Bold-faced correlations are significant at p < 0.01. Variables are defined in Appendix 1. 12 Australian Journal of Management 00(0) Table 4. Baseline regression: Strategic deviation and investment inefficiency. (1) (2) Pooled Robust STRAT_DEV 0.003*** 0.001** (0.001) (0.000) MTB 0.000*** 0.000*** (0.000) (0.000) SIZE 0.016*** 0.009*** (0.001) (0.000) CFL_VOL 0.000 –0.000 (0.000) (0.000) SALES_VOL 0.032*** 0.001 (0.006) (0.001) INV_VOL 0.003*** 0.001*** (0.001) (0.000) TANG 0.077*** 0.089*** (0.008) (0.003) CFOSALES –0.028*** –0.000*** (0.010) (0.000) SLACK –0.000** –0.000*** (0.000) (0.000) LOSS –0.016*** –0.004*** (0.002) (0.001) AGE –0.014*** –0.008*** (0.002) (0.001) ZSCORE –0.020*** –0.006*** (0.001) (0.000) DIV –0.023*** –0.011*** (0.002) (0.001) OPCYCLE –0.007*** –0.002*** (0.002) (0.001) |DACC| 0.162*** 0.045*** (0.011) (0.003) First step control SALESG 0.060** 0.112*** (0.029) (0.024) Industry effects Yes Yes Year effects Yes Yes SALESG*Industry effects Yes Yes SALESG*Year effects Yes Yes Constant 0.097*** 0.1094*** (0.025) (0.024) Observations 56,133 56,133 Adjusted R 0.290 0.570 Note: This table shows the regression estimates of the association between strategic deviation (STRAT_DEV) and investment inefficiency (INV) for the pooled model (column 1) and for the robust model (column 2). Both models include the first step control variable, SALESG, and the interactions between SALESG and Fyear and SALESG and Industry (Chen et al., 2022; Jackson, 2022). Robust standard errors clustered at firm-level are in parentheses. **, *** denote a two-tailed p-value of less than 0.10, 0.05 and 0.01, respectively. Variables are defined in Appendix 1. Ranasinghe and Habib 13 Table 5. Endogeneity tests. Panel A: Fixed effect model. INV STRAT_DEV 0.004*** (0.001) Other control variables Yes First step control SALESG 0.078*** (0.001) Constant –0.117*** (0.015) Industry effects No Year Effects Yes SALESG*Year effects Yes Firm effect Yes Observations 56,133 Adjusted R 0.08 Panel B: Entropy balancing models. Covariate matching table. Before balancing After balancing Treat Control Control Mean Variance Skewness Mean Variance Skewness Mean Variance Skewness MTB 3.534 39.270 5.826 3.235 16.910 6.185 3.534 39.270 5.826 SIZE 4.921 5.048 0.213 5.588 4.811 0.135 4.921 5.048 0.213 CFL_VOL 0.676 19.830 13.810 0.452 14.320 17.700 0.676 19.830 13.810 SALES_VOL 0.227 0.050 2.270 0.196 0.036 2.502 0.227 0.050 2.270 INV_VOL 0.144 2.614 86.320 0.173 4.871 75.040 0.144 2.614 86.320 TANG 0.258 0.050 1.074 0.231 0.043 1.487 0.258 0.050 1.074 CFOSALES –0.012 0.048 –1.688 0.025 0.042 –1.850 –0.012 0.048 –1.688 SLACK 4.527 235.600 6.585 3.538 108.900 8.303 4.527 235.600 6.585 LOSS 0.381 0.236 0.493 0.334 0.223 0.703 0.381 0.236 0.493 AGE 2.900 0.285 –0.647 2.933 0.280 –0.744 2.900 0.285 –0.647 ZSCORE 1.133 3.501 –2.193 1.143 2.092 –2.252 1.133 3.501 –2.193 DIV 0.284 0.203 0.958 0.282 0.202 0.971 0.284 0.203 0.958 OPCYCLE 4.658 0.771 –0.563 4.754 0.499 –0.799 4.658 0.771 –0.563 |DACC| 0.105 0.011 1.993 0.100 0.011 2.070 0.105 0.011 1.993 SALESG 0.160 0.404 5.630 0.179 0.319 6.217 0.160 0.404 5.630 Entropy balancing regression Variable INV STRAT_DEV 0.004*** (0.0007) Controls Yes (Continued) 14 Australian Journal of Management 00(0) Table 5. (Continued) Entropy balancing regression Variable INV First step control SALESG 0.074** (0.033) Constant 0.085*** (0.0198) Industry effects Yes Year Effects Yes SALESG*Industry effects Yes SALESG*Year effects Yes Observations 56,133 Adjusted R 0.301 Panel C: 2SLS. 1st stage 2nd stage 1st stage 2nd stage Variables regression regression regression regression DV = SRAT_DEV DV = INV DV = SRAT_DEV DV = INV Geographical Proximity index Tightness/Looseness index PROXIMITY 0.001*** (0.000) LOOSENESS 0.057*** (0.013) SRAT_DEV_PRED 0.015* 0.202*** (0.009) (0.055) MTB 0.000*** 0.000 0.001*** –0.000*** (0.000) (0.000) (0.000) (0.000) SIZE –0.048*** 0.020*** –0.061*** 0.031*** (0.004) (0.001) (0.004) (0.003) CFL_VOL 0.002 0.001*** 0.011** –0.001** (0.002) (0.000) (0.002) (0.001) SALES_VOL 0.179*** 0.0324*** 0.202*** –0.006 (0.037) (0.005) (0.034) (0.014) INV_VOL 0.000 0.003*** –0.009** 0.005*** (0.003) (0.000) (0.004) (0.001) TANG –0.631*** 0.040*** –0.571*** 0.169*** (0.052) (0.008) (0.045) (0.034) CFOSALES –0.838*** –0.020* –0.976*** 0.156*** (0.052) (0.011) (0.056) (0.055) SLACK 0.008*** –0.000 0.006*** –0.001*** (0.001) (0.000) (0.001) (0.000) LOSS –0.018 –0.021*** –0.027 –0.020*** (0.019) (0.003) (0.018) (0.004) AGE –0.024* –0.022*** –0.010 –0.017*** (0.015) (0.002) (0.013) (0.003) (Continued) Ranasinghe and Habib 15 Table 5. (Continued) Panel C: 2SLS. 1st stage 2nd stage 1st stage 2nd stage Variables regression regression regression regression DV = SRAT_DEV DV = INV DV = SRAT_DEV DV = INV Geographical Proximity index Tightness/Looseness index ZSCORE –0.141*** –0.024*** –0.129*** 0.002 (0.006) (0.001) (0.006) (0.007) DIV –0.109*** –0.044*** –0.089*** –0.027*** (0.018) (0.003) (0.017) (0.007) OPCYCLE 0.295*** –0.013*** 0.319*** –0.072*** (0.011) (0.003) (0.010) (0.018) |DACC| 0.485*** 0.207*** 0.444*** 0.122*** (0.066) (0.010) (0.064) (0.029) Industry effects Yes Yes Yes Yes Year effects Yes Yes Yes Yes Constant 0.025 –0.341** (0.196) (0.142) Observations 40,951 40,951 55,172 55,172 Adj R 0.176 – 0.310 Kleibergen-Paap rk LM 0.000 0.000 statistic: p-value Weak identification test: Cragg-Donald Wald F-stat. 203.44 19.14 Stock-Yogo critical value 16.38 16.38 Panel D: Two-step system generalized method of moments (GMM). Variables (1) STRAT_DEV 0.012*** (0.003) MTB 0.0010 (0.001) SIZE 0.0446*** (0.005) CFL_VOL 0.0013 (0.001) SALES_VOL 0.0358** (0.017) INV_VOL 0.0048*** (0.001) TANG 0.1193*** (0.039) CFOSALES 0.1192** (0.059) SLACK –0.0011** (0.000) (Continued) 16 Australian Journal of Management 00(0) Table 5. (Continued) Panel D: Two-step system generalized method of moments (GMM). Variables (1) LOSS –0.0832*** (0.012) AGE –0.0395 (0.028) ZSCORE –0.0400*** (0.006) DIV –0.0265** (0.012) OPCYCLE –0.0594*** (0.014) |DACC| 0.2827*** (0.080) First stage control SALESG 0.0724 (0.137) AR(1) (p value) 0.000 AR(2) (p