Brand Equity and Firm Sustainable Performance : The Mediating Role of Analysts ’ Recommendations

Current literature has overlooked the signaling effects of the brand on a firm’s sustainable performance through financial analysts. This study posits that financial analysts may serve as the information bridge connecting brand equity and a firm’s sustainable performance by providing professional recommendations of stock investments to public investors. Using a longitudinal archived dataset of Chinese listed firms, we found that: (1) brand equity improves the level of analysts’ recommendations for a focal firm’s stock, and also reduces the inconsistency of analysts’ recommendations; (2) industrial competition further strengthens the positive impact of brand equity on analysts’ recommendation level and strengthen its negative impact on recommendation inconsistency; (3) analyst recommendations mediate the relationship between brand equity and a firm’s sustainable performance in terms of abnormal return, systematic and idiosyncratic risk. These findings emphasize the importance of financial analysts’ recommendations in influencing the value of brand equity on sustainable firm performance.


Introduction
The way brand equity impacts firm sustainable performance has drawn considerable interest from academic scholars and practitioners [1].Brand equity is defined as the "added value" with which a brand endows a product [2].There are two different perspectives on evaluating brand equity.First, from the perspective of customers: customer-based brand equity refers to the different reactions of customers due to their brand knowledge [3].Second, from the perspective of firms: brand equity is evaluated based on the firms' product (e.g., premium price, high market share, penetration speed, etc., achieved compared to non-branded products) or financial performance (e.g., abnormal market return and lower risks).Increasing literature proposes that firm financial performance provides a more systematic and integrative channel to evaluate the value of brand equity [1,4], which also directly responds to the investors' demands of marketing accountability [1,5].
Previous literature has shown two major views on how brand equity influences firm sustainable performance: cognitive psychology and signaling theory, the first view, which has been adopted in most empirical brand studies, argues that brand equity can alter customers' perceptions and preferences if product information is delivered effectively.However, this view makes an implicit assumption that the market allows product information clearly and fully communicated between firms and their customers [6].By recognizing the fact of market imperfection that causes information asymmetry, the second view highlights the signaling effect of brand equity, which helps differentiate firms from competitors and generates superior firm sustainable performance [6][7][8].However, this view holds the assumption that customers make rational decision and that they accurately understand the signal of brand equity.This assumption largely overlooks the importance of signals' clarity and credibility.
According to signaling theory, the extent to which signals are delivered and correctly communicated in markets depends on the clarity and credibility of these signals [7].
Based on the above, three potential research gaps exist in the current literature on brand equity.First, existing studies have paid limited attention to the signaling effect of brand equity as well as the clarity and credibility of such signals on firm sustainable performance.Second, firm sustainable performance is essentially a risk-and-return balance, while most of the research has paid much more attention to return, ignoring the risk aspects [9].However, risk reflects performance stability and directly impacts a firm's financial ability and costs [1], thereby representing a significant aspect of firm sustainable performance.Third, few studies have investigated the boundary conditions of industrial factors, such as industrial competition, on influencing the signaling effect of brand equity.Industrial competition indicates the level of market uncertainty and risks, which substantially impact the effectiveness of signal and information processing in marketplaces [10].
To address these research gaps, this study adopts the signaling theory to empirically examine how brand equity affects firms' sustainable performance through the recommendations of financial analysts.Financial analysts are stock specialists, employed by investment banks and brokerage firms.They issue research reports regarding firm products, strategies, management quality and performance forecasts along with their recommendations on stock investments [11,12].Thus, financial analysts serve as information intermediaries between firms and investors by reducing the information asymmetry [13] and their recommendations aim to increase investors' utilities, because they can obtain information from both public and private channels that are not available for common investors.
Accordingly, our study has three research questions: (1) Does brand equity influence analysts' recommendations?(2) Does industrial competition affect the impact of brand equity on analysts' recommendations?(3) Do analysts' recommendations serve as a mediating mechanism in transferring the benefits of brand equity into firm sustainable performance?We investigated two important aspects of analysts' recommendations: level and inconsistency.The former indicates the analysts' invest opinions about whether a given stock is worth buying or selling [13,14] and the latter refers to the degree of disagreement on the recommendation levels to a given firm among analysts [13,15].Our main arguments are that brand equity has a positive impact on the level of analysts' recommendations and a negative impact on the inconsistency of analysts' recommendations, that such relationships are further strengthened by industrial competition and that analysts' recommendations mediate the relationship between brand equity and firm sustainable performance (return and risk).We used panel data of Chinese listed firms during the period of 2004 to 2016 to test our hypotheses, and data was collected from several sources (i.e., CSMAR (China Stock Market and Accounting Research database, the database official website is www.gtarsc.com),RESSET database (Database official website is www.resset.cn,which is one of major financial research database in China), WBL (World Brand Lab, the database official websites is www.worldbrandlab.com),etc.).We employed the generalized method of moments (GMM) to analyze the data in order to account for the potential endogeneity bias, because GMM has the best correcting effect comparing to other methods (i.e., control variable, lagged independent variable, firm or industry fixed effects, instrumental variable and lagged dependent variable) [16].
Our study contributes to the current literature in several ways.First, we newly uncovered an important mechanism through which brand equity influences a firm's sustainable performance-analysts' recommendations.As such, our study deviated from most existing research that has investigated brand equity based on cognitive preference views.Second, our study extended the research stream in the marketing-finance interface by closely connecting marketing asset (i.e., brand equity) and financial factors (i.e., analysts' recommendations) in jointly affecting firm sustainable performance.Third, our study further explored important boundary conditions on the relationship between brand equity and analysts' recommendation by examining the moderating role of industrial competition.Overall, our findings provide novel insights into the important role brand equity plays in the assessment systems of analysts to develop their stock recommendation, and in turn, affect firms' sustainable performance.

