key: cord-0699184-0uk718sl
authors: Edeling, Alexander; Srinivasan, Shuba; Hanssens, Dominique M.
title: The marketing–finance interface: A new integrative review of metrics, methods, and findings and an agenda for future research
date: 2020-09-19
journal: nan
DOI: 10.1016/j.ijresmar.2020.09.005
sha: 0f2baa06dbceded9d6a689782f6b01523fd1b3e0
doc_id: 699184
cord_uid: 0uk718sl

The marketing–finance interface is an important research field in marketing, helping demonstrate the accountability of marketing within companies and building a necessary interdisciplinary bridge to finance and accounting research. Since the first comprehensive review article by Srinivasan and Hanssens (2009), the marketing–finance field has broadened considerably, as has research in finance and accounting. This updated systematic review of extant and new research integrates research in marketing, finance, and accounting into an overarching marketing–finance research framework. We discuss new methodological developments and offer solutions to recent technical debates on the event-study method and Tobin’s q. Motivated in part by a survey of marketing–finance researchers, the article identifies and synthesizes four key emerging research areas: digital marketing and firm value, tradeoffs between “doing good” and “doing well,” the mechanisms of firm-value effects, and feedback effects. The article closes with a future research agenda for this dynamic research field and offers key conclusions.

We ask several key research questions. First, how can the metrics that have emerged since 2009 be categorized into a conceptual marketing-finance framework? Second, which dataanalytic advances, such as machine-learning-supported textual analysis, have had a major impact on the field since 2009? How can the methodological debates since 2009 be resolved? Importantly, which generalizable results can be drawn from the countless empirical studies? We organize these generalizations along the following four themes, motivated in part by a survey we conducted among researchers from marketing, finance, and accounting: (1) digital marketing and firm value, (2) tradeoffs between "doing good" and "doing well," (3) the mechanisms of firm-value effects, and (4) feedback effects. Finally, what directions should future research take against the backdrop of a business environment that is moving away from a pure shareholder-focused approach?

In addressing these questions, we offer several contributions. For researchers, we provide an overview of metrics, methods, and findings and an agenda for future research. Our review shows that finance and accounting researchers tackle marketing-related topics, but do so using different approaches in data collection and analysis. We explore avenues for enhancing marketing's position among the business disciplines. For marketing managers, we provide insights into the strongest drivers of firm value. Our review also sheds light on the potential of marketing to reconcile the objectives of multiple stakeholders (customers, shareholders, employees, and communities) . For the investor community (analysts and investors), we provide insights into how to incorporate information from various marketing actions/signals in their investment decisions and studies that cite these articles; (3) conducted a keyword search on Google Scholar (e.g., "marketing firm value," "marketing stock return", "marketing stock risk"); and (4) applied a journal-byjournal search from 2009, the publication year of SH, to April 30, 2020.

The search led to the identification of 286 empirical articles, 227 (or 79.4%) of which were published in or after 2009 (Web Appendix B provides a reference list of all included studies). (3) the number of studies dealing with marketing-finance topics outside the marketing discipline is considerable, with 59 studies (or 20.6%) in total-among those, finance has the largest share (28 articles), followed by accounting and management/strategy (15 each). Thus, while marketing-finance research has been growing rapidly in the marketing discipline, it has also spread (or developed in parallel) to related disciplines, in particular the foundational field of finance, where the focus has been on innovation, advertising, digital metrics, and, particularly, corporate social responsibility (CSR).

The distribution of articles by investigated industry shows that 73.1% are from a mix of industries, 5.9% from pharmaceuticals, 4.9% from high-tech, 4.2% from consumer durables, 2.4% from automobiles, and the remaining 9.1% from other single industries (.3% of articles do not report the industry). Thus, a majority of articles use samples obtained across industry sectors, which enhances the generalizability of the findings. Regional distribution is skewed, with 89.2% of studies conducted in the US. Thus, despite research showing that global equity markets are integrated (Park, 2004) , generalizations to other international markets are lacking, offering an op-J o u r n a l P r e -p r o o f portunity for further research.

To add to the insights from our systematic review of previous work, we conducted two online surveys, one among researchers who have worked on marketing-finance topics and one among those who have published event studies and/or Tobin's q-related work in major finance and accounting journals. The surveys took place from August to October 2019 and included questions about the most important past and future marketing-finance topics (survey 1), the relevance of marketing topics for finance/accounting researchers (survey 2), the influence of digitization on the research fields (both surveys), respondents' own use of different research methods (survey 1), and recent methodological controversies within the marketing-finance research field with respect to event studies and the appropriateness of Tobin's q (both surveys). For surveys 1 and 2, we obtained usable answers from 66 and 46 respondents, respectively, with 51 and 26 individuals completing the full questionnaire.

The value of our study lies in its breadth of coverage of marketing variables and methodological approaches. Regarding the marketing variables, we evaluate the full set of variables suggested by SH (marketing assets and actions) and Moorman and Day (2016) (marketing organization, including capabilities, human capital, configuration, and culture). Regarding methodological approaches, we incorporate models with other financial outcome variables (e.g., analyst/investor behavior) and feedback models from the stock market to marketing decision making. In contrast with the studies of Conchar, Crask, and Zinkhan (2005) , Edeling and Fischer (2016) , and Sorescu et al. (2017) , which focus specifically on certain variables and/or methods, in the spirit of SH, our study presents a comprehensive state-of-the-art picture of the marketing-finance research field (see Web Appendix C for a comparison of review studies).