value) 0.445 Industry effects Yes Year effects Yes SALESG*Year effects Yes Observations 56,133 Note: This table shows the endogeneity tests. Panel A presents the fixed effect estimates. Panel B shows the results from the entropy balanced test. Panel C shows the 2SLS models. Panel D presents two-step system GMM results. Robust standard errors clustered at firm-level are in parentheses. *, **, *** denote a two-tailed p-value of less than 0.10, 0.05 and 0.01, respectively. Variables are defined in Appendix 1. GMM = generalized method of moments. Then we perform 2SLS estimates and present the results in Panel C of Table 5. First, we use geographical proximity (PROXIMITY) to local largest strategic-deviant firm as an instrumental variable. Prior research suggests that decision making strategies are similar among companies in the local community (Pool et al., 2015). Therefore, it is more likely that the focal firm pursues a deviant strategy if that firm is geographically closer to the largest strategic-deviant firm in the city. The largest deviant firm in the city is known as the local-largest deviant firm. Following prior research (Dong et al., 2021; Pool et al., 2015), we first find the latitudes and longitudes for all the observations. We then determine the latitudes and longitudes of the largest strategic-deviant firm at the city level. Then we use the Vincenty’s formula to calculate the geographical proximity between the focal firm and the largest strategic-deviant firm based on longitudes and latitudes of the focal firm and the local largest strategic-deviant firm. Column 1 reveals that the instrument PROXIMITY is positively and significantly associated with STRAT_DEV (β = 0.001, p < 0.01), supporting our assertion that geographical proximity to the local largest strategic-deviant firm increases focal firm’s strategic deviation. As shown in the second stage, the STRAT_DEV_PRED is positively and significantly associated with INV (β = 0.015, p < 0.10). The LM statistic reveals that the excluded instruments are ‘relevant’. The Cragg–Donald F-statistic of 203.44 is higher than the Stock and Yogo (2005) critical value of 16.38, implying that our instrument does not suffer from weak identification. Ranasinghe and Habib 17 Second, following Harrington and Gelfand (2014), we use the US state-level tightness-loose- ness index (LOOSENESS) as the instrumental variable. Harrington and Gelfand (2014) argue that the extent of enforcement of laws and regulations among the states are different and hence, the tolerance level of deviation from law implementation differs between the states. Provaty et al. (2022) argue that the state-level looseness (i.e. if the state’s tolerance for lack of law enforcement is high) has a positive association with strategy deviation. We, therefore, use LOOSENESS index as our second instrumental variable. We take the view that when the LOOSENESS score is high, there is more tolerance towards strategic deviation, hence, we expect a positive association between LOOSENESS and STRAT_DEV. As shown in Column 3, we indeed find a positive association between LOOSENESS and STRAT_DEV (β = 0.057, p < 0.01). In Column 4, we find the coeffi- cient on STRAT_DEV_PRED positive and significant (β = 0.202, p < 0.01). The Cragg–Donald F-statistic, however, is much lower than the first instrument. Finally, we use the two-step system GMM (generalized method of moments) approach adopted by Arellano and Bover (1995) and Blundell and Bond (1998) to validate our results documented in Table 4. This should also alleviate any concerns with reverse causality concern. The coefficient on STRAT_DEV is positive and significant (coefficient 0.012, p < 0.01). Given that errors in levels are serially uncorrelated, we expect significant first-order serial correlation, but insignificant sec- ond-order correlation in the first-differenced residuals. Test results reported at the bottom of Panel D confirm the desirable statistically significant AR(1), and statistically insignificant AR(2). Based on these endogeneity tests above, we conclude that our results are not affected by endo- geneity concerns. 4.4. Strategic deviation and investment inefficiency: Cross-sectional tests Table 6 presents the cross-sectional test results for the association between strategic deviation and investment inefficiency. As discussed in Section 2, we consider corporate governance (board inde- pendence and institutional ownership), product market competition, and the quality of the informa- tion environment proxied by analyst following and financial statement comparability as the three cross-sectional settings that are likely to moderate the relation between strategic deviation and investment inefficiency. Regression results are based on a sub-sampling procedure which allows the coefficients on all the control variables to vary between the groups. Columns 1 and 2 present regression estimates of the association between STRAT_DEV and INV for the board independence ( BIND ) context. The low- (high-) BIND sub-sample consists of firm-year observations that have board independence (i.e. proportion of independent directors on board) lower (higher) than the industry-year median, respectively. We do not find empirical evidence to support our assertion for board independence. Columns 3 and 4 present regression results for the low- and high-institutional ownership ( INSTOWN ) sub-sample. The low- (high-) INSTOWN sub-sample includes observations where the INSTOWN is lower (higher) than the industry-year median. As shown in Column 3, we find a positive and marginally significant (β = 0.004, p < 0.10) association between STRAT _ DEV and INV for the low- INSTOWN sub- sample, suggesting that strategic deviation increases investment inefficiencies in the presence of weaker external monitoring (i.e. low INSTOWN ). The coefficient on STRAT _ DEV for the high- INSTOWN sub-sample is insignificant. The results of board independence and institutional ownership, therefore, provides partial support for H2. Then, we examine the moderating effect of product market competition on the association between STRAT _ DEV and INV . We proxy product market competition using the fluidity meas- ure (FLUID) developed by Hoberg et al. (2014). The data are sourced from the Hoberg-Phillips Data Library. The low- (high-) product-market-competition sub-sample includes observations that 18 Australian Journal of Management 00(0) Table 6. Strategic deviation and investment efficiency: Cross-sectional tests. 