Signaling Effect of Brand Equity
Facing information asymmetry caused by market imperfection, customers tend to collect information in order to reduce perceived risks and uncertainties.However, information is usually either difficult to access or very costly [8].Under such circumstances, brand equity representing importance market signals can effectively influence customers' decisions if these signals meet two criteria-clarity and credibility [7].Brand clarity means the content of product information is without ambiguity [8], which can increase credibility.Brand credibility refers to customers' perceptions of firms in having both the ability (i.e., expertise) and willingness (i.e., trustworthiness) to continuously deliver what has been promised in brands statements [8].If and when firms do not deliver what has been promised, brand equity will erode.
Apart from brand equity, firms can also use tools of the marketing mix to communicate information to customers, such as advertisement [17] and warranty [18].However, the signaling effect of brand equity differs greatly from that of the marketing mix.First, since brand equity is a long-term accumulated asset resulting from the interactions of firms' previous branding activities, strategies and customers, it should have a higher level of signal credibility than that of short-term marketing mix tools.Second, brand equity is a multi-faceted complex construct [19], and thus, shows a lower level of signal clarity because of the difficulties in assessing the consistency in firms' previous marketing activities and brand investments.Therefore, brand equity represents signals with high credibility but low clarity for most of the customers.
According to signaling theory, credibility plays a considerably important role in imperfect markets because of information asymmetry.When the credibility and clarity of signals are enhanced, brand equity can influence customers' purchase decisions by increasing customers' perceived quality, reducing their perceived risks and information costs (i.e., including both information searching and processing cost) [7].

The Influence of Brand Equity on Stock Analysts' Recommendations
Previous literature has shown that firms' intangible assets (e.g., brand equity) have significant impacts on analysts' recommendations in terms of analysts' forecast choices, accuracy and coverage [20][21][22].For example, White, Lukas and Hill (2007) proposed that the regularity and comprehensiveness of firms' reporting on their intangible value (i.e., brands, customer relationship and market knowledge) increase stock analysts' assessment accuracy [20][21][22].Barth, Kasznik and Mcnichols (2001) found that firms with more intangible value (e.g., Research and development (R&D) and brand value) have a greater analysts' coverage because intangible assets are often difficult to identify, and hence, estimates require a greater effort of analysts [20].Furthermore, perceived quality, one of the important sub-dimension of brand equity, has been found to largely influence the analyst earnings forecast [23,24].Moreover, Luo (2010) found that a positive (negative) change in product competitiveness increases the likelihood of beating (missing) analyst earning forecasted target [22].
Drawing on the previous literature, we argue that strong brand equity increases the level of analysts' recommendations.Brand equity can signal the level of expected firm performance, based on which analysts' develop their potential stock recommendations [13,25,26].Specifically, strong brand equity potentially contributes to future firm performance: (1) brand equity is a valuable, rare, non-imitable and non-substitutable firm resource; hence, it represents an important source of a firm's sustainable competitive advantage [27].Thus, brand equity can be an indicator or signal of a firm's future superior performance.Srivastava, Shervani and Fahey (1998) also asserted brand equity as one of the most important market-based assets, which could increase firms' financial performance in terms of cash flow (on both magnitude and speed) [1].(2) Strong brand equity indicates a good customer relationship for a firm, which is beneficial for generating superior financial and strategic performance in the future [28].
Strong brand equity usually results from a firm's superior performance with its customers (e.g., customer satisfaction, customers' willingness to pay, repurchase rate, etc.), signaling a positive future financial performance [29].Existing research finds that strong brand equity also enhances a firm's market and strategic performance, because customers often have an affective bond with firms possessing strong brand equity and they are more willing to interact with and repurchase from the firm [30,31].Thus, brand equity represents firm assets that can generate sustainable competitive advantage and superior customer relationship, contributing to the firm's superior performance in the future.Analysts develop their stock recommendations based on several firm factors, among which positive financial and strategic performance is of most significance [28].Therefore, strong brand equity has a positive influence on analysts' recommendation.We advance the following hypothesis: H1: Brand equity increases the level of analysts' recommendations.
Market-based assets (e.g., brand equity) signals a firm's sustainable and stable financial performance, which indicates less vulnerability and volatility, thus facilitating a consensus agreement among analysts [1].
Brand equity enhances the sustainability and stability of a firm's financial performance for two main reasons.First, brand loyalty, as one of its sub-dimensions [23], is a firm's valuable, rare and non-imitable asset that helps a firm raise the barriers of industry entry and reduces the risks of financial loss caused by fierce competition [27].Second, brand loyalty can buffer the fluctuation of a firm's financial performance in the face of strong competition and market uncertainty.Brand loyalty reduces customers' sensitivity to new marketing tactics of competitors and the likelihood of customers' switching to competing brands, which greatly helps the firm effectively address market uncertainties and dynamism [32].Furthermore, previous research has shown that customers' brand loyalty is negatively associated with the churn rate [33].
In view of the above, the strong customer base developed by brand equity enhances the stability and sustainability of a firm's future financial performance.Therefore, analysts are more likely to reach a consensus regarding their stock recommendations with respect to the focal firm.Hence, brand equity is expected to be negatively associated with the inconsistency of analysts' recommendations [26].
H2: Brand equity decreases the inconsistency of analysts' recommendations.

