< PreviousA PROFESSIONAL DEVELOPMENT JOURNAL for the CONSULTING DISCIPLINES 10 JULY | AUGUST 2019 t h e v a l u e e x a m i n e r weak cybersecurity are recent developments. We believe that the expert should decide the best approach but start with the cost to mitigate as a decrease to the value of the subject company. CYBERSECURITY RISK IN THE REPORTING STANDARDS There are three areas where this risk must be identified if it exists and a fourth if an outside expert is used. The first discussion is in the limiting conditions or scope limitations where the risk is identified and its effect on the results quantified. The second discussion is in the description of the company and how it does business where the risk is identified. The third discussion is in the valuation method used where the value adjustment is made. Lastly, if a third-party specialist is used, the report discussion should disclose what reliance was placed upon that work and who is responsible for the work done by the specialist. CONCLUSION While it is not necessary to be a cybersecurity expert, cybersecurity may now have a material impact on valuation. The magnitude of this impact can be determined by addressing cybersecurity in the management interview, the development of a value, and the reporting of a value. Raymond Hutchins is the managing partner for CyberCecurity LLC, a full-service cybersecurity and privacy firm headquartered in Denver, Colorado that has an international client base. He and his partner, Mitch Tanenbaum, have over fifty years of IT, cybersecurity, and privacy experience. They are highly respected national cybersecurity leaders who speak and write frequently on the subject of cybersecurity, privacy, and compliance. They are the first cybersecurity professionals to introduce cybersecurity due diligence to the accounting and valuation professions. E-mail: Dave Miles, CPA, CVA, CGMA, is the business valuation manager at ValuSource. For the last nineteen years, he has worked on developing software, developing web applications, publishing datasets, and providing valuation expertise to both customers and the ValuSource team. Typically, he identifies the “what” and “why” of a valuation technique or database and then passes on that knowledge. VEUltimate Training and Membership Subscription Unlimited Continuing Professional Education Membership Dues Recertification Fees One Monthly Fee: (multi-user options available) To simplify your professional development, we are excited to introduce the Ultimate Training and Membership Subscription where, for a flat monthly fee, NACVA members have available to them a new membership level to receive unlimited CPE with zero added or hidden costs. Designed to allow you to precisely budget your annual CPE expense and obtain 100% of your CPE from NACVA/CTI. In addition, the subscription fee covers your annual membership dues and tri-annual recertification fees. Options available for both single- and multi-users. All subscription options require a one-year commitment. For $335 per month (for first user) you receive these benefits: Paid Practitioner membership dues Paid recertification fees Unlimited complimentary registration to all our live training events Complimentary registration to all our conferences, including the Annual Consultants’ Conference and Industry Specialty SuperConferences Unlimited complimentary registration to all live webinars Unlimited access to our CPE On-Demand library of nearly 700 webinars Complimentary self-study materials in an electronic format (nominal expense for hard copies) Access to all other paid CPE resources, including The Value Examiner® and QuickRead® quizzes, etc. Access to training in a variety of fields of study (registered with the National Association of State Boards of Accountancy) to meet your NACVA and other credential requirements, including your CPA license Surgent CPE: NASBA qualified self-study courses to meet the needs of every CPA, tax practitioner, and financial professional. Select courses are available 24/7 in a variety of fields of study covering ethics, accounting, auditing, and taxes. To sign up for the Ultimate Training and Membership Subscription, and address any questions, or contact Member/Client Services at (800) 677-2009. No other association in the business valuation or financial litigation profession provides the wealth and depth of resources like NACVA. Choose the Ultimate Triple Play Subscription to receive everything included in all three other subscriptions, plus Damages Advocate calculation software. This subscription gives you access to everything we offer. Available for $625 per month for the first user. Visit of what is included in each subscription level. Ultimate KeyValueData® Titanium Subscription— $250 per month for first user (access to 21 separate databases, reports, libraries, and presentations) Ultimate Software Subscription— $90 per month for first user (licenses to five valuation and report writing software packages, plus technical support) Add to your Ultimate Training and Membership Subscription: 3/4/19 11:33AMA PROFESSIONAL DEVELOPMENT JOURNAL for the CONSULTING DISCIPLINES 12 JULY | AUGUST 2019 t h e v a l u e e x a m i n e r VALUATION /////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////// /////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////// The Role of Managerial Ability in Firm Valuation By Davit Adut, PhD; Marinilka Barros Kimbro PhD; Marc Picconi, PhD; and Philipp Schaberl, PhD In this article, we provide an explanation based on Ohlson (1995) and empirical evidence that managerial ability (MA) significantly influences the relationship between market value (MV) and accounting fundamentals, such as book value of equity (BV), net income (NI), cash flows (CF), and accruals (ACCR). We define MA as a mathematical ratio of inputs to outputs and provide a novel approach to measuring MA. Valuation analysts usually use discounted CF or residual income methods. These models usually adjust for MA in their discount rates. Although valuation analysts already evaluate the quality of management in their evaluation of risk by direct inputs, this approach is not standardized and is ad hoc as it relates to evaluation of accounting fundamentals. This approach, although widely used, might contain biases specific to the analyst. Generally Accepted Accounting Principles place great emphasis on and urges systematic approaches to asset valuations as well as other accounting information. Consistent with this approach, our measure presents a more systematic valuation method that can be used by analysts. We argue that this should be of value to the business valuation community. In a related article, Hambrick et al., (1995) provide evidence that managerial discretion differs widely among industries. Our measure provides an industry adjusted approach to MA and provides insights to a problem recognized by security analysts. We argue that these two important points present an incremental contribution to business valuation techniques and can be utilized by valuation analysts. The management literature has explored the value of MA and concludes that human capital (talent) adds additional value to the firm (e.g., Hambrick and Mason, 1984). Indeed, numerous studies provide evidence in support of the argument that strong MA has a positive effect on firm performance, and therefore firm value. For example, Demerjian et al., (2013) document a positive association between MA and earnings and ACCR persistence. In light of these findings, the natural question emerges as to whether investors price accounting fundamentals differently conditional on a firm’s level of MA. In other words, is a dollar of income more valuable when it is generated by a high- ability manager? The purpose of this article is to examine this question. Moreover, at the end of the article, we use two real-world examples to demonstrate how estimating the relationship between accounting fundamentals and firm value by applying an appropriate level of MA can reduce the error when estimating firm value. Based on a sample of 150,348 publicly traded firm-year observations for the twenty-nine-year period 1987–2015, we use the robust Theil-Sen (TS) estimation methodology to estimate the relationships between MV and several accounting fundamentals. We find that the association between MV and NI is stronger for high-ability managers. To obtain a better understanding whether this result is driven by the CF or ACCR component of NI, we repeat this analysis after decomposing NI into CF and ACCR. Again, we find that CF and ACCR are more strongly associated with MV for firms with high-ability managers. Based on Ohlson (1995) and Demerjian et al., (2012) we propose a positive relationship between MA and earnings persistence as an explanation for our findings. MEASURING MANAGERIAL ABILITY Demerjian, et al., (2012) develop a two-step process which measures MA based on how effective managers are at converting a given set of resources into revenue. In the first step of the process, Demerjian et al., use data envelopment analysis (DEA) to estimate firm efficiency. DEA is a nonparametric method that uses multiple inputs and outputs to measure the relative efficiency of decision-making units (DMUs) or firms within their industries. DEA creates an efficient frontier of observed production points from linear programming to maximize a ratio of outputs to inputs. DEA assigns a value of one to the most efficient DMUs on the frontier and a value less than one to less efficient DMUs. To estimate firm efficiency A PROFESSIONAL DEVELOPMENT JOURNAL for the CONSULTING DISCIPLINES t h e v a l u e e x a m i n e r JULY | AUGUST 2019 13 scores, Demerjian et al., use sales revenue as the sole output variable and the following seven input variables that are all derived from firms’ publicly available financial reports which are widely used by valuation analysts: (1) net property, plant, and equipment; (2) cost of goods sold; (3) selling, general, and administrative costs; (4) capitalized operating leases; (5) net research and development; (6) purchased goodwill; and (7) other intangible assets. In the second step, the resulting FirmEfficiency measure from step one is regressed against several firm characteristics1 to obtain a measure of MA, i.e., the residual firm efficiency not accounted for by firm characteristics. Demerjian, et al., (2012) validate their MA measure (MA score) by comparing it with other previously developed—and arguably imperfect— proxies for MA like CEO pay, earnings quality, and CEO turnover, and find that their MA score is superior to these proxies.2 We obtain the annually estimated FirmEfficiency scores from Peter Demerjian’s website. FirmEfficiency is the first stage, DEA-based measure of total firm efficiency, with values ranging from zero to one.3 Next, following Adut et al., (2019) we estimate the following Tobit regression model across firms by year and Global Industry Classification Standard (GICS) sector to obtain the measure of MA that we use in this analysis.4 These Tobit regressions are estimated annually for GICS sectors which allow us to obtain a MA score that is relative to other firms in the same industry which is a more novel approach than current methods in business valuation: FirmEfficiencyi,t = α + β 1 Ln(Assets) i,t + β2 MarketSharei,t + β3 D_Pos_FCFi,t + β4 FirmAge i,t + β5 BusSegConcentrationi,t + β6 D_ForeignCurrencyi,t + εi,t (1) 1 Demerjein et al., (2012) uses a linear estimation and regresses the total firm efficiency score from the first step DEA against: firm size, firm market share, cash availability, life cycle, operational complexity, and foreign operations. The residual from this estimation is the MA score. 2 These proxies of MA likely introduce measurement error, because they also contain factors beyond managers’ control and thus do not exclusively measure MA. 3 We want to thank Peter Demerjian for providing the MA data on his 4 A copy of the working paper is available from the authors upon request. Please reach out to where Ln(Assets) is the natural log of assets; MarketShare is the percentage of revenues earned by the firm in its Fama and French (1997) industry in year t; D_Pos_FCF is an indicator variable coded as one when a firm has nonnegative free CF and zero otherwise; FirmAge is the number of years the firm has been listed on Compustat at the end of year t; BusSegConcentration is the ratio of individual business segment sales to total sales, summed across all business segments for year t (if segment data is not available, the firm is assigned a concentration of one); D_ForeignCurrency is coded as one when a firm reports nonzero value for foreign currency adjustments in year t, and zero otherwise.5 The residual from Equation (1) represents MA, i.e., the part of firm efficiency that is not due to firm characteristics. Since this is a firm specific score, business analysts can use the corresponding score for the corresponding firm and can obtain individual scores. Alternatively, the analyst can obtain the firm-year specific variable MA score directly from Peter Demerjian’s website. Please see the “Practitioner’s Note” in the appendix available online at com/2019/19-JA-Appendix/. DATA AND EMPIRICAL ANALYSIS Univariate Analysis We obtain financial statements data from Compustat and merge this data with the previously discussed measure of MA. The Full sample consists of all annual firm-year observations with non-missing data required for our analysis. The Full sample contains 150,348 publicly traded firm-year observations for the twenty-nine-year period 1987–2015. The average sample year contains five,184 firm-years. The number of observations per sample-year range from a high of 6,843 for the year 1999, to a low of 805 for the year 1987. We allocate firms into Low, Medium, and High MA terciles by year. 5 For a more detailed discussion about how to measure MA please see Demerjian et al., (2012). A PROFESSIONAL DEVELOPMENT JOURNAL for the CONSULTING DISCIPLINES 14 JULY | AUGUST 2019 t h e v a l u e e x a m i n e r TABLE 1: DESCRIPTIVE STATISTICS AND UNIVARIATE COMPARISON Low-MAMed-MAHigh-MAHigh - Low MV 1.9791.9463.325 1.345*** BV0.5450.5460.602 0.056* NI (0.154)(0.084)(0.062) 0.092*** CF(0.031)0.0110.029 0.061*** ACCR(0.121)(0.092)(0.091) 0.030** Table 1 shows time-series averages of annually estimated means. Except for MA, all variables are measured in $ millions. ***, **, * indicates statistically significant difference at the one percent, five percent, and ten percent level, respectively. Statistical significance is estimated with two-tailed t-tests based on the parameter’s distribution across the twenty-nine sample years (1987 to 2015). All variables are unscaled and