Fig. 2 depicts the framework used to categorize hitherto investigated metrics at the marketing-finance interface. Using the frameworks of SH, Edeling and Fischer (2016) , and Katsikeas et al. (2016) , we begin from a marketing-finance value chain that links marketing actions through marketing assets, market, and accounting performance with behavior of the investor community, culminating in financial-market performance metrics. Figure 2 about here == We combine this operational marketing-finance value chain with the view of Moorman and Day (2016) , who regard "marketing organization" as fundamental to the achievement of marketing excellence. Marketing organization is the strategic foundation for the functioning of the conversion of marketing actions into firm value along the marketing-finance value chain. It consists of four dimensions: capabilities ("complex bundles of firm-level skills and knowledge and firm adaptation to marketplace changes"), configuration ("organizational structures, metrics, and incentives/control systems that shape marketing activities"), human capital ("marketing leaders and employees … [who] create, implement, and evaluate a firm's strategy"), and culture ("values, norms, and behaviors that facilitate a focus on the market over time") (Moorman & Day, 2016, p. 6) .

For each of the eight focus categories, Fig. 2 shows the intensity with which marketingfinance researchers have investigated relationships between marketing variables and financialmarket metrics. Studies have considered both directions, from marketing to finance variables and, to a lesser degree, feedback effects from the financial market to marketing organization, actions, and assets. The numbers in parentheses show the number of studies that have dealt with a category or an exemplary variable. As one study can include more than one marketing or financialmarket variable, the sum of studies across marketing and finance categories exceeds the number J o u r n a l P r e -p r o o f Journal Pre-proof of empirical studies included in the review (286). The numbers show that the majority of studies have investigated either marketing action (216) or marketing asset relationships (126), while marketing organization is a considerably under-explored topic. In organizational studies, researchers have investigated configurational themes (41) the most, followed by human capital (20) and capabilities (14), but culture only sporadically (4).

The emphasis on classic marketing action and asset topics is also mirrored in the free-text answers to a survey question on the most important marketing-finance interface topics in the past (see Fig. 1 , panel B). The only organizational topic that appears on the list of the most-oftenmentioned themes is chief marketing officer (CMO)/top management team (4 mentions). Overall, the top marketing variables that have been investigated by at least 10 articles are advertising expenditures (49), customer satisfaction (33), new product introduction (27), CSR activities (22), R&D expenditures (18), alliances (17), and customer-based brand equity (17). On the financialmarket side, studies with financial-market performance (391) strongly outnumber studies focusing on the behavior of financial-market participants (74). Of the eight most-often-investigated metrics with double-digit studies, six belong to the former group including stock returns (195), Tobin's q/market-to-book ratio (71), idiosyncratic risk (29), systematic risk (25), market capitalization (13), and cash flow (13), while only two metrics are part of the latter category with trading volume (11) and institutional stock ownership (10).

Using a two-color scheme, Fig. 2 also shows recent metric developments within the marketing-finance interface field. All sub-category variables investigated at the time of SH (2009) are depicted in black ("old"). Conversely, sub-category variables analyzed since 2009 are in red ("new"). In each sub-category, the first mentioned variable is the most frequently used across all years, while the second is a new variable since 2009, either with the highest or second-highest frequency if the most frequently used happens to be a new variable. Three insights emerge. First, J o u r n a l P r e -p r o o f only one new category has surfaced since 2009-the organizational culture dimension. Second, several sub-categories have emerged in the past 10 years, including customer management in the "actions" category, general offline and online buzz in the "assets" category, analysts in the "behavior" category, and metrics for young firms and debt-related metrics in the "financial-market performance" category. Third, two sub-categories have not seen any development of new metrics since 2009-observable customer behavior within marketing assets and equity-and debt-related metrics within financial-market performance. Overall, the majority of metrics have been introduced since the publication of SH (see Web Appendix D for a detailed list of metrics, including their categorization as "old" or "new" and number of occurrences). In addition, 53.0% of variables (55.3% of marketing variables, 46.7% of financial variables) have appeared in only one study. Fig. 2 includes only information on focal-firm effects, that is, relationships between a firm's marketing variables and its own financial-market outcomes. However, several attempts have analyzed competitor effects in areas such as innovation, promotion, and advertising ; customer data breaches (Martin, Borah, & Palmatier, 2017) ; online negative chatter about product recalls (Borah & Tellis, 2016) ; and celebrity endorsements (Knittel & Stango, differ considerably in terms of several important criteria (see Panel A of Table 1 ). 3 First, finance/accounting studies mainly focus on interesting empirical phenomena using unique databases, such as daily advertising data (Focke, Ruenzi, & Ungeheuer, 2020) and Amazon.com product reviews (Huang, 2018) . The theory section in marketing articles is often lengthier, and a higher percentage of studies are explicitly based on finance theories such as stakeholder theory (Wies et al., 2019) , psychology-based concepts such as associative network theory (Borah & Tellis, 2016) , and interdisciplinary theories such as news value theory (Stäbler & Fischer, 2020) . Table 1 about here == Second, key differences emerge in the direction of investigation and the foci regarding variables. The vast majority of marketing articles investigate the traditional direction from marketing to financial-market metrics, thus contributing to the goal of "nailing down marketing's impact" (Hanssens, Rust, & Srivastava, 2009, p. 115) . Conversely, a much higher percentage of studies in finance/accounting ask how financial variables affect firms' marketing decision making. We further observe asymmetry in metric foci: marketing studies are broad on the marketing side but narrow on the financial-market side (with a focus on the "big 4" firm-value metrics stock return, Tobin's q, and idiosyncratic and systematic risk). Finance/accounting studies have less breadth on the marketing side (focus on innovation, advertising, online metrics, and, particularly, CSR) but higher variety on the finance side (an especially strong focus on investor behavior metrics such as trading volume). Finance/accounting articles also investigate more relationships per study, relying on multi-method designs, a trend that has only recently been adopted by marketing researchers for robustness and breadth of insights (Moorman et al., 2019) .