The Moderating Role of Industrial Competition
Industrial competition reflects the degree of rivalry in the industry [33]; we argue that industrial competition may influence the effect of brand equity on analyst' recommendations.In highly competitive industries, customers face increased market uncertainty and risks of purchase decisions.Brand equity plays a more salient role in signaling a firm's product quality and sound performance, and thus reduces customers' perceived uncertainty and risks and enhances customers' purchase intention [33,34].Therefore, the stronger association between brand equity and a firm's future financial performance makes analysts more relying on brand equity to evaluate firms and recommend investment choices to public investors.
On the contrary, in less competitive industries, a few firms control the majority of the resources and markets tend to be highly concentrated.In such situations, the effectiveness of marketing activities and market-based assets on firm financial performance is decreased [35].Firms are less paid-off from their investments on brand equity, reducing the signaling role of brand equity for analysts' recommendations.Thus, we propose two hypotheses as below: H3a: As the industrial competition increases, the positive relationship between brand equity and analysts' recommendations level is strengthened.
H3b: As the industrial competition increases, the negative relationship between brand equity and analysts' recommendations inconsistency is strengthened.

The Mediating Role of Analysts' Recommendations
Brand equity, as one of the most important firms' important intangible assets, contributes to shareholder value in terms of both return and risk [36,37].For example, Mortanges and Riel (2003) indicated that brand equity, which is measured by "strength" and "status", significantly impacts firm value [38].Rego, Billett and Morgan (2009) found that a firms' customer-based brand equity (CBBE) is negatively associated with firms risk and had a stronger variance explaining capability than the existing financial model [9].Despite the positive role of brand equity, it is a challenge for public investors to identify and assess the complex intangible assets [1,21].Brand equity may not automatically transform into firm value because of information asymmetric between firms and public investors.
Information asymmetry often exists between firms and public investors [13].In the case of brand equity, information asymmetry arises from the lack of adequate level of signal clarity, which makes brand equity difficult and costly to be analyzed for public investors.First, though brand equity represents a key signal of a firm's future growth, it is rarely disclosed by firms.Thus, the lack of timely and public disclosure reduces the signal clarity of brand equity and exacerbates the information asymmetry between firms and public investors.Second, public investors often face difficulties in fully understanding brand equity because they lack access to important information channels and an adequate level of cognitive resources to accurately interpret mess information [35,39].Finally, brand equity is a complex and multi-faceted firm asset that has been developed through long-term accumulation, which increases signal credibility of brand equity but decreases its clarity and makes it difficult to be evaluated by public investors [19].To overcome difficulties of signal clarity, public investors tend to rely on external cues [40] such as analysts' recommendations to support their evaluation and investment decisions [35,40].
We argue that analysts' recommendations increase the signal clarity of brand equity and help reduce the information asymmetry between firms and public investors.First, brokerage firms spend lots of money and resources analyzing stocks, and public investors are willing to pay for investment suggestions (i.e., analysts' recommendations) because of the induced benefits of these suggestions usually overweight cost [41].Thus, financial analysts provide clear recommendations of stocks to investors, increasing the clarity of brand signal.Second, financial analysts can get access to not only public information, but also private information channels and resources that are usually not available for outsiders (e.g., public investors) [42].Moreover, financial analysts are specialized in information collection and processing, and they have spent plenty of time and effort on extracting useful information that can effectively predict firms' future cash flow from various signals of brand equity [43,44].Finally, analysts' forecast accuracy can be easily assessed by public investors based on prior records of analysts.Investors often use their assessments of analysts to help understand the complicated phenomena of brand equity.Therefore, financial analysts provide recommendations with clarified and credible information of brand equity, which can support investors' investment decisions by reducing investors' perceived uncertainty, and their risk and costs in information searching and processing.
According to prior accounting and finance literature, the analyst's recommendation do influence firm performance.For instance, Srinivasan and Hanssens (2009) asserted that analysts' recommendations have significant impacts on firms' sustainable performance because analysts' recommendations are more persuasive than other information channels for public investors due to their specialized reputation [45].Womack (1996) found that analysts' recommendations (both selling and buying recommendations) influenced the stock price significantly, both in a short (i.e., a few days) and long (i.e., several months) period [41].Jegadeesh et al. (2004) showed that analysts' recommendations influence firms' stock prices by helping public investors assess firm performance more accurately [25].
Current research has shown that analysts play an important intermediates role between firms and the public when relevant information is ambiguous and difficult to evaluate.For example: analysts' recommendations mediated the corporate social performance influence on their financial performance [44].In addition, other research has emperically found that Internet Technology (IT) investment has an impact on firm sustainable performance through the mediation of analysts' recommendations [43].Therefore, based on previous literature, we propose that analysts' recommendations can serve as the intermediate transferring the benefits of brand equity into firm sustainable performance.Thus, we develop the hypothesis as below: H4: Analysts' recommendations (level and inconsistency) mediate the relationship between brand equity and firm sustainable performance (return and risk).

Research Framework
We depict our conceptual model in Figure 1. Brand equity is positively associated with the level of analysts' recommendations, and negatively with the inconsistency of analysts' recommendations.Such relationships are more pronounced in highly-competitive industries.Moreover, analysts' recommendations mediate the relationship between brand equity and firm sustainable performance in terms of return and risks.We considered several control variables.For instance, we needed to control the influences of finance and accounting factors on firm sustainable performance such as firm size, Return on asset (ROA), ROA volatility, dividend, leverage, and liquidity.Additionally, we controlled the influences of analysts' personal characteristics, such as analysts' tenure, forecast accuracy and coverage.Lastly, we controlled for CEO demographic variables, such as CEO gender, age and salary.
Current research has shown that analysts play an important intermediates role between firms and the public when relevant information is ambiguous and difficult to evaluate.For example: analysts' recommendations mediated the corporate social performance influence on their financial performance [44].In addition, other research has emperically found that Internet Technology (IT) investment has an impact on firm sustainable performance through the mediation of analysts' recommendations [43].Therefore, based on previous literature, we propose that analysts' recommendations can serve as the intermediate transferring the benefits of brand equity into firm sustainable performance.Thus, we develop the hypothesis as below: H4: Analysts' recommendations (level and inconsistency) mediate the relationship between brand equity and firm sustainable performance (return and risk).