1 2 3 4 5 6 7 8 9 10 BIND INSTOWN COMPETITION ANALYST FCOMP Low High Low High Low High Low High Low High STRAT_DEV 0.001 0.003 0.004* –0.000 0.001 0.004*** 0.003* 0.001 0.004*** –0.001 (0.001) (0.002) (0.002) (0.002) (0.001) (0.001) (0.002) (0.001) (0.001) (0.001) MTB 0.000*** 0.000* 0.000 0.000* 0.000*** 0.000 –0.000** 0.000*** 0.000 0.000 (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) SIZE 0.017*** 0.010*** 0.010*** 0.008*** 0.012*** 0.014*** 0.010*** 0.001 0.017*** 0.011*** (0.001) (0.002) (0.002) (0.002) (0.001) (0.001) (0.002) (0.002) (0.001) (0.001) CFL_VOL 0.001 0.001 0.003* 0.005* 0.001 –0.000 0.002* –0.001 0.001 0.003*** (0.001) (0.001) (0.001) (0.003) (0.001) (0.000) (0.001) (0.001) (0.001) (0.001) SALES_VOL 0.045*** 0.017 0.033** 0.040** 0.044*** 0.027*** 0.037*** –0.006 0.047*** 0.002 (0.013) (0.018) (0.015) (0.019) (0.007) (0.008) (0.011) (0.013) (0.011) (0.010) INV_VOL 0.004*** –0.005*** 0.002 –0.026 0.003*** 0.002 0.002*** –0.030*** –0.001 0.003*** (0.001) (0.001) (0.001) (0.016) (0.000) (0.002) (0.000) (0.010) (0.001) (0.001) TANG 0.018 0.033* 0.011 0.022 0.092*** 0.076*** 0.040** 0.062*** 0.035** 0.060*** (0.016) (0.018) (0.023) (0.022) (0.011) (0.013) (0.016) (0.016) (0.015) (0.014) CFOSALES –0.024 0.001 –0.028 –0.019 0.023* –0.037*** 0.013 –0.050** –0.018 –0.085*** (0.018) (0.023) (0.025) (0.023) (0.013) (0.013) (0.017) (0.020) (0.016) (0.017) SLACK –0.000* –0.000 –0.000* –0.001*** –0.000** –0.000*** –0.001** 0.000 –0.000 –0.001*** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) LOSS –0.008 0.007 0.000 0.001 –0.015*** –0.014** –0.019*** 0.006 –0.008* –0.013*** (0.006) (0.007) (0.007) (0.007) (0.003) (0.004) (0.005) (0.007) (0.004) (0.005) AGE –0.023*** –0.029*** –0.015** –0.027*** –0.009*** –0.014*** –0.012*** –0.006 –0.014*** –0.014*** (0.006) (0.007) (0.008) (0.007) (0.003) (0.003) (0.004) (0.004) (0.004) (0.004) ZSCORE –0.026*** –0.028*** –0.022*** –0.030*** –0.022*** –0.018*** –0.038*** –0.024*** –0.024*** –0.035*** (0.003) (0.003) (0.003) (0.004) (0.002) (0.002) (0.003) (0.004) (0.002) (0.003) DIV –0.011** –0.029*** –0.013** –0.003 –0.015*** –0.023*** –0.013*** –0.041*** –0.021*** –0.032*** (0.005) (0.005) (0.006) (0.006) (0.003) (0.004) (0.004) (0.005) (0.004) (0.004) OPCYCLE –0.013*** –0.010* –0.023*** –0.005 0.001 –0.010*** –0.012*** –0.003 –0.015*** –0.006* (0.004) (0.005) (0.007) (0.005) (0.003) (0.003) (0.004) (0.004) (0.004) (0.004) |DACC| 0.171*** 0.164*** 0.120*** 0.195*** 0.115*** 0.190*** 0.203*** 0.249*** 0.148*** 0.196*** (0.027) (0.032) (0.037) (0.033) (0.015) (0.017) (0.021) (0.027) (0.020) (0.019) (Continued) Ranasinghe and Habib 19 Table 6. (Continued) 1 2 3 4 5 6 7 8 9 10 BIND INSTOWN COMPETITION ANALYST FCOMP Low High Low High Low High Low High Low High First step control SALESG 0.408*** 0.108 –0.185** 0.686*** 0.048 0.096* 0.109 0.272 –0.029 0.191** (0.068) (0.091) (0.087) (0.049) (0.038) (0.050) (0.087) (0.181) (0.058) (0.096) Industry effects Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Year effects Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes SALESG*Industry Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes effects SALESG*Year effects Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Constant 0.127*** 0.085 0.164** 0.175*** 0.072** 0.064* 0.221*** 0.218*** 0.066* 0.332*** (0.032) (0.060) (0.071) (0.046) (0.032) (0.033) (0.045) (0.057) (0.039) (0.045) Observations 10,460 8,226 7,078 8,090 24,594 28,610 15,477 12,289 17,195 16,118 Adjusted R 0.272 0.260 0.316 0.333 0.301 0.262 0.323 0.268 0.321 0.259 Note: This table presents the cross-sectional test results. Columns 1 and 2 shows the association between STRAT_DEV and INV for Low and High board independence (BIND) sub-samples. Columns 3 and 4 shows the association between STRAT_DEV and INV for Low and High institutional ownership (INSTOWN) sub-samples. Columns 5 and 6 show regression estimates for the association between STRAT_DEV and INV for the Low and High competition (COMPETITION) sub-sample, when the competition is measured using product market fluidity. Columns 7 and 8 show the association between STRAT_DEV and INV for the Low and High sub-samples of number of analysists following (ANALYST). Columns 9 and 10 show the regression estimates of the association between STRAT_DEV and INV for the Low and High sub-samples of financial statement comparability (FCOMP). Robust standard errors clustered at firm-level are in parentheses *, **, *** denote a two-tailed p-value of less than 0.10, 0.05 and 0.01, respectively. Variables are defined in Appendix 1. 20 Australian Journal of Management 00(0) have a FLUID value lower (higher) than the industry-year median FLUID. The coefficient on STRAT _ DEV is positive and significant for the high-product-market-competition sub-sample, (β = 0.004, p < 0.01) (Column 6), while this is insignificant for the low-product-market-competi- tion sub-sample (Column 5). This finding suggests that the escalated risks and agency problems stemming from high competition are associated with sub-optimal investments by deviant firms and supports signal precision view of H3. Finally, to ascertain the moderating effects of the information environment on the association between STRAT _ DEV and INV , first we divide the sample into low and high analyst following. The low- (high-) analyst-following sub-sample includes observations with analyst following ( LNANALYST ) lower (higher) than the industry median for each year. Results show that the coef- ficient on STRAT _ DEV is positive and significant for the low-analyst-following sub-sample (β = 0.003, p < 0.10) (Column 7). However, the coefficient is insignificant for the high-analyst- following