Finally, we observe substantive differences in the way marketing and finance/accounting articles establish causality in econometric studies (Angrist & Pischke, 2009 ). Other than event studies constructed as quasi experiments (see subsequent discussion), marketing researchers have relied mainly on panel-data and instrumental-variable approaches to account for endogeneity (Germann, Ebbes, & Grewal, 2015) or explicitly modeled the temporal causal nature or structure of relationships using vector autoregressive (VAR) modeling (Colicev et al., 2018) . Identification is a paramount issue for finance/accounting scholars. Thus, they often rely on exogenous shocks that act as quasi-natural experiments (e.g., an Indian law that makes CSR expenditures mandatory for firms above a certain size; Manchiraju & Rajgopal, 2017) . The methods to analyze such data are either regression discontinuity (Manchiraju & Rajgopal, 2017) or difference-in-differences (He & Tian, 2013) . Accounting researchers have also begun using field experiments to evaluate the effect of news articles on liquidity and stock returns (Lawrence et al., 2018) . In general, marketing researchers can benefit from monitoring the ongoing discussion in finance and accounting about the potential to show causal effects (e.g., Gow, Larcker, & Reiss, 2016) .

SH propose four research methodologies, based on the Fama-French model, that previous work has used with different frequencies (see Fig. 1 , panel C): short-term (91, 19.8%) and longterm (11, 2.4%) event studies, stock return response models (75, 16.3%), calendar time portfolio models (30, 6.5%), and persistence (VAR) models (16, 3.5%). These approaches generally rely on the efficient market hypothesis in finance, which states that investors fully and immediately incorporate any new information that has value relevance. In addition to these models, marketing-finance researchers have applied firm-value-level models for variables such as Tobin's q ) (80, 17.4%), risk/volatility models (Han, Mittal, & Zhang, 2017) (44, 9.6%) , models for other financial outcome variables such as credit ratings or trading J o u r n a l P r e -p r o o f volume (Anderson & Mansi, 2009; Focke et al., 2020) (54, 11 .8%), feedback models that incorporate the reverse effect from stock-market performance to marketing actions (Park, Chintagunta, & Sun, 2019 ) (46, 10.0%), marketing-based valuation models (McCarthy & Fader, 2018) (7, 1.5%), and structural equation models (Zuo, Fisher, & Yang, 2019) (5, 1.1%). 4 In what follows, we discuss several of these methods, including some recent developments in marketing and finance/accounting research (see Panel B of Table 1 ).

The Fama-French four-factor model, a foundational model, recognizes systematic sources of cross-sectional differences among firms' stock returns: the size factor, the market-to-book value factor, the market risk factor (the original three factors by Fama and French [1993] ), and the momentum factor (added by Carhart [1997] ). Four-factor models have been used to compute abnormal returns and to calculate systematic and idiosyncratic risk, which then serve as input in different firm-value models (Hsu, Fournier, & Srinivasan, 2016) . Recently, Fornell, Morgeson, and Hult (2016) estimated the stock returns to customer satisfaction using a new five-factor model suggested by Fama and French (2015) . This model augments their original three factors with a profitability factor (difference in returns between diversified portfolios of stocks with strong versus weak profitability) and an investment factor (difference in returns between diversified portfolios of stocks of low and high investment firms). In the past five years, the discussion about factor models in the finance literature has accelerated. In particular, Hou, Xue, and Zhang (2015) develop their own factor pricing model that also relies on the market, size, profitability, and investment factor. Recently, Daniel, Hirshleifer, and Sun (2020) introduced a novel behavioral fac-tor model based on investor psychology. Researchers have also turned to machine-learning approaches to identify relevant asset pricing factors and have shown that these methods outperform traditional four-factor models in return prediction (Gu, Kelly, & Xiu, 2020) . These rapid developments (of which we can only provide an overview) are accompanied by increasing skepticism toward attempts to find significant drivers of abnormal stock returns, that is, pricing anomalies (e.g., Harvey, Liu, & Zhu, 2016) . We advise marketing researchers to closely follow these developments (see, e.g., the 2020 special issue on "New Methods in the Cross-Section" in Review of Financial Studies).

Event-study approaches have had a solid trajectory with increased use over time, likely because they allow for an inference of causality in quasi-experimental settings. Event studies are used to measure short-or long-term value relevance of a discrete event (e.g., .