Research Framework
We depict our conceptual model in Figure 1. Brand equity is positively associated with the level of analysts' recommendations, and negatively with the inconsistency of analysts' recommendations.Such relationships are more pronounced in highly-competitive industries.Moreover, analysts' recommendations mediate the relationship between brand equity and firm sustainable performance in terms of return and risks.We considered several control variables.For instance, we needed to control the influences of finance and accounting factors on firm sustainable performance such as firm size, Return on asset (ROA), ROA volatility, dividend, leverage, and liquidity.Additionally, we controlled the influences of analysts' personal characteristics, such as analysts' tenure, forecast accuracy and coverage.Lastly, we controlled for CEO demographic variables, such as CEO gender, age and salary.

Sample
Data collection: the data for our study was collected from three sources.The data of brand equity was obtained from the World Brand Lab (WBL).We collected the stock price, risk free, stock analyst's recommendation level and inconsistency, accounting and financial data from China Stock Market & Accounting Research (CSMAR).We collected Fama-French three factors data from RESSET.
Sample screening and processing: our sample excluded financial firms and B shares, because financial firms are very different from other firms and B share stock are trading in a different currency from RMB (i.e., B shares are trading in U.S. dollar in the Shanghai Stock Exchange and trading in HK dollar in the Shenzhen Stock Exchange).We winsorized some financially related

Sample
Data collection: the data for our study was collected from three sources.The data of brand equity was obtained from the World Brand Lab (WBL).We collected the stock price, risk free, stock analyst's recommendation level and inconsistency, accounting and financial data from China Stock Market & Accounting Research (CSMAR).We collected Fama-French three factors data from RESSET.
Sample screening and processing: our sample excluded financial firms and B shares, because financial firms are very different from other firms and B share stock are trading in a different currency from RMB (i.e., B shares are trading in U.S. dollar in the Shanghai Stock Exchange and trading in HK dollar in the Shenzhen Stock Exchange).We winsorized some financially related variables such as ROA, leverage and liquidity at 1% and 99% percentiles in order to avoid the influence of extreme values.We lagged independent variable, moderator, mediators and all control variables for one year, in order to avoid reverse causality and reduce the potential endogeneity bias [46,47].
After merging data from different sources, we finally obtained 833 firm-year observations, with a time window spanning from 2004 to 2016.There are 217 firms in the final sample, which constituted a sample rate of 25.22%.The most durable firms in the sample had 12 years of consecutive observations, and the least durable ones had only a one-year observation.