sub-sample (Column 8). This supports the view that strategic deviation increases infor- mation opaqueness and, hence, managerial opportunism and consequently sub-optimal investments. The result, therefore, supports H4a. We also investigate whether the association between STRAT _ DEV and INV is stronger when the financial statement comparability () FCOMP is lower. We follow De Franco et al. (2011) and measure financial statement comparability using time-series regression. We use INDC _ OMP , the median financial statement comparability for all firms j in the same industry as firm during year t, as our measure of financial statement com- parability. The low- (high-) FCOMP sub-sample includes observations that have an INDC _ OMP value lower (higher) than the industry-year median. As shown in Column 9, the coefficient on STRAT _ DEV is positive and significant for the low-FCOMP sub-sample only (β = 0.004, p < 0.01), thereby supporting H4b. 4.5. Robustness tests In our first robustness test, we develop an alternative investment inefficiency measure (_ ALTINV _) INEFF following Eisdorfer et al. (2013). First, we determine the actual invest- ments for each firm-year by scaling the gross CAPEX by the book value of total assets at the begin- ning of the year. We then determine expected investment where the expected investment is calculated as the median investment in the four-digit SIC industries. Then we estimate the abnor- mal investments by subtracting actual investments from expected investments. We use the absolute value of abnormal investment and denote this as ALTI __ NV INEFF. Table 7 shows the coeffi- cient on ALTI __ NV INEFF positive and significant (β = 0.313, p < 0.01) (Column 1). Second, we use an alternative measure of strategic deviation ( ALTS __ TRAT DEV ) and re- estimate our baseline model and report the result in Column 2 in Table 7. We follow prior research (Dong et al., 2021; Finkelstein and Hambrick, 1990) and construct ALTS __ TRAT DEV , exclud- ing R&D intensity and advertising intensity from the STRAT _ DEV measure. We regress INV on ALTS __ TRAT DEV , a set of control variables, the first stage control variable (SALESG), industry and year indicators, SLAESG*industry effects and SALESG*year effects. We find sup- port for the assertion that STRAT _ DEV and INV are positively associated using this alternative proxy of strategic deviation (β = 0.002, p < 0.05). 4.6. Strategic deviation, investment and firm value Finally, we examine whether the investment inefficiencies that arise due to strategic deviation have negative consequences for firm value. We use the following regression to test this proposition: Ranasinghe and Habib 21 Table 7. Strategic deviation and investment efficiency: robustness tests. (1) (2) ALT_INV_INEFF INV STRAT_DEV 0.313*** (0.044) ALT_STRAT_DEV 0.002** (0.001) MTB 0.000 0.000*** (0.000) (0.000) SIZE 0.205*** 0.015*** (0.039) (0.001) CFL_VOL 0.074*** 0.000 (0.027) (0.000) SALES_VOL 2.493*** 0.030*** (0.306) (0.005) INV_VOL –0.0155 0.003*** (0.016) (0.001) TANG 13.235*** 0.077*** (0.754) (0.008) CFOSALES –1.063 –0.029*** (0.943) (0.009) SLACK 0.004* –0.002* (0.002) (0.000) LOSS –0.813*** –0.016*** (0.181) (0.002) AGE –0.673*** –0.014*** (0.176) (0.002) ZSCORE 0.258*** –0.021*** (0.082) (0.001) DIV –1.351*** –0.023*** (0.108) (0.002) OPCYCLE 0.715*** –0.006*** (0.240) (0.002) |DACC| 5.887*** 0.160*** (0.710) (0.011) First step control SALESG – 0.057** (0.029) Industry effects Yes Yes Year effects Yes Yes SALESG*Industry effects – Yes SALESG*Year effects – Yes Constant –2.737 0.099*** (2.242) (0.025) Observations 64,510 59,423 Adjusted R 0.069 0.278 Note: This table shows the pooled regression estimates of the association between strategic deviation and investment using alternative measures. Column 1 shows the association between STRAT_DEV and ALT_INV_INEFF. Column 2 shows the association between ALT_STRAT_DEV and INV. Robust standard errors clustered at firm-level are in parentheses. *, **, *** denote a two-tailed p-value of less than 0.10, 0.05 and 0.01, respectively. Variables are defined in Appendix 1. 22 Australian Journal of Management 00(0) VALUE =+ ββ STRAT __ DEVI ++ ββ NV STRAT DEVI * NV it ,, +10 12 it it ,, 3 it (3) + λλ CONTROLS ++ INDUSTRYY λε EAR + ii,, t jj t kk,, ti t ∑ ∑∑ ∑ i j k Where VALUE is firm value proxied by Peters and Taylor (2017) investment-Q (QTOT) and TOBINQ measures and other variables are defined as before. We expect the coefficient β to be negative and significant if the capital market discounts the sub-optimal investments undertaken by the strategically deviant firms. We present the regression results in Table 8. Columns 1 and 2 report the result for QTOT and TOBINQ measures respectively. We find the coefficient on the interactive variable STRAT_DEV*INV negative and significant for both QTOT (coefficient -0.043, p < 0.01) and TOBINQ (β = -0.345, p < 0.01) suggesting that the capital market discounts the investments undertaken by the deviant firms. 5. Conclusion We explore the association between strategic deviation and investment inefficiency. It is important to understand the repercussions of deviating from the industry-level strategy. At the same time, it is of vital importance to know the direct antecedents of investment inefficiencies. We argue that firms deviating from industry norms are prone to increased information asymmetry and hence, are able to engage in self-serving behaviour as manifested in inefficient investments. Using a sample of U.S. data, we find support for our prediction. Then we examine the moderating role of weak monitoring, high competition and a low-quality information environment. Our findings suggest that the positive association between strategic deviation and investment inefficiency is pronounced for firms exposed to weak monitoring and with a low-quality information environment and for firms operating in a highly competitive environment. Our results remain robust to possible endo- geneity concerns. By examining the association between strategic deviation and investment inef- ficiency, we