The intuition behind event-study methodology is that, given market efficiency, perfect information, and rational investors (Fama, 1991) , the effect of a relevant event should be immediately reflected in stock prices. Both marketing and finance researchers have added to the development of this method during the past years. In finance, research has introduced the so-called event-study regression, an alternative to the standard analysis of cumulative abnormal returns (e.g., Beber & Pagano, 2013; Boehmer, Jones, & Zhang, 2013) . Hock and Raithel (2020) recently applied this method to investigate celebrity endorsement scandals. The approach can be superior if the researcher is interested in the effect of direct firm reactions to certain marketing-related events.

A technical issue that often arises in event studies is cross-sectional correlation among abnormal returns when the event days are clustered for sample firms, leading to over-rejection of the null hypothesis of no mean event effect (SH). The calendar time portfolio approach automatically accounts for this problem, but with the limitations that individual-firm abnormal returns J o u r n a l P r e -p r o o f cannot be computed and the test power is generally low (Sorescu et al., 2017) . Kolari and Pynnönen (2010) suggest a new t-statistic that overcomes the misleading-inference problem of crosssectional correlation in single-and multiple-event windows. Feng, Morgan, and Rego (2020) recently applied this t-statistic in their study of unprofitable customer management strategies.

Another development in the finance literature that was transferred to the marketing-finance interface is the application of the event-study approach to stock price risk instead of returns (Carlson, Fischer, & Giammarino, 2010) . Thomaz and Swaminathan (2015) use this approach to analyze the impact of marketing alliances on firm risk.

There have also been recent event-study-related advancements in marketing research. Standard practice (60.4% of studies in our sample) is to eliminate confounded events (e.g., due to earnings announcements) in short-term event studies. Sorescu et al. (2017) argue in favor of retaining such events in the sample to increase statistical power and to decrease the subjectivity in the choice of confounding announcements. They provide evidence that retaining versus deleting confounded events does not significantly alter results. Park et al.'s (2019) approach contributes to the question of what proportion of abnormal returns are really unpredicted. They decompose abnormal returns to drug approval into a proportion predicted by pharmaceutical managers based on their pre-approval market research and a proportion not predicted by managers before approval (i.e., new information).

A fundamental decision when conducting event studies is how the abnormal return (AR) or cumulative abnormal return (CAR), when using event windows of more than a day, should be calculated. Skiera et al. (2017) develop an innovative and simple solution to the phenomenon that most marketing events are likely to affect only a firm's operating business (OB) and not the other two components of shareholder value (SHB): non-operating assets (NOA, e.g., excess cash) and debt (DEBT). They derive mathematically that if the assumption of a sole effect on OB holds, the J o u r n a l P r e -p r o o f CAR on the operating business (CAR OB ) is equal to the standard CAR SHV divided by a firmspecific "leverage effect" OB/(OB -NOA + DEBT), which describes the relative change in SHV for a 1% change in operating business. Applying this simple formula to three previously published event studies, the researchers show that CAR OB results can differ fundamentally from the standard CAR SHV results, including even a change in sign.

Our survey results on the question whether the implications from Skiera et al.'s (2017) study are likely to be adopted by marketing-finance and finance/accounting researchers are noteworthy.

While the average score for the first group is 3.6 (SD = 1.1, n = 17), the second group is more skeptical (M = 2.6, SD = 1.5, n = 10). Skiera et al. themselves argue that "events typically explored in … finance, such as regulatory changes, natural disasters, and mergers and acquisitions, are likely to influence both the value of the operating business and non-operating assets and debt" (p. 644). Our suggestion to researchers who want to conduct marketing event studies resembles Skiera et al.'s call to elaborate more on which parts of SHV are influenced by marketing events.

First evidence (albeit not in an event-study setting) shows that marketing assets such as customer satisfaction and brand equity can have a significant effect on debt-related metrics such as credit ratings (Anderson & Mansi, 2009; Himme & Fischer, 2014) and cash holdings (Bharadwaj, Hanssens & Rao, 2020; Larkin, 2013) , so a general statement is not possible. Given that the components of the leverage effect are publicly available, calculating CAR OB , comparing it with CAR SHV , and interpreting potential differences should be standard practice in marketing-finance research (for a recent application, see Lim, Tuli, & Dekimpe, 2018) .

Stock return response models typically measure incremental value relevance of continuous marketing metrics that are not fully reflected in contemporaneous accounting performance. These models are used to establish whether investors perceive information on changes in marketing ac-J o u r n a l P r e -p r o o f tivity, such as advertising spending, as contributing to a change in the projection of future cash flows. Importantly, these models are based on the efficient market hypothesis and recognize that investors react only to new information, which is operationalized as the difference between the actual and expected level of the independent variable (e.g., Edeling & Fischer, 2016; SH) . Yet empirical marketing-finance researchers need to consider the distinction between unexpected changes and levels of marketing actions, which many still ignore (e.g., Lariviere et al., 2016) . Recently, Mizik (2014) suggested a method that builds on stock return response modeling that enables the analysis of the total long-term financial consequences of marketing assets and their decomposition into immediate and future effects.

Calendar time portfolios measure mispricing, or "the extent to which the financial markets fail to react to information that has long-term profit implications or overreact to information that does not have long-term profit implications" (Jacobson & Mizik, 2009, p. 837 ). Research has extended this approach to incorporate the Fama-French five-factor model (see Fornell et al., 2016) .