Measurement
Dependent variables: the measure of firm sustainable performance should capture both abnormal return and risks (systematic and idiosyncratic risks) [45].The return aspect concerns about the speed and magnitude of cash flow and risk refers to the volatility and vulnerability of firm performance [1].The abnormal return was measured by the difference between firms' expected and actual returns in this study.Moreover, we employed the Fama-French three model to calculate the expected return: where R it is daily returns on stock of firm i on day t in fiscal year, R ft is daily risk-free return on day t, R mt is daily returns on a value-weighted market portfolio on day t, SMB t represents small minus big in market capitalization, and HML t represents high minus low in market to book ratio.Systematic risk refers to the variability commonly shared by a firm and the whole market.Systematic risk was measured by using a β 1i that is the slope coefficient of market factors [48].Idiosyncratic risk refers to firm-specific variability that is not covariant with the entire stock market.It is measured by using the standard deviation of residual in formula (1).
Independent variable: brand equity was measured using the ranking of China's 500 Most Valuable Brands from the WBL.While this ranking list contains both listed and non-listed firms, we included only the listed-firms into our sample because of the lack of availability of financial data of non-listed firms.
The WBL, established in New York City in 2003, is a leading independent consultancy specialized in brand valuation and marketing strategy in the world.WBL has created its own tools for assessing the added value of the brand and has applied it to over 1000 large-scale companies worldwide.WBL provides information about brand equity rank, brands' name, institutions' name, brands' value, influence (three tiers: world, China, local), listed (indicate whether it is a listed firm), birthland (the provinces where firms are established) and main business.In most cases, only one stock id code is associated with one brand name.
Unlike Interbrand, which evaluates firms' future earnings using a discount rate method, the WBL can predict both the firms' future earnings and the percentage of earnings generated by brands by using its unique brand value added (BAV) tools.The WBL uses formula (2) below to evaluate the brand value: where E refers to a weighted average of five years' annual earnings, which include the past three years (including the current year) and the next two years.The WBL uses the "Economic added value" (EVA) method to predict firms' annual earnings that indicate the net profit minus the equity cost of the firm's capital.BI refers to brand value added index, which refers to the percentage of earnings generated by brand.It is generated from the BAV tool, originally created by the WBL.S refers to the brand strength coefficient in consideration of the industrial and markets' unique characteristics, consisting of eight factors-industry features, external support, brand recognition, brand loyalty, leadership, brand management, ability to expand and brand innovation.Taking industry features as an example, the brand strength coefficient should be adjusted depending on specific industries in consideration of levels of competition and entry barrier.
Mediators: Level of analysts' recommendations is measured by using the median value of the levels of all analysts' recommendation (latest forecast in the year-end) for a given firm.There are five standard recommendation levels in CSMAR database, and we coded them as the following: 5 = "Buy", 4 = "Overweight", 3 = "Hold", 2 = "Underweight" and 1 = "Sell".Inconsistency of analysts' recommendations refers to the degree of disagreement on the recommendation levels to a given firm among analysts.It is measured by using the standard deviation of analysts' recommendation levels.
Moderators: Industrial competition.In order to measure industrial competition, we employed the calculation method of the Herfindahl index.Following the previous literature [33], we calculated the sum of squared market shares of the firms in the industry based on sale revenues from industrial classifications used by Chinese Security Regulatory Commission (CSRC).Following the previous Chinese research [49], we used the two-digit code in the manufacturing industry and one-digit code in others.That is, industrial sales Herfindahl j = ∑ i I s ij 2 , where s ij is the ratio of the firm's sales to the total sales of industry j to which firm i belongs [50].We used Herfindahl j multiple (−1) as the indicator of industrial competition; thus, the larger the indicator the stronger the industrial competition.Control variables.In order to rule out the alternative explanations that may influence our conclusions, we included several financial and accounting related control variables, such as return on assets (ROA), ROA volatility, firm size, dividend, leverage, liquidity and age.ROA was measured as the ratio of net profit to the total assets.ROA volatility was measured as the moving average for the past five years of the standard deviation of ROA.Firm size is a fundamental firm characteristic that may affect a firm's assets accumulation and sustainable performance [51].According to Dang, Li and Yang (2018), there are three types of proxies variable to measure firm size: total assets (appropriate for control the influence of firm size in product markets), total sales (appropriate for control the influence of firm size in stock market) and market value of equity (appropriate for control the influence of firm size in terms of firms' resources that can generate profits) [51].In this study, we chose to measure firm size using the natural logarithm of total assets to control the influence of firm resources on a firm's ability to develop brand equity and generate financial performance.Dividend was measured by the ratio of cash dividend to firm market capitalization.Firm leverage was measured as the ratio of long-term debt to the total assets.Firm age refers to the number of years since a firms' initial public offering in Shanghai/Shenzhen Stock Exchange.Assets were measured as the natural logarithm of firms' total assets.
In addition, we controlled the analyst demographic variables, which may influence both analysts' recommendation and firm sustainable performance, such as their tenure, coverage and forecast accuracy.Analyst tenure is measured by the analysts' working duration counted by the months since their initial forecast.Analysts' tenure reflects their forecasting experience and recommendation accuracy [13,26].Previous literature has indicated that analysts' experiences are positively associated with earning higher forecasting accuracy [52,53].For instance, analysts with more firm-specific experience can improve stock price efficiency reflecting in decreased accrual anomaly [54].Analyst coverage is measured as the number of analysts who follow the stock of a given firm.According to prior literature, analysts have greater incentives to cover firms with more intangible assets [20]; hence, analyst coverage influences forecasting accuracy.Forecast accuracy represents analysts' earning forecasting errors, and is measured by the difference between the latest analyst's median value of forecasted earning and the actual earnings scaled by corresponding stock price.Forecast accuracy of analysts potentially influences the effect of stock recommendations of each analyst [20,55].Loh and Mian (2006) find analysts' reputation on forecasting accuracy strongly indicates the level of possible earnings of analysts' recommended stock [56].
Furthermore, corporate governance is also an important factor affecting analysts' recommendations and firm sustainable performance.According to prior research, corporate governance can be investigated from several aspects such as market competition [57], CEO tournament [58], CEO compensation incentives [59] and mutual monitoring among executives [60].In our study, we included industrial competition or market competition in the model, as it directly affects firm performance and the effectiveness of analysts' recommendations.In addition, we controlled several CEO factors to reflect various aspects of corporate governance, such as Female CEO, CEO age and CEO salary.Female CEO is a binary variable that indicates whether a given firm's CEO is female.Previous literature revealed that firms run by female CEO are more conservative, less risk-taking and less indebted [61].CEO age was measured as how old the CEO is, which may influence CEO turnover and firm risk [62].Previous literature found that CEO age is negatively associated with firms' acquisitions [63].CEO salary captures the cash compensation of CEO's remuneration.[64].Previous literature used cash compensation to test tournament theory [64].
Finally, we controlled the time and industrial fixed effects.Table 1 summarizes all variables' measurement and data sources.Systematic risk.This reflects the variability in a firm's stock return associated with macroeconomic event that affect the entire stock market.Based on the Fama-French three factor model, it is the slope of market factor.

RESSET, CSMAR Idiosyncratic risk
Idiosyncratic risk.This refers to the variability that is not explained by changes in the average market portfolio return, but instead by firm-specific events.Based on the Fama-French three factor model, it is the standard deviation of residual.

Brand equity
Brand equity value.The brand equity value based on the China's 500 Most valuable Brands List, as assessed by the WBL.The ranking included both listed and non-listed firms in China.We used the listed firms only, then took the logarithm of value to represent the brand equity.

Level of analyst recommendations
Analyst recommendation level.There are five standard recommendations in CSMAR; we coded them as follows: 5 = Buy; 4 = Overweight; 3 = "Neutral/Hold"; 2 = "Underweight"; 1 = "Sell".We used the median value of all recommendation for a given firm to represent the analyst recommendation level.We used the latest forecast of the end of each year.

Inconsistency of analyst recommendations
Analyst recommendation inconsistency.