unwinsorized. Table 1 shows time-series averages of the annually estimated means for a firm’s MV, BV, NI, CF from operations, and ACCR by MA tercile. Using time-series averages (rather than pooled averages) has the advantage that each sample year receives an equal weight. The appendix at the end of the article shows a list of variable definitions. Except for MA, all variables are unscaled and measured at the firm-year level in $ millions at the end of the fiscal period. A comparison between the Low-MA tercile in Column (1) with the High-MA tercile in Column (3) shows that firms with high-ability management tend to be more valuable and more profitable in terms of NI, CF, and ACCR. Column (4) shows the difference between the Low and High-MA terciles. We estimate statistical significance with a two-tailed t-test based on the parameters’ distribution across the twenty-nine sample years. For all variables examined, the difference between the High and Low-MA tercile is statistically significant at the ten percent level or better. A PROFESSIONAL DEVELOPMENT JOURNAL for the CONSULTING DISCIPLINES t h e v a l u e e x a m i n e r JULY | AUGUST 2019 15 MAMVBVNICFACCR MA 0.1110.0380.0910.0710.068 MV 0.2330.340-0.454-0.423-0.262 BV 0.1250.5310.1360.0490.207 NI 0.2300.2660.3650.7970.661 CF 0.1190.1340.2030.6700.181 ACCR 0.1150.0520.1990.420-0.214 The top (bottom) triangle of Table 2 shows the time-series averages of the annually estimated Pearson (Spearman) correlation coefficients. Except for MA (MA), all variables are measured in $ millions. Bold font indicates that the correlation is statistically significant at the five percent level or better. Statistical significance is estimated with two-tailed t-tests based on the parameter’s distribution across the twenty-nine sample years (1987 to 2015). Table 2 shows the time-series averages of annually estimated correlation coefficients. The top (bottom) triangle is based on Pearson (Spearman) correlation coefficients. For ease of exposition, the following discussion will focus on the Spearman correlation coefficient shown in the bottom triangle. Bold font indicates that the coefficient is statistically significant at the five percent level or better. As shown in the MA column of Table 2, all variables are positively associated with MA. Consistent with the results presented in Table 1, this finding suggests that firms with high-ability management tend to be more valuable and more profitable in terms of NI, CF, and ACCR. Moreover, the correlation between MA and MV or NI, is almost twice as large as the correlation between MA and the other variables. Multivariate Analysis Ohlson (1995) provides the theoretical basis for the valuation model we are using in this study. Specifically, Ohlson (1995) models MV as a function of accounting fundamentals and so called “other” information which includes value relevant information that is not-yet captured by the accounting system. Following prior literature (e.g., Barth et al., 1998) we model MV as a function of BV and NI as shown in Equation (2). Specifically, we estimate the following cross-sectional regression each year: MVi = β0 + β1 BVi + β2 NIi + ε (2) Ohlson (1995) shows analytically that the NI-coefficient is increasing with earnings persistence and decreasing with the cost of capital.6 Given that firms with high-ability managers have more persistent earnings (see Demerjian et al., 2013) and/or lower cost of capital, we hypothesize that the NI-coefficient for firms with high-ability management is larger than the NI-coefficient for firms with low-ability managers. The widely used Ordinary-Least-Squares (OLS) regression approach suffers from two problems. First, the coefficient estimates can be strongly influenced by outlier observations. Second, heteroscedastic residuals create the need to scale variables (e.g., Ohlson and Kim 2015).7 Moreover, it is common 6 For a more detailed explanation, please see Equation (7) in Ohlson (1995), p. 670 ff. 7 In this context, the issue of heteroscedasticity refers to the circumstance in which the variability of the dependent variable is unequal across the range of values of the independent variables that predict it. TABLE 2: PEARSON AND SPEARMAN CORRELATIONS A PROFESSIONAL DEVELOPMENT JOURNAL for the CONSULTING DISCIPLINES 16 JULY | AUGUST 2019 t h e v a l u e e x a m i n e r practice to winsorize or truncate data to mitigate the influence of outliers. To overcome these issues, we follow prior literature and use the TS estimation approach. The methodology under the TS approach is intuitive and straightforward. Rather than running a single OLS regression based on an entire sample, the TS approach runs repeated OLS regressions on thousands of small sub-samples comprised of n randomly selected observations, where n equals the number of estimated parameters (i.e., intercept and coefficients). Given the small sample size for each of the