Persistence models involving time-series methods are well suited to analyze stock price data and their sensitivity to new marketing information (e.g., Colicev et al., 2018) . They are flexible to accommodate dynamics, feedback loops from investors to managers, and deviations from the efficient market hypothesis. In addition, they can flexibly incorporate risk and other performance variables. In the past few years, marketing-finance research has applied panel-VAR models, which exploit cross-sectional variation when long time series are not available. Kang et al. (2016) use annual data on more than 4,500 firms across 19 years (i.e., a large cross-section and short time series) to model the interplay among CSR, corporate social irresponsibility (CSI), and firm value using a structural panel-VAR model that allows contemporaneous effects among some of the endogenous variables. Recently, Huang and Trusov (2020) introduced interactions in a panel-J o u r n a l P r e -p r o o f VAR model to investigate how the interrelationship between firm financial performance and executive compensation varies with productivity and customer satisfaction levels. Overall, persistence models continue to remain in the methods toolkit for marketing-finance researchers, especially as granular weekly or even daily data (Colicev et al., 2018 ) become more available.

Firm-value-level models include all models that connect a firm-value dependent variable in levels (i.e., cash flow, market capitalization, Tobin's q, and market-to-book ratio) with an independent marketing variable. The frequent and growing use of level models is surprising given criticism that autocorrelation leads to downward-biased standard errors and false inferences (Edeling & Fischer, 2016; ). The most-often used metric, Tobin's q, is under special scrutiny. In a thought-provoking study, Bendle and Butt (2018) question the validity of marketing-finance studies that use accounting-based approximations of Tobin's q (AATQ) as the dependent firm-value variable. The core criticism is that market-based assets such as selfcreated customer satisfaction or brand equity go unrecorded in firms' accounting reports, leading to a biased measure of a firm's replacement value in the denominator of the AATQ formula. The authors formally show that performance-neutral marketing decisions such as converting advertising expenditures (cash) into brand equity (unrecorded asset) increase AATQ, which results in an over-estimation of the value relevance of marketing.

Considering that Tobin's q (and the related market-to-book ratio) is the second most often used firm-value variable (after stock return, see section 3) in our review, we offer several observations and potential remedies to the discussion. First, the criticism of AATQ is not new. Mizik and Jacobson (2009) , in their commentary to SH, already mention the measurement error in the denominator of AATQ as a critical flaw. Edeling and Fischer's (2016) meta-analysis shows that elasticities derived from "intangibles-to-tangibles" models are substantially larger than the elas-J o u r n a l P r e -p r o o f ticities derived from stock return models. Second, the criticism is not limited to the marketing field. Finance researchers are frank about the fact that "proxies of Tobin's q are imperfect measures of firm value" (Gurun & Butler, 2012, p. 586) . Third, despite these limitations, and similar to researchers in marketing, finance and accounting researchers continue to use Tobin's q in their models. Use of the keyword "Tobin's q" in Google Scholar led to 266 identified articles in the top three finance journals and 81 articles in the top three accounting journals (both according to the UT Dallas list) since 2010. However, the majority of these studies use Tobin's q as a regressor that measures firms' future growth or investment opportunities rather than as the dependent variable. Not surprisingly, respondents in our finance/accounting survey concur that stock returns should be preferred to Tobin's q as a performance metric (M = 4.22, SD = .67; scale from 1 [completely disagree] to 5 [completely agree]). The only slightly lower score for marketing-finance researchers (M = 3.76, SD = 1.09) for this question shows that the community is largely aware of the shortcomings of Tobin's q. Finally, similar to the event-study-related suggestions, we urge researchers to justify the choice of firm-value metrics (previous use in a top marketing journal is not a justification) and to use more than one metric for robustness. One of these metrics could actually be a variant of the traditional AATQ, a measure called "Total q" developed by Peters and Taylor (2017) in the finance field. This easily computable measure accounts for intangible capital in the denominator as the sum of so-called knowledge capital (based on R&D expenditures) and organization capital (based on selling, general, and administrative expenditures). The metric, applied recently in the marketing literature by Du and Osmonbekov (2019) and popular in the finance literature (301 citations on Google Scholar as of May 1, 2020), is closer to the true Tobin's q measure than AATQ and thus could overcome many of the problems Bendle and Butt (2018) discuss.

J o u r n a l P r e -p r o o f At a more strategic level, emerging from the marketing-finance literature is a new approach to firm valuation, marketing-based valuation, that monetizes the expected value of a firm's customer relationships as a proxy for its future financial outlook. This is achieved by deriving the customer equity of the firm, or the sum of expected net revenues from current customers and future customer acquisitions. 5 Gupta, Lehmann, and Stuart (2004) 

In addition to the developments in data analysis methods, important progress has been made with respect to data collection methods in the past 10 years. The most important digitizationtriggered novelty is that marketing-finance researchers augment structured numeric marketing data (e.g., advertising expenditures, customer satisfaction ratings) with unstructured textual data 5 The tangible parts of firm value, such as plants and equipment, should be included separately.