contribute to both the strategic management and the investment literature. Although there are a number of studies that examine the determinants of investment inefficiencies, there is no evidence on the association between deviant strategies and investment inefficiencies. This study fills this gap in the literature. Business strategy is a choice that managers have to make, and, hence, it is important to under- stand the managerial incentives behind choosing a particular business strategy. As elaborated above, deviating from industry peers results in negative consequences, such as high information asymmetry, high risk and uncertainty. On the other hand, conforming to industry peers bring about homogeneity and is preferred by capital market participants due to low information processing costs and low uncertainty (DiMaggio and Powell, 1983). Litov et al. (2012) argue that pursuing unique strategies brings about economic rents which are associated with firm value in the long term. In contrast, Navissi et al. (2017) find that sub-optimal investments resulting from different strategic choices adversely affect future performance. Similarly, Dong et al. (2021) find that the capital market discounts the value of cash holdings of firms pursuing non-conforming strategies. Given these mixed implications, managers face a paradox in selecting a strategy (Litov et al., 2012). Despite the negative implications, managers might pursue deviant strategies when the costs of being different is lower than the benefits of conforming to industry standards. In other words, such firms may confront information asymmetry and thereby negative capital market consequences in the short term, but may enjoy first-mover benefits in the long term. Alternatively, the impact of strategic deviance on firm-level outcomes may be stronger in firms with peculiar characteristics. Ranasinghe and Habib 23 Table 8. Strategic deviation, investment and firm value. (1) (2) QTOT TOBINQ t + 1 t + 1 STRAT_DEV 0.396*** 3.392*** (0.050) (0.258) INV 0.055*** 0.188*** (0.006) (0.028) STRAT_DEV*INV –0.043*** –0.345*** (0.014) (0.072) MTB 0.000* 0.001*** (0.000) (0.000) SIZE 0.232*** 0.872*** (0.006) (0.029) CFL_VOL 0.020*** 0.096*** (0.002) (0.011) SALES_VOL 0.235*** 1.093*** (0.043) (0.209) INV_VOL –0.002 –0.013 (0.006) (0.025) TANG –0.431*** –8.865*** (0.066) (0.298) CFOSALES 0.197*** 0.877** (0.076) (0.350) SLACK 0.004*** 0.064*** (0.001) (0.006) LOSS –0.124*** –0.339*** (0.015) (0.076) AGE 0.049** –0.177* (0.023) (0.102) ZSCORE –0.022** –0.458*** (0.009) (0.053) DIV –0.256*** –1.207*** (0.027) (0.117) OPCYCLE –0.009 –0.367*** (0.015) (0.078) |DACC| 0.682*** 2.263*** (0.061) (0.291) Industry effects Yes Yes Year effects Yes Yes Constant 0.187 4.540*** (0.312) (0.824) Observations 50,470 50,263 Adjusted R 0.244 0.387 Note: This table shows the association between investment and firm value when strategic deviance exists. Column 1 shows results when investment-Q model is used, following Peters and Taylor (2017). Column 2 shows the results when firm value is measured using TOBINQ. Robust standard errors clustered at firm-level are in parentheses. *, **, *** denote a two-tailed p-value of less than 0.10, 0.05 and 0.01, respectively. Variables are defined in Appendix 1. 24 Australian Journal of Management 00(0) For example, Tang et al. (2011) find that firms with dominant CEOs pursue deviant strategies and encounter extreme performance. In this study, we focus on whether strategically deviant firms are associated with sub-optimal investments as a first step to understanding the implications of strate- gic choice. Acknowledgement We acknowledge the Associate Editor, Professor Tom Smith, Deputy Editor (Finance), Associate Professor Chelsea Liu and the anonymous reviewer for their valuable comments. Funding The author(s) received no financial support for the research, authorship, and/or publication of this article. ORCID iDs Dinithi Ranasinghe https://orcid.org/0000-0002-2102-9481 Ahsan Habib https://orcid.org/0000-0003-2433-3961 Notes 1. As we use scaled investment as our dependent variable, control for the independent variable from the first stage regression and include several interaction terms as suggested by Jackson (2022) for perform- ing the one-stage regression model, we continue to use the term investment in(efficiency) for the remain- der of the paper but use INV instead of INV_INEFF. 2. However, Ye et al. (2021) find that firms pursuing a deviant strategy have more firm-specific information impounded into their stock prices and hence such firms have less synchronous stock price movement. This occurs because strategically deviant firms issue more managerial earnings forecasts and have a higher level of block ownership than the nondeviant firms. 3. We replace missing R&D and advertising expenses with zero consistent with prior research. 4. We do not use first step controls and the industry and year interactions in the 2SLS model because we use IVREG2 in STATA to estimate the 2SLS model, which supposedly makes standard error corrections and accounts for estimation issues (see Chen et al., 2022). 5. Firm location data (latitudes and longitudes) come from Professor Bill McDonald’s website (https:// www3.nd.edu/~mcdonald/). 2 2 (( COSa22 SINb −+ bC 1)) ( OSaS 12 INaS − INa1 12 COSa COSb () 21 − b ) 6. , we Distancea = 3963.* 19 rctan SINa SINa COSa COSa COSb () − b 12 12 21 multiply the value by negative one to convert into proximity. Distance 7. See http://hobergphillips.tuck.dartmouth.edu/. 8. See Appendix 2 for a detailed explanation of financial statement comparability calculation. 