Model Specification
Firstly, in order to test the influence of brand equity on analysts' recommendations and the moderating effect of industrial competition, we used the following formula (3) and ( 4): where Level it is the analysts' recommendations level, Incon it is analysts' recommendations inconsistency, Comp it refers to industrial competition, Equity it refers to brand equity, Controls it refers to all control variables, α 0 and θ 0 are all intercepts.βc and θ c are control variables' coefficient vectors, ε it1 is the error term, its variance is σ ε1 2 , and i refers to firm i, while t refers to year t.In order to control the penitential bias (i.e., heteroscedasticity, serial correlation), we employed the generalized method of moments (GMM) to estimate our model, which is based on moment condition instead of the density function.We used the White Heteroskedasticity-and autocorrelation-consistent (HAC) covariance matrix, e.g., Φ HAC , [formula ( 5) and ( 6)].Where ω is White error vectors, q is band, k is the kernel function, Z t is a K*P matrix [46].
Φ HAC = Γ (0) + [ ∑j=1 T−1 k(j,q)Γ (j) + Γ (j)] ( 5) To test the main effect of brand equity on the level (Hypothesis 1) and inconsistency of analysts' recommendations (Hypothesis 2), we regressed the recommendation level and inconsistency on brand equity.To test the moderation effect of industrial competition (Hypothesis 3), we followed Aiken and West's (1991) [65] approach to include the interaction between brand equity and industrial competition as specified by Equations ( 3) and (4).To avoid the multicollinearity problem, we mean-centered the variables before generating the interaction terms.
In order to test the mediation effect of analysts' recommendations between brand equity and firm sustainable performance (Hypothesis 4), we followed Baron and Kenny's (1986) [66] three-step method.In the first step, we regressed mediators (i.e., level and inconsistency of analyst recommendation) against the independent variable (i.e., Brand equity) as specified by Equations ( 3) and (4).In the second step, we regressed dependent variables (i.e., abnormal return, systematic and idiosyncratic risk) against independent variable as specified by Equations ( 7)- (9).Finally, in the third step, we regressed dependent variables against the independent variable and mediators as specified by Equations ( 10)- (12).(12) where the Ab it refers to abnormal return, Srisk it refers to systemic risk and Irisk it refers to idiosyncratic risk.

Endogeneity Correction
To address the potential endogeneity problem in the estimation model, we took several measures.First, in order to alleviate the potential reverse causality, we lagged our independent variable, moderator and all control variables for one period [67,68], i.e., the data on independent variable (i.e., brand equity), moderator (i.e., industrial competition) and controls variables were from 2004 to 2015, while data on mediators (i.e., analysts' recommendations) and dependent variable (i.e., firm sustainable performance) were from 2005 to 2016.Second, we employed a hackman two-stage model [69] to account for the sample selection bias.Therefore, we ran a logit model in the first stage to calculate the inverse mills ratio (Mills) using a full sample and included Mills in the following second stage models to test the mediating effect of analysis recommendation level and inconsistency and moderating effect of industrial competition using brand equity sample.We used consistent control variables in both two-stage models.See the correlation matrix and the first stage results in Appendix A Tables A1  and A2.Third, in accordance with previous studies [70], we employed a GMM estimation specification that had the greatest correction power on endogeneity bias compared to other single methods (e.g., instrument variable, fixed effect model, lagged dependent variables and control variables) [16].Fourth, following previous literature [51,[58][59][60], we included a set of firm-, CEO-, and analyst-level control variables that might affect brand equity and firm performance simultaneously.Firm-level controls include firm size, ROA, ROA volatility, leverage, dividend, liquidity and firm age.CEO-level controls involved CEO gender, CEO age and CEO salary.Analyst-level controls captured analysts' tenure, analysts' coverage and analysts' forecasting accuracy.These control variables potentially reduce the potential likelihood of endogeneity bias caused by omitted variables that confound the relationship between the independent variable and the dependent variable [71].

Result
Table 2 shows all variables descriptive statistics and correlations.The correlations between brand equity and analyst recommendation level and inconsistency were 0.23 (p < 0.05) and −0.07 (p < 0.05), respectively.The correlations between brand equity and abnormal return, systematic risk and idiosyncratic risk were 0.09 (p < 0.05), −0.23 (p < 0.05) and −0.15 (p < 0.05), respectively.We further investigated the multicollinearity problem by computing the variance inflation factors (VIFs).The maximum VIF obtained in any of models was 2.83 (M8), and the mean value was 1.66, all of which were much lower than the suggested cutoff of 10.00 [72].Therefore, multicollinearity was not a problem in our results.
Table 3 presents the GMM regression results of the main effects and moderating effects.Model 1 and Model 4 report the main effect of brand equity on the level and inconsistency of analysts' recommendations respectively.Model 2-3 and Model 5-6 report the effects of industrial competition and the interaction term of industrial competition and brand equity sequentially.Table 4 presents the GMM regression results of the mediation effect of analysts' recommendations.Model 7, 9, 11 reports the direct effects of brand equity on abnormal return, idiosyncratic risk and systematic risk, respectively.Model 8, 10, 12 reports the effect of brand equity on abnormal return, idiosyncratic risk and systematic risk when controlling two mediators-level and inconsistency of analysts' recommendations.

The Effect of Brand Equity on the Level and Inconsistency of Analysts' Recommendations
In Model 1, the result shows the main effect of brand equity on the level of analyst recommendations is significantly positive (b = 0.048, p < 0.01), which supports Hypothesis 1.In Model 4, it shows the main effect of brand equity on the inconsistency of analyst recommendation is significantly negative (b = −0.019,p < 0.05), which supports Hypothesis 2.

The Moderating Effect of Industrial Competition
Hypothesis 3a states that industrial competition positively moderates the relationship between brand equity and the level of analysts' recommendations.Model 3 shows that the interaction between brand equity and industrial completion has a significantly positive effect on the level of analyst recommendation (b = 0.027, p < 0.05), supporting Hypothesis 3a.
Hypothesis 3b states industrial competition negatively moderates the relationship between brand equity and inconsistency of analyst recommendation.Model 6 shows that the interaction between brand equity and industrial completion has a negative effect on the inconsistency of analyst recommendation (b = −0.012,p < 0.01), supporting Hypothesis 3b.