J o u r n a l P r e -p r o o f (Berger et al., 2020) . Table 2 contains previous applications of automated textual analysis 6 in the domain and categorizes them according to the data used and the extracted text information. While a majority of applications use consumer-generated social media data (e.g., Bartov, Faurel, & Mohanram, 2018; Borah & Tellis, 2016) , we observe a recent trend toward the study of firmgenerated text data from 10-K reports (e.g., Frennea, Han, & Mittal, 2019) and press announcements (e.g., Dotzel & Shankar, 2019). Other used sources are newspapers (Solomon, 2012; Xiong & Bharadwaj, 2013) and books. Sorescu et al. (2018) use the publicly available Google Books Ngram Viewer tool to identify the historical usage of certain words in published books. Table 2 about here == These studies extract different types of text information from the given data, which can broadly be divided into simple volume of documents (e.g., reviews as in Tirunillai and Tellis [2012] ) or words (Sorescu et al., 2018 ) and more complex classification outcomes. We distinguish among sentiment classification (distinguishing between positive and negative sentiment), content classification (assigning pre-defined category labels), and topic modeling (exploratively identifying general topics; see Berger et al., 2020; Hartmann et al., 2019) . Marketing-finance research has predominantly dealt with sentiment analysis and content classification, while using topic modeling only sporadically. While topic modeling relies on unsupervised machine-learning techniques such as Latent Dirichlet allocation, sentiment and content classification can be executed using either lexicon-based methods or supervised machine-learning approaches (Hartmann et al., 2019) , with applications in marketing favoring the latter. However, researchers in accounting and finance are less enthusiastic about the often-complex machine-learning algorithms because of 6 Automated textual analysis is distinct from manual textual analysis that is also applied at the marketing-finance interface (e.g., Bayer, Tuli, & Skiera, 2017) . Manual coding reaches its limits with extensive text data, as is customary in marketing and finance.

J o u r n a l P r e -p r o o f their potential lack of transparency (Loughran & McDonald, 2016) .

Given the multitude of empirical findings at the marketing-finance interface and in the interest of space, we present an overview of findings in this section and selected findings based on their importance for the field (section 5.2). Table 3 summarizes the findings for the marketing  financial-market effects for the 10 most-often analyzed marketing and finance variables. We focus on this classic direction because feedback effects have been investigated much less often (see section 5.2.4). Note that the entries (1) are based on regression coefficients; (2) contain main effects only; and (3) distinguish among significantly positive (+), non-significant (0), significantly negative (-), and non-studied relationships (gray fields). A study can contribute more than one effect if findings are inconsistent across methods (e.g., Fornell et al., 2006) or samples (e.g., Balasubramanian, Mathur, & Thakur, 2005) . We find that though these are the most-often studied relationships, 78 of the 140 cells are still gray, indicating a large potential for future research. Because stock return is the only finance outcome variable that is covered for each of the marketing variables, we rely on stock-return effects for our generalizable conclusions. 7 We identify four groups of marketing variables to organize the findings overview. The first group comprises variables with only positive or neutral stock-return effects-that is, the marketing action variable new-product introductions as well as the marketing asset variables customer satisfaction, customer-based brand equity, product quality, financial brand equity, earned social media volume, and positive sentiment. The second group includes advertising expenditures and 7 "Real" empirical generalizations are derived from quantitative meta-analyses (Hanssens, 2018) , which, unlike our qualitative systematic review, allows for a mean effect size as in Edeling and Fischer (2016) . Table 3 about here ==

We select four major topic areas based on our judgment of dominant issues in the field and a synthesis of the answers to our survey questions on important extant and future topics (see Fig. 1 , panels B and D): digital marketing and firm value (section 5.2.1); tradeoffs between doing good and doing well (5.2.2); mechanisms of firm-value effects, for which we cover selected moderation and mediation findings (5.2.3); and feedback effects from the financial market to marketing decisions (5.2.4). For each topic area, we offer several synthesizing statements based on emerging empirical patterns. Web Appendix F includes a list of selected studies published since 2009 within each area, including a summary of their main findings.

Despite the dominant role of digital marketing in marketing research (Lamberton & Stephen, 2016) , surprisingly few studies have examined the relationship between digital marketing and firm value. A quote from our survey offers a potential explanation: "Marketing-finance topics tend to be more strategic whereas research on digitization from firm perspective has been mostly tactical." Regarding firms' online communication actions, the limited published evidence suggests that the firm-value impact of online advertising lies between the effect of offline national and regional advertising, but negative interaction effects of the three media types hint at weak communication integration or a ceiling effect of the impact of advertising in general (Sridhar et al., 2016) . Bayer et al. (2020) show that paid search advertising has a more positive effect on sales than offline advertising, consistent with paid search being closest to the actual purchase decision and having enhanced targeting abilities. They find that display advertising has a relatively more positive effect on Tobin's q than offline advertising, consistent with its long-term effects.

In the social media sphere, firms' owned social media has both direct and indirect (via its ef- . However, social media is a stronger predictor of stock returns and stock risk than more traditional online buzz metrics, such as online search and web traffic (Luo, Zhang, & Duan, 2013) . Twitter tweets and Amazon product reviews are especially important predictors of abnormal returns (Bartov et al., 2018; Huang, 2018) .

Owned social media is a driver of firm value, with potential asymmetries for positive and negative sentiment and likely spillovers on rivals.

Data privacy is an evolving field in digital marketing research (Martin & Murphy, 2017) .