9. See the Appendix 1 for the detailed variable calculation. References Abdoh H and Varela O (2017) Product market competition, idiosyncratic and systematic volatility. Journal of Corporate Finance 43: 500–513. Al-Hadi A, Habib A, Taylor G, et al. (2021) Financial statement comparability and corporate investment efficiency. Meditari Accountancy Research 29: 1283–1313. Arellano M and Bover O (1995) Another look at the instrumental variable estimation of error-components models. Journal of Econometrics 68: 29–51. Babar M and Habib A (2021) Product market competition in accounting, finance, and corporate governance: A review of the literature. International Review of Financial Analysis 73: 101607. Ranasinghe and Habib 25 Bebchuk LA and Stole LA (1993) Do short-term objectives lead to under-or overinvestment in long-term projects? The Journal of Finance 48: 719–729. Bentley KA, Omer TC and Sharp NY (2013) Business strategy, financial reporting irregularities, and audit effort. Contemporary Accounting Research 30: 780–817. Bentley-Goode KA, Omer TC and Twedt BJ (2019) Does business strategy impact a firm’s information envi- ronment? Journal of Accounting, Auditing & Finance 34: 563–587. Biddle GC, Hilary G and Verdi RS (2009) How does financial reporting quality relate to investment effi- ciency? Journal of Accounting and Economics 48: 112–131. Blundell R and Bond S (1998) Initial conditions and moment restrictions in dynamic panel data models. Journal of Econometrics 87: 115–143. Brennan MJ and Subrahmanyam A (1995) Investment analysis and price formation in securities markets. Journal of Financial Economics 38: 361–381. Carpenter MA (2000) The price of change: The role of CEO compensation in strategic variation and deviation from industry strategy norms. Journal of Management 26: 1179–1198. Chen M-J and Hambrick DC (1995) Speed, stealth, and selective attack: How small firms differ from large firms in competitive behavior. Academy of Management Journal 38: 453–482. Chen T, Xie L and Zhang Y (2017) How does analysts’ forecast quality relate to corporate investment effi- ciency? Journal of Corporate Finance 43: 217–240. Chen W, Hribar P and Melessa S (2018) Incorrect inferences when using residuals as dependent variables. Journal of Accounting Research 56: 751–796. Chen W, Hribar P and Melessa S (2022) On the use of residuals as dependent variables. Journal of Financial Reporting 7: 69–83. Choi JH, Choi S, Myers LA, et al. (2019) Financial statement comparability and the informativeness of stock prices about future earnings. Contemporary Accounting Research 36: 389–417. Christodoulou D, Ma L and Vasnev A (2018) Inference-in-residuals as an estimation method for earnings management. Abacus 54: 154–180. Chung KH, McInish TH, Wood RA, et al. (1995) Production of information, information asymmetry, and the bid-ask spread: Empirical evidence from analysts’ forecasts. Journal of Banking & Finance 19: 1025–1046. Deephouse DL (1996) Does isomorphism legitimate? Academy of Management Journal 39: 1024–1039. Deephouse DL (1999) To be different, or to be the same? It’s a question (and theory) of strategic balance. Strategic Management Journal 20: 147–166. De Franco G, Kothari SP and Verdi RS (2011) The benefits of financial statement comparability. Journal of Accounting Research 49: 895–931. DiMaggio PJ and Powell WW (1983) The iron cage revisited: Institutional isomorphism and collective ration- ality in organizational fields. American Sociological Review48: 147–160. Dong X, Chan KC, Cui Y, et al. (2021) Strategic deviance and cash holdings. Journal of Business Finance & Accounting 48: 742–782. Eisdorfer A, Giaccotto C and White R (2013) Capital structure, executive compensation, and investment efficiency. Journal of Banking & Finance 37: 549–562. Fama EF and Jensen MC (1983) Separation of ownership and control. The Journal of Law and Economics 26: 301–325. Ferreira MA and Matos P (2008) The colors of investors’ money: The role of institutional investors around the world. Journal of Financial Economics 88: 499–533. Financial Accounting Standards Board (FASB) (2010) Statement of financial accounting concepts No. 8. Conceptual framework for financial reporting. Norwalk, CT: FASB. Finkelstein S and Hambrick DC (1990) Top-management-team tenure and organizational outcomes: The moderating role of managerial discretion. Administrative Science Quarterly 35: 484–503. Francis JR, Pinnuck ML and Watanabe O (2014) Auditor style and financial statement comparability. The Accounting Review 89: 605–633. Geletkanycz MA and Hambrick DC (1997) The external ties of top executives: Implications for strategic choice and performance. Administrative Science Quarterly 42: 654–681. 26 Australian Journal of Management 00(0) Hainmueller J (2012) Entropy balancing for causal effects: A multivariate reweighting method to produce balanced samples in observational studies. Political Analysis 20: 25–46. Harrington JR and Gelfand MJ (2014.) Tightness–looseness across the 50 united states. Proceedings of the National Academy of Sciences 111: 7990–7995. Hoberg G, Phillips G and Prabhala N (2014) Product market threats, payouts, and financial flexibility. The Journal of Finance 69: 293–324. Holmström B (1999) Managerial incentive problems: A dynamic perspective. The Review of Economic Studies 66: 169–182. Imhof MJ, Seavey SE and Watanabe OV (2022) Competition, proprietary costs of financial reporting, and financial statement comparability. Journal of Accounting, Auditing & Finance 37: 114–142. Irvine PJ and Pontiff J (2008) Idiosyncratic return volatility, cash flows, and product market competition. The Review of Financial Studies 22: 1149–1177. Jackson A (2022) Residuals from two-step research designs. Accounting & Finance 62: 4345–4358. Jensen MC (1986) Agency costs of free cash flow, corporate finance, and takeovers. The American Economic Review 76: 323–329. Jensen MC and Meckling WH (1976) The theory of the firm: managerial behavior, agency costs, and owner- ship structure. Journal of Financial Economics 3(4): 305–360. Jiang F, Kim KA, Nofsinger JR, et al. (2015) Product market competition and corporate investment: Evidence from China. Journal of Corporate Finance 35: 196–210. Kang YJ, Lee H-Y, Park H-Y, et al. (2021) Social ties, managerial overconfidence, and investment efficiency. Finance Research Letters 46: 102300. Kim J-B, Li L, Lu LY, et al. (2016) Financial statement comparability and expected crash risk. Journal of Accounting and Economics 61: 294–312. Laksmana I and Yang Y-w (2015) Product market competition