The Mediating Effect of Analysts' Recommendations
With regard to Hypothesis 4, it examines the mediating role of analyst recommendation level and inconsistency between brand equity and firm sustainable performance (i.e., abnormal return, systematic risk and idiosyncratic risk).According to Baron and Kenny (1986) [66], three conditions must be met to establish mediation: (1) brand equity must significantly affect analyst recommendation level and inconsistency; (2) analyst recommendation level and inconsistency must significantly affect firm sustainable performance; (3) when mediators are controlled in the model, the coefficients of brand equity become either nonsignificant or less significant.As shown in Model 1 and 4 in Table 3, brand equity significantly affects analyst recommendation level and inconsistency.In addition, the results in Model 8, 10, and 12 in Table 4 show that analyst recommendation level and inconsistency significantly affect firm abnormal return, idiosyncratic and systematic risks, respectively.As reported in Model 8 Table 4, after including mediators in the model, the impact of brand equity on abnormal return is no longer significant (from b = 0.002, p < 0.05 to b = 0.001, p > 0.10 in Model 8), supporting a full mediating role of recommendation level and inconsistency in relationship between brand equity and abnormal return.
After controlling analysts recommendation level and inconsistency, the significant impact of brand equity on idiosyncratic risk was reduced (from b = −1.14× 10 −3 , p < 0.01 to b = −0.91 × 10 −3 , p < 0.01 in Model 10), in support of a partial mediation of analysts recommendation level and inconsistency; the effect of brand equity on systematic risk was also reduced (from b = −0.025,p < 0.01 to b = −0.021,p < 0.05 in Model 12), in support of a partial mediating effect of analysts recommendation level and inconsistency in relationship between brand equity and systematic risk.Therefore, the results support Hypothesis 4 in that analysis recommendation level and inconsistency fully mediate the relationship between brand equity and abnormal return, and partially mediate the relationship between brand equity and firm risks (both idiosyncratic risk and systematic risk).
Furthermore, we conducted a Sobel test (1982) [73] for mediation to assess whether the indirect mediation effects were statistically significant.The calculation formula is , where a and S a are the coefficient and standard error, respectively, for the impact of the independent variable on the mediator, and b and S b are coefficient and standard error, respectively, for the impact of mediator on the dependent variable.In Tables 3 and 4, the results show that brand equity's direct effects on abnormal return were not significant (0.001, p < 0.01), while its indirect effect through two mediators was (0.027 × 0.004) = 1.08 × 10 −4 .The direct effect of brand equity on idiosyncratic risk was −0.91 × 10 −3 , its indirect effect through mediators was [0.027 × (−0.002) + (−0.012 × 0.003)] = −0.90× 10 −4 .In addition, brand equity' direct effect on systematic risk was −0.021, and its indirect effect was [0.027 × (−0.041) + (−0.012 × 0.043)] = −1.62 × 10 −3 .The results of the Sobel test show that all mediation effects were significant, except one: the mediation effect of analysts' recommendation inconsistency in brand equity-idiosyncratic risk link (Z = 1.46, p > 0.10).Therefore, these results provide further support for Hypothesis 4 in the mediating role of analysts' recommendations in the relationship between brand equity and firm sustainable performance.
According to the results in Table 3, we found that analyst tenure and analyst forecast accuracy did not have a significant impact on analyst recommendation level (b = 0.001, p > 0.05; b = −0.196,p > 0.05, respectively) and analyst recommendation inconsistency (b = 0.002, p > 0.05; b = −0.212,p > 0.05, respectively).Analyst coverage significantly and positively related to the analyst recommendation level (b = 0.006, p < 0.05) and negatively related to analyst recommendation inconsistency (b = −0.002,p < 0.10).This is consistent with previous literature [20], in that analysts would expand greater effort to follow firms with more intangible assets in order to increase their forecast accuracy.Firm size had a positive significant impact on both analyst recommendation level (b= 0.081, p < 0.01) and inconsistency (b = 0.023, p < 0.01).ROA, dividend, liquidity and CEO salary all had a significant positive influencing analyst recommendation level, while firm level significantly and negatively influenced the analyst recommendation level.ROA volatility significantly and positively influenced analyst recommendation inconsistency, while dividend and CEO salary significantly and negatively influenced inconsistency.

Conclusions
In our study, we assembled longitudinal data from multiple data sources to test our hypothesis.In order to reduce the potential endogeneity bias, we lagged one period for all our explanatory variables, employed a GMM model specification and controlled a set of variables that may influence both brand equity and firm sustainable performance.In addition, we used a Heckman two-stage model to account for the sample selection bias.Building on signaling theory, this study examined how brand equity affects firm sustainable performance (i.e., abnormal return, idiosyncratic risk and systematic risk) through the mediating effect of analysts' recommendations, as well as the moderating role of industrial competition.Based on archived data from multiple sources on Chinese Listed firms, the findings highlighted the mediation role of analyst recommendation level and inconsistency in transferring the benefits of brand equity into firm sustainable performance, as well as the importance of the industrial competition in strengthening the benefits of brand equity.These findings provided several contributions to the theory and managerial practices.