Customer data breaches appear to harm the stock return of both the affected firms and their competitors (Kashmiri, Nicol, & Hsu, 2017; . Firms can mitigate the negative consequences by giving customer control of their data via opt-out options and by investing in stronger marketing capability and IT know-how in the top management team (Kashmiri et al., 2017) .

Finding 3. Data breaches can have severe negative firm-value effects on focal firms and, to a lesser degree, rival firms.

Marketing-finance research has extensively examined whether creating value for shareholders versus for three other important stakeholder groups (customers, employees, and society as a whole) is actually a tradeoff or whether mutual benefits are possible. The positive empirical evidence for marketing's main stakeholder group, customers, and their satisfaction is overwhelming (see section 5.1), based not only on the widely used ACSI database but also on a customer satisfaction measure from Interbrand (Colicev et al., 2018) and Amazon customer reviews (Huang, 2018) . However, the routes by which customer satisfaction increases firm value, via loyalty intention (another mindset metric; Larivière et al., 2016) , earnings surprises (Fornell et al. 2016), or analyst recommendations (Luo, Homburg, & Wieseke, 2010) , need further investigation, as do industry-specific heterogeneity (Larivière et al., 2016) and interactions with other mindset met-J o u r n a l P r e -p r o o f rics (Himme & Fischer, 2014) Finding 4. In general, positive changes for the customer stakeholder group in terms of higher customer satisfaction are associated with positive shareholder effects.

The yearly abnormal return to the intangible asset employee satisfaction (measured via Fortune's "100 Best Companies to Work for in America" ranking) is 3.5% (Edmans, 2011) , and thus only approximately one-third of the yearly abnormal return to customer satisfaction of 10.8% (Fornell et al., 2016) . Green et al.'s (2019) recent study finds that firms with improving online employee reviews on Glassdoor.com also outperform firms that experience declines in these crowdsourced ratings. Firms' activities to improve their human capital must be considered in parallel to their customer-focused marketing initiatives. Vomberg, Homburg, and Bornemann (2015) find significant human-capital effects on Tobin's q, cash flow, and cash flow volatility only when interacted with the customer-based brand equity of a firm. Similarly, Groening, Mittal, and Zhang (2016) show that positive (negative) actions toward employees are an especially positive (negative) signal for investors when they co-occur with positive (negative) customer-related achievements. In addition, online job postings by firms, especially those that signal hiring for growth, have positive effects on stock prices (Gutierrez et al., 2020) .

Finding 5. Preliminary evidence suggests that employee satisfaction has a positive effect on firm value and a positive interaction with a firm's brand and customer activities.

The stakeholder group of society/communities, which is often the target of CSR activities, has the most questionable mutual benefit with shareholders (see also section 5.1). For example, the announcement of cause-related marketing initiatives can even decrease shareholder wealth (Woodroof et al., 2019) . Similarly, the introduction of an Indian law that forces firms that fulfill certain criteria to spend at least 2% of their net income on CSR has led to a significant firm-value loss for affected firms (Manchiraju & Rajgopal, 2017) . Again, marketing plays an important role in mitigating the negative consequences through better corporate reputation (Woodroof et al., J o u r n a l P r e -p r o o f 2019) or by helping firms reap positive firm-value benefits through higher advertising expenditures and a better prior reputation (Servaes & Tamayo, 2013) . Kang et al.'s (2016) study on the interplay of CSR and CSI, which together can be termed "corporate social behavior," shows that firms aim to weaken the detrimental effects of CSI through a subsequent increase of CSR but are not rewarded by the stock market. Dai, Liang, and Lilian (2020) identify a key interaction effect, in which buyers' and suppliers' collaborative CSR efforts improve firm value of both firms.

Finding 6. Evidence of the shareholder-value effect of investing in CSR is highly mixed and contingent on a firm's marketing and CSI activities, as well as other firms' CSR behavior in the value chain.

A critical point raised in the marketing-finance survey is the perception that most studies use overly simplistic models and that researchers should go "beyond studies that show how marketing action/asset affects stock return" (quote from one researcher surveyed). Our review shows that this perception may not be accurate though. In addition to the methodological advances we outline, the majority of studies (66.1%) use moderation-effects models, with a significantly higher proportion since 2009 (69.6% vs. 52.5%; chi-square test, p = .048). Use of mediation models is still low (9.1%) but has grown considerably since 2009 (11.0% vs. 1.7%; p = .085).

The number of interaction effects investigated at the marketing-finance interface is vast. In particular, each event study that uses a regression analysis to identify drivers of abnormal returns is essentially an interaction analysis with many moderating variables. Standard regression models include various interaction effects as well, as Nath and Bharadwaj's (2020) recent study on the contingency effects of the CMO presence-firm value relationship shows. Thus, full coverage of all moderation effects would go beyond the scope of this article. The situation for mediation findings is somewhat less cluttered, but again a broad discussion of effects is not feasible. Nevertheless, we include in Web Appendices G-I three tables that report moderation and mediation find-J o u r n a l P r e -p r o o f ings. Whenever possible, we select studies that deal with our two themes "digital marketing and firm value" (depicted in green) and "tradeoffs between doing good and doing well" (depicted in yellow). The tables help identify areas that are still under-researched in terms of interaction and mediation effects. Web Appendix G, which includes selected moderation studies for the marketing  financial-market direction, shows, for example, that, so far, no studies have examined the interaction between configuration and human capital. Web Appendix H, which entails all interaction studies in the opposite direction, shows even more uncovered (gray) areas in need of investigation (e.g., how the behavior of financial-market participants coupled with financial-market performance affects marketing decisions). Mediation studies are more evenly distributed across our defined marketing and financial-market categories, as the full coverage in Web Appendix I shows. However, we were unable to identify studies that include metrics related to the marketingorganization categories configuration, human capital, or culture as mediators. In addition, with one exception (Malshe & Agarwal, 2015) , all mediation studies follow the direction marketing  financial market, thus calling for future research to show a chain of effects in feedback studies.