and corporate investment decisions. Review of Accounting and Finance 14: 128–148. Lin Y-E, Li Y-W, Cheng TY, et al. (2021) Corporate social responsibility and investment efficiency: Does business strategy matter? International Review of Financial Analysis 73: 101585. Litov LP, Moreton P and Zenger TR (2012) Corporate strategy, analyst coverage, and the uniqueness para- dox. Management Science 58: 1797–1815. McMullin JL and Schonberger B (2020) Entropy-balanced accruals. Review of Accounting Studies 25: 84– Malmendier U and Tate G (2005) CEO overconfidence and corporate investment. The Journal of Finance 60: 2661–2700. Miles RE and Snow CC (1978) Organizational Strategy, Structure, and Process. New York: McGraw Hill. Mintzberg H (1978) Patterns in strategy formation. Management Science 24: 934–948. Modigliani F and Miller MH (1958) The cost of capital, corporation finance and the theory of investment. The American Economic Review 48: 261–297. Myers SC (1977) Determinants of corporate borrowing. Journal of Financial Economics 5: 147–175. Navissi F, Sridharan V, Khedmati M, et al. (2017) Business strategy, over-(under-) investment, and manage- rial compensation. Journal of Management Accounting Research 29: 63–86. Peters RH and Taylor LA (2017) Intangible capital and the investment-q relation. Journal of Financial Economics 123: 251–272. Pool VK, Stoffman N and Yonker SE (2015) The people in your neighborhood: Social interactions and mutual fund portfolios. The Journal of Finance 70: 2679–2732. Porter ME (1980) Competitive Strategy: Techniques for Analyzing Industries and Competitors. New York: The Free Press. Porter ME (1986) Changing patterns of international competition. California Management Review 28: 9–40. Porter ME (1996) What is strategy? Harvard Business Review, November–December, 37–55. Provaty S, Amin S and Hasan MM (2022) Strategic deviation and debt maturity structure. Finance Research Letters 50: 103317. Ranasinghe and Habib 27 Rajkovic T (2020) Lead independent directors and investment efficiency. Journal of Corporate Finance 64: Roulstone DT (2003) The relation between insider-trading restrictions and executive compensation. Journal of Accounting Research 41: 525–551. Shleifer A and Vishny RW (1989) Management entrenchment: The case of manager-specific investments. Journal of Financial Economics 25: 123–139. Shleifer A and Vishny RW (1997) A Survey of Corporate Governance. The Journal of Finance 52(2): 737–783. Stock J and Yogo M (2005) Asymptotic distributions of instrumental variables statistics with many instruments. In: Andrews DWK, Stock JH and Rothenberg TJ (eds) Identification and Inference for Econometric Models: Essays in Honor of Thomas Rothenberg, pp. 109–120. Cambridge: Cambridge University Press. Stoughton NM, Wong KP and Yi L (2017) Investment efficiency and product market competition. Journal of Financial and Quantitative Analysis 52: 2611–2642. Tang J, Crossan M and Rowe WG (2011) Dominant CEO, deviant strategy, and extreme performance: The moderating role of a powerful board. Journal of Management Studies 48: 1479–1503. Ye K, Juan JX and Zhang B (2021) Strategic deviation and stock return synchronicity. Journal of Accounting, Auditing & Finance 36: 172–194. Appendix 1 Variable definitions. Variable Definition Dependent variable INV INV is total investments, calculated as the sum of research and development (R&D) expenditure, capital expenditure (CAPEX) and acquisition expenditure (AQC), less cash receipts from sale of property, plant and equipment (SPPE), and scaled by the lagged total assets (AT). Explanatory variable STRAT_DEV Strategic deviation, measured as the extent to which a firm’s strategy deviates from industry average. See Section 3.2.2. for detailed estimation procedure. Control variables MTB Market-to-book ratio, measured as market value of equity (PRCC_F*CSHO) scaled by book value of equity (CEQ). SIZE The natural logarithm of the market value of equity. CFO_VOL Standard deviation of the cash flow from operations (OANCF) deflated by average total assets for the period t-5 to t-1. SALES_VOL Standard deviation of the sales deflated by average total assets for the period t-5 to t-1. INV_VOL Standard deviation of the investment defined in Section 3.2 for the period t-5 to t-1. TANG The ratio of net property, plant and equipment to total assets. CFOSALES The ratio of the cash flow from operations to sales. SLACK The ratio of cash and short-term investments (CHE) to net property, plant and equipment (PPENT). LOSS An indicator variable coded 1 for firms incurring losses, and zero otherwise. AGE The natural logarithm of firm age estimated as the number of years since the firm’s initial appearance in Compustat annual file. (Continued) 28 Australian Journal of Management 00(0) Appendix 1. (Continued) Variable Definition ZSCORE (3.3*Pretax Income/Total Assets + Sales/Total Asset + 0.25*Retained Earnings/ Total Assets + 0.5*Working Capital/Total Assets) (Rajkovic, 2020). DIV An indicator variable that equals one for a dividend paying firm, and zero otherwise. OPCYCLE The natural log of receivables (RECT) to sales plus inventory (INVENT) to cost of goods sold (COGS) multiplied by 360. |DACC| Absolute value of performance adjusted discretionary accruals, estimated using the below equation for all firms in the same industry with at least 15 observations for each industry-year       ACC 1 ΔΔ SALESR − ECEIV PPE it , it ,, it it , = γγ + + γ 0 1 2       TA TA TA TA (4) it ,, −− 1 it 1 it , − −− 1 it , 1       + γε ROA +… 31 () it ,, − it where ACC is total accruals calculated as earnings before extraordinary items and discontinued operations minus operating cash flows; TA is total assets in year t-1; ΔSALES is change in sales from year t-1 to year t; ∆RECEIV is change in accounts receivable from year t-1 to year t; PPE is gross property plant & equipment; ROA is return on assets measured as earnings before extraordinary items and discontinued operations for the preceding year divided by total assets for the same year. The coefficient estimates from Equation (4) are used to