sub-samples, the n-parameters estimated based on a given sub-sample may be heavily influenced by outliers. However, since the TS-parameter is estimated as the median across the parameter estimates generated by the thousands of randomly generated sub-samples, any undue influence from outlier observations or outlier parameters is largely removed. We employ the TS estimation method following an approach similar to Schaberl and Sellers (2017). Given the robustness of this approach, we use unscaled, unwinsorized firm-year level variables. First, for each sample year, we randomly draw 10,000 sub-samples with n-observations, with replacement. Second, for each sample year, we calculate the median for each parameter estimate in the regression. Third, we use two-tailed t-tests based on the distribution of the annually estimated parameters across the twenty-nine sample years to determine statistical significance for each parameter estimate. TABLE 3: TS COEFFICIENT ESTIMATES FOR BV AND NI MVi = β0 + β1 BVi + β2 NIi + ε (1)(2)(3)(4)(5) FullLowMedHighHigh - Low Intercept 5.1919.3783.4005.331 <0.01<0.01<0.01<0.01 BV 1.3731.2811.3201.5430.262 <0.01<0.01<0.01<0.01<0.01 NI 2.0250.8921.7874.0773.184 <0.01<0.01<0.01<0.01<0.01 Table 3 shows time-series averages of annually TS coefficient estimates. For each sample year, we randomly draw 10,000 sub-samples with n-observations, with replacement. Next, for each sample year, we calculate the median for each parameter estimate in the regression. Statistical significance is estimated with two-tailed t-tests based on the parameter’s distribution across the twenty-nine sample years (1987 to 2015). All variables are unscaled and unwinsorized. Table 3 shows the time-series averages of annually estimated TS regression coefficients. Column (1) shows the results based on the Full sample with 150,348 firm-years, while the results presented in Columns (2), (3), and (4) are based on the Low, Med, and High-MA terciles. Each tercile contains approximately 50,100 firm-year observations. To indicate statistical significance, p-values are reported in italics under the time-series averages of the coefficient estimates. The results presented in Column (5) show the difference and the statistical significance of the difference between the Low and High-MA column. The results presented in Column (1) show that BV and NI are significantly positively associated with MV. Consistent with our expectations, the NI-coefficient for the High-MA tercile is significantly larger than the NI-coefficient for the Low-MA tercile (4.077 vs. 0.892). In fact, the NI-coefficient for the High-MA tercile is over four times the size of the NI-coefficient for the Low-MA tercile. However, inferences based on averages A PROFESSIONAL DEVELOPMENT JOURNAL for the CONSULTING DISCIPLINES t h e v a l u e e x a m i n e r JULY | AUGUST 2019 17 (2) (1) - 1 2 3 4 5 6 7 8 19 87 19 88 19 89 19 90 19 91 19 92 19 93 19 94 19 95 19 96 19 97 19 98 19 99 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 Figure 1: NI-cofficient for High vs. Low MA High-MALow-MA can potentially be driven by outliers. To demonstrate that this result is consistent across time, Figure 1 shows that the NI-coefficient is larger for the High-MA tercile in every year of the sample period. In short, the results presented in Table 3 and Figure 1 indicate that investors perceive a dollar of NI as relatively more valuable when it is generated by a firm with high-ability management in charge. Sloan (1996) has documented that the CF component of NI is more persistent than the ACCR component of NI. Given that NI = CF + ACCR, our Equation (2) implicitly forces the coefficient for CF and ACCR to be equal. To allow the coefficients—and the implied persistence—to vary, we decompose NI into its cash and ACCR components and MV as a function of BV, from operations CF and ACCR as shown in Equation (3). Specifically, we estimate the following cross-sectional regression each year: MVi = β0 + β1 BVi + β2 CFi + β3 ACCRi + ε (3) TABLE 4: TS COEFFICIENT ESTIMATES FOR BV, CF, AND ACCR. MVi = β0 + β1 BVi + β2 CFi + β3 ACCRi + ε (1)(2)(3)(4)(5) AllLowMedHighHigh - Low Intercept 4.8928.2183.4984.826 <0.01<0.01<0.01<0.01 BV 1.3071.1861.2681.5200.333 <0.01<0.01<0.01<0.01<0.01 CF 2.9351.5432.6034.9533.410 <0.01<0.01<0.01<0.01<0.01 ACCR 1.3400.4401.0382.6272.187 <0.01<0.01<0.01<0.01<0.01A PROFESSIONAL DEVELOPMENT JOURNAL for the CONSULTING DISCIPLINES 18 JULY | AUGUST 2019 t h e v a l u e e x a m i n e r Table 4 shows time-series averages of annually TS coefficient estimates. For each sample year we randomly draw 10,000 sub-samples with n-observations, with replacement. Next, for each sample year, we calculate the median for each parameter estimate in the regression. Statistical significance is estimated with two-tailed t-tests based on the parameter’s distribution across the swenty-nine sample years (1987 to 2015). All variables are unscaled and