Overall, various moderation and mediation effects have been studied at the marketing-finance interface, but opportunities to investigate other mechanisms remain.

Several recent studies have investigated investor attention and expectations as explanatory mechanisms for the shareholder-value effects of marketing metrics. Xiong and Bharadwaj (2013) show that corporate news, combined with advertising activities, can have an effect on proxies of investor attention (i.e., firm ticker Google search and trading volume). Similarly, Warren and Sorescu (2017b) investigate the role of the investor community as recipients of corporate information and find that their attention (measured by trading volume, institutional holdings, and analyst following) is driven by concurrent announcements of new product introductions with other firm news. Marketing variables also shape investor expectations of corporate announcements.

Specifically, investors react less to (i.e., are less surprised by) new product introductions if a firm has introduced many new products in the past or if the volume of past news sentiment is high (Warren & Sorescu, 2017a) . Chen, Wu, and Yang (2019) identify a similar anticipation mechanism in their study on the financial value of fintech innovations.

Finding 7. Abnormal returns to firms' marketing activities should be regarded not solely as representative of the net present value of future cash flows but also as an indicator of how salient the information is and the extent to which investors have formed previous expectations of it.

Few studies have addressed the "mechanism question" using experimental investigations of individual investor behavior. Wiles et al. (2010) use lab experiments to validate their findings regarding the stock return effect of deceptive advertising in the pharmaceutical industry. Recently, Wang et al. (2019) corroborated their study findings of a significantly higher valuation for Chinese home-name stocks in two lab experiments, using investor identification and motivation to self-enhance as mediator and moderator, respectively. Further experimental research is required to validate aggregate-level findings with individual-investor-level behavior.

One mediating/moderating party we highlight here is financial analysts, who act as information intermediaries within the marketing-firm value chain. Luo et al. (2010) show that analyst recommendation level and dispersion mediate the relationship between customer satisfaction and stock returns and systematic and idiosyncratic risk. Luo and de Jong (2012) show similar effects of other analyst-related variables (coverage, earnings forecast) on the advertising-firm value relationship. Likewise, Du and Osmonbekov (2019) show that analyst coverage plays a moderating role, in that advertising only affects firm value positively when firms are not covered by analysts, implying that analyst coverage and firm advertising are substitutes.

Finding 8. Financial analysts play an important mediating and moderating role in the marketingfirm value relationship.

The number of studies investigating reverse causality from the stock market to firm marketing decisions has grown significantly since 2009 (see section 4). We categorize these studies as research on (1) myopic management, (2) effects from financial-market metrics to marketing decisions, and (3) effects of the status of the company (public vs. private). Myopic management, or the practice of "overemphasiz[ing] strategies with immediate payoffs at the expense of strategies with superior but more distant payoffs" (Mizik, 2010, p. 594) , occurs frequently (20%) and has severe long-term negative consequences that are actually worse than the effects of accruals-based earnings manipulation (Kothari, Mizik, & Roychowdhury, 2016; Mizik, 2010) . Recent research finds that marketing myopia is especially common among firms that use share repurchases to increase short-term earnings per share and that its long-term negative consequences extend beyond the financial market to the product market in terms of a higher number or product recalls (Bendig et al., 2018) . The good news is that firms can reduce myopic management by increasing marketing's relevance within the organization, including by having a powerful marketing department and a CEO who has a marketing background (Srinivasan & Ramani, 2019) .

Finding 9. Myopic management has negative stock-market consequences, and firms should aim to introduce organizational structures to reduce its occurrence.

Extant and new research has established that managers adapt their managerial decision making in response to stock price return and volatility signals (Chakravarty & Grewal, 2011; Focke et al., 2020) . Adding to this work, Park et al. (2019) find that firms react to a higher unpredicted proportion of abnormal returns to new drug approvals by increasing the marketing budgets for these products. This managerial practice of "listening" to the stock market is rewarded in the product market in this case, as post-approval advertising sales elasticities are also higher. In a related study, Mian, Sharma, and Gul (2018) find that firms increase their advertising expenditures significantly when the general investor sentiment (measured by variables such as number of ini-J o u r n a l P r e -p r o o f tial public offerings in the market) is high. However, in contrast with Park et al., that study shows that such a practice should be avoided, as advertising effectiveness is lower when sentiment is high. Finally, a recent meta-analysis in the organizational behavior literature (Porto & Foxall, 2019) documents that, across a large sample of nearly 12,000 public firms in the US and UK, financial gains only partially feed back to subsequent marketing investments. Thus, "marketing is effective in generating financial gains, but it is not a sustainable activity, requiring managers to inject more money to do more marketing" (Porto & Foxall, 2019, pp. 138-139) .

Finding 10. Firms react