estimate the non-discretionary component of total accruals (NDAC) for our sample firms. The discretionary accruals are then the residuals from equation (4). First step controls SALESG Sales change from the last year. Instrumental variables LOOSENSS US state-level tightness–looseness index following Harrington and Gelfand (2014). The authors used nine items to create a composite index. Four items reflect strength of punishment: (i) the legality of corporal punishment in schools, (ii) the percentage of students hit/punished in schools, (iii) the rate of executions from 1976 to 2011, and (iv) the severity of punishment for violating laws (i.e., selling, using, or possessing marijuana). Two items reflect latitude/permissiveness: (i) access to alcohol (i.e., ratio of dry to total counties per state) and (ii) the legality of same-sex civil unions. Institutions that reinforce moral order and constrain behaviour were assessed with two items: (i) state-level religiosity and (ii) percentage of individuals claiming no religious affiliation. The final indicator was the percentage of total population that is foreign (p. 7991). PROXIMITY Geographical proximity to the local largest strategic deviant firm (see footnote 4 in-text). Moderating variables BIND Board independence is the proportion of independent directors on board, i.e. the number of independent directors on the board divided by the total number of directors on the board. INSTOWN Institutional ownership is the proportion of top five institutional ownership relative to total ownership. COMPETITION Product market threat proxied by the Fluidity score obtained from Hoberg- Phillips Data Library (Hoberg et al. (2014). ANALYST The natural log of number of analysts following. (Continued) Ranasinghe and Habib 29 Appendix 1. (Continued) Variable Definition FCOMP Firm-year level accounting comparability, which is the industry median of comparability combinations for firm i and other firms in the same two-digit SIC in a given year (see Appendix 2 for detailed estimation). Firm value Measured by scaling firm value by the sum of physical and intangible capital using measures Peters and Taylor (2017) methodology as follows: QTOT (5) Qtot =+ VK /( phy Kint ) it ,, it it ,, it where Qtot is measured by scaling firm value by the sum of the physical and intangible capital. V is firm’s market value defined as market value of equity (PRCC_F*CSHO) plus book value of debt (DLTT + DLC), minus current assets (ACT). Kphy is the replacement value of physical capital (PPEGT). Kint is the replacement cost of intangible capital, which is the sum of a firm’s externally purchased (INTAN) and internally created intangible capital. If INTAN is missing, we set the value to zero. Internally created intangible capital is the sum of knowledge capital (G) and organizational capital (O). (6) GG =− () 1 ∂+ RD & it ,& RD it ,, −1 it (7) OO =− () 1 ∂+ SG & A it ,& SG Ai ,, ti −1 t G is the end-of-period stock of knowledge capital, ∂ is depreciation rate, i, t R&D and R&D is the real research and development expenditure (XRD) for the i,t year. We replace missing XRD with zero following Peters and Taylor (2017). Following Peters and Taylor (2017) we use a 15% depreciation rate for ∂ . R&D Based on the suggestion Peters and Taylor (2017), we set G = 0. O is the i,0 i, t organizational capital, ∂ is depreciation rate, and SG&A is the selling, general SG&A i,t and administrative expenses (XSGA minus XRD minus RDIP) for the year. If XRD is higher than XSGA but is less than COGS, or XGA missing, we measure SG&A i,t as XSGA with no additional adjustments. We set XSGA, XRD and RDIP to zero if missing, following Peters and Taylor (2017). Following Peters and Taylor (2017) we use a 20% depreciation rate for ∂ . We further set O = 0 following Peters SG&A i,0 and Taylor (2017). TOBINQ Market value of equity plus book value of total assets minus book value of equity, divided by book value of total assets at the beginning of fiscal year. Appendix 2 Computation of financial statement comparability We follow De Franco et al. (2011) in measuring financial statement comparability. As presented in equation (8) using firm i ’s 16 previous quarters of earnings and stock returns: EARNINGS =+ αβ RETURN +∈ (8) it ,, ii it it , where EARNINGS is the quarterly net income before extraordinary items scaled by the beginning- of-period market value of equity, and RETURN is the raw stock return during quarter t. and β denote the mapping of firm i ’s economic events (i.e. returns) onto its accounting numbers (i.e. earnings). Firm from the same two-digit industry as firm i has and β , which reflects its ∝ 30 Australian Journal of Management 00(0) economic events to accounting numbers. Next, predicted earnings of firm and are calculated using their respective accounting functions with firm i’s economic events, � � E() EARNINGS =∝ + β RETURN (9) iit ,, i it , � � (10) E() EARNINGS =∝ + β RETURN ij ,,t j it , where E() EARNINGS is firm i’s predicted earnings based on firm i’s accounting function and iit ,, firm i’s return in period t, and E() EARNINGS is firm j’s predicted earnings based on firm j’s ij ,,t accounting function and firm i’s return in period t. The negative value of the average absolute dif- ference between the predicted earnings using firm i’s and firm j’s accounting functions is the pair- wise comparability between firms i and j, FCOMP . ij ,,t (11) FCOMPE =− () EARNINGS − E() EARNINGS ij ,,t ∑ iit ,, ij ,,t t−15

Journal

Australian Journal of Management – SAGE

Published: Jan 1, 2023

Keywords: Information asymmetry; investments; product market competition; strategic deviation

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Strategic deviation and investment inefficiency

Strategic deviation and investment inefficiency

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Strategic deviation and investment inefficiency

Strategic deviation and investment inefficiency

References (55)

Abstract

Journal

Recommended Articles

There are no references for this article.

Our policy towards the use of cookies