Theoretical Contribution
First, our study contributes to brand equity literature by uncovering the signaling role of brand equity in enhancing firm sustainable performance.Most of the existing literature has investigated the benefits of brand equity from the view of cognitive psychology [9,21,74].Extending on that, this study adopted the view of signaling theory, and focused on the signaling effect of brand equity on analysts and firm sustainable performance-which is an overlooked aspect of brand equity in existing literature [7].By examining the mediating role of analysts' recommendations, this study provides an understanding of how brand equity represents an important market signal that would affect firm sustainable performance.Overall, the findings highlight that brand equity enhances firm sustainable performance via its signaling effects on analysts.
Second, this study enriches the research on the marketing-finance interface by exploring the important role of analysts' recommendations in linking market-based assets (i.e., brand equity) with firm sustainable performance.The role of financial analysts is under-studied in brand equity research [21].Our study responds to Luo, Homburg and Wieseke's (2010) [13] call for more research on the intermediate role of financial analysts in bridging firms' marketing and financial information (e.g., firm sustainable performance).Our findings indicate that financial analysts' recommendations are significant mediating mechanisms transferring the benefits of brand equity into firm sustainable performance.As such, this study revealed the important role of financial analysts in linking a new key marketing metric-brand equity with firms' sustainable value.
Third, this study further contributes to the link between brand equity-analysts' recommendations by examining an important industrial contingency-industrial competition.Prior studies have investigated the boundary condition of culture difference (i.e., collectivism culture versus individual) and uncertainty avoidance [8], yet few have looked the influence of industrial factors on the relationship between market-based assets and analysts forecast.Adding into this line, our findings indicate that industrial competition strengthens the positive effect of brand equity on the level of analysts' recommendations, and further reduces the negative effect of brand equity on analysts' recommendations inconsistency.Therefore, these findings emphasize a stronger financial impact of brand equity in highly-competitive industries.

Managerial Implication
The findings of this study have several managerial implications.First, Chief marketing offers (CMO) or marketing managers are under numerous pressures to demonstrate the financial accountability of marketing strategies [4,5,46].Our findings indicate that brand equity generates financial returns for firms and also reduces firms' performance risks.Hence, firm expenses devoted to developing brand equity are strategic investments that meet the requirements of financial accountability.Firms should be confident on allocating resources to invest on brand equity and also strengthen the communication (e.g., information exchange and flow) between marketing and financial departments to further help evaluate the overall outcomes of branding activities.
Second, managers should be aware of the important role of analysts in explaining the benefits of brand equity on firm sustainable performance.Brand equity is a complex phenomenon for most public investors, who rely heavily on analyst recommendations to make stock investment decisions.In view of this, in firms with strong brand equity, managers are encouraged to disclose sufficient information about brand equity to financial analysts, especially in competitive industries.Meanwhile, managers should proactively interact with financial analysts to help them include and appropriately interpret the importance of brand equity in the stock evaluation and forecast processes.Such behaviors can support analysts to draw adequate recommendations based on brand equity, thereby enhancing the impact of brand equity on firm sustainable performance.

Limitation and Future Studies
Our study had some limitations that indicate the potential for future research.First, our findings were limited by the sample of firms listed only in the Shanghai/Shenzhen Stock Exchange in WBL.There were a significant proportion of non-listed firms and also a small number of firms listed in foreign countries (i.e., U.S.A., Singapore) in the WBL.Considering the enormous differences between listed and non-listed firms and the emerging market (i.e., China) and the developed market (i.e., U.S. & Singapore), conclusions in our study might not be able to be generalized into other contexts.Second, the archival data we used to measure brand equity is not able to capture its sub-dimensions.Future research is encouraged to further explore the relative impacts of brand equity's sub-dimensions on analysts' forecasts and firm value.Third, we employed signaling theory as the theoretical base in our research and specifically focus on the role of analysts.More research will be helpful in exploring other important market players (e.g., institutional investors) [75] who may have similar influences like analysts.
Fourth, because of the importance role of corporate governance in affecting firm performance and analysts' recommendations, future research can extend our study by controlling more governance-related factors, for example, CEO tournament [58], levels of CEO equity incentives [59], variation between CEOs' holdings of equity incentives [59] and mutual monitoring among the executives [60].Previous research indicates industrial tournament incentives for CEOs significantly affect firm performance and risk [58].Core and Guary (1999) found that firm performance can be governed using compensation incentives: setting optimal levels of CEO equity incentive and actively managing levels among CEOs' holding equity by varying their incentive grants [59].Li (2014) found that mutual monitoring among the executives is positively related to future firm performance, and incentivizing executives to monitor the CEO increase the effectiveness of mutual monitoring [60].Mutual monitoring among the executives is measured by the compensation gap between CEO and that of the second-highest-paid executive scaled by CEO total compensation.As such, we encourage future studies to explore these salient governance factors to further extend our conclusions.
Fifth, there are other important variables that may affect the influence of recommendations from different analysts, such as brokerage firm size (represent for available resources) [53], talent, status (e.g., star analyst) [76], etc. Dang et al. [77] argued that the analyst's talent is a more important factor than the other variables studied in previous research (i.e., experience, brokerage affiliation and task complexity), and the analysts' high innate ability can effectively reduce the information asymmetry between corporate insiders and outside investors [78].Recommendations of star analysts may be more influential for public investors than non-star analysts [76].Future studies can explore different characteristics of analysts and their relative effects of recommendations from which the benefits of firm assets transfer into sustainable performance.
Finally, although our study empirically supports the connection between brand equity (in year t − 1) and analysts' recommendations (in year t), secondary data failed to capture whether every analyst actually factored brand equity into their assessment system to establish recommendations.Future research could conduct deep interviews or survey research with analysts to deeply explore how brand equity affects their stock recommendations.Finally, we lacked data to capture firms' social performance, which may have influenced both brand equity and firms' sustainable performance [79].Future research can improve our research by including the social performance variables, such as the corporate social responsibility score or philanthropy donation amount.

Table 2 .
Correlation and descriptive statistics.

Table 3 .
The effects of brand equity on analysts' recommendations level and inconsistency.

Table 4 .
The mediation effect of analysts' recommendation level and inconsistency.