unwinsorized. Table 4 shows the time-series averages of annually estimated TS regression coefficients. Column (1) shows the results based on the Full sample with 150,348 firm-years, while the results presented in Columns (2), (3), and (4) are based on the Low, Med, and High-MA terciles. To indicate statistical significance, p-values are reported in italics under the time-series averages of the coefficient estimates. The results presented in Column (5) show the difference and the statistical significance of the difference between the Low and High- MA column. Consistent with the idea that CF is more persistent than ACCR, the result presented in Column (1) shows a CF-coefficient that is more than twice as large as the ACCR-coefficient (2.935 vs. 1.340); untabulated results confirm that this difference is statistically significant (p < 0.001) for the Full sample as well as for each of the three MA-terciles. Moreover, this result is consistent across time. Specifically, the CF-coefficient is larger than the ACCR-coefficient in ninety-three percent, 100 percent, ninety-seven percent, and ninety-three percent of the sample years for the Full, Low-MA, Med-MA, and High-MA samples, respectively. The results presented in Columns (2) and (4) show that the CF-coefficient is significantly larger for the High- MA tercile relative to the Low-MA tercile (4.953 vs. 1.543). Similarly, the ACCR-coefficient is significantly larger for the High-MA tercile (2.627 vs. 0.440). As shown in Column (5), both of these differences between the coefficients across the High vs. Low-MA terciles are statistically significant. (2) - 2 4 6 8 10 19 87 19 88 19 89 19 90 19 91 19 92 19 93 19 94 19 95 19 96 19 97 19 98 19 99 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 Figure 2: CF-cofficient for High vs. Low MA High-MALow-MAA PROFESSIONAL DEVELOPMENT JOURNAL for the CONSULTING DISCIPLINES t h e v a l u e e x a m i n e r JULY | AUGUST 2019 19 (2) (1) - 1 2 3 4 5 6 7 19 87 19 88 19 89 19 90 19 91 19 92 19 93 19 94 19 95 19 96 19 97 19 98 19 99 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 Figure 3: ACCR-cofficient for High vs. Low MA High-MALow-MA To demonstrate that these results are consistent across time, Figure 2 (Figure 3) shows the CF-coefficient (ACCR-coefficient) for the High and Low-MA tercile for every year of the sample period. As shown in Figure 2, the CF-coefficient for the High-MA tercile is larger than the CF-coefficient for the Low-MA tercile in every sample year. Similarly, Figure 3 shows that the ACCR-coefficient is relatively larger for the High-MA tercile in every sample year but 2013. In short, the results presented in Table 4 and Figures 2 and 3 indicate that investors perceive a dollar of CF or ACCR as relatively more valuable when generated by a firm with high-ability management. Although a more thorough explanation of our findings is beyond the scope of this article, we propose the previously documented positive relationship between MA and earnings persistence as a likely explanation for our results (Demerjian et al., 2013). Another possible explanation is that firms with high ability managers have lower cost of capital, or higher growth and, therefore, a stronger association between MV and income. We are not aware of a study that has directly tested the relationship between MA and the cost of equity capital. However, the recent literature has documented that MA is positively associated with the quality of a firm’s information environment (Baik et al., 2018) and a firm’s credit ratings (Cornaggia et al., 2017), and negatively associated with bank loan prices (DeFranco et al., 2017). Taken together, these recent findings strongly suggest that firms with high-ability managers face lower cost of capital, which, in turn, should yield a larger earnings coefficient. Determining whether MA’s influence on earnings persistence or the cost of capital drives its impact on firm valuation could prove a fruitful avenue for future research. CONCLUSION In this study, we investigate whether MA influences the strength of the relationship between MV and several accounting fundamentals. To measure MA, we use a novel metric of MA based on Demerjian et al., (2012). Utilizing a sample of 150,348 publicly traded firm-year observations for the twenty-nine-year period, 1987– 2015, we employ the robust TS estimation methodology to estimate the relationships between MV and several accounting fundamentals. We find that the association between MV and NI is stronger for high- ability managers. To better understand the driver of our results, we repeat our analysis decomposing NI into its CF and ACCR components. We find that both components, CF and ACCR, have a stronger impact on MV for firms with high-ability managers. In short, our findings suggest that a dollar of income (be it CF or ACCR) is more valuable when generated by a firm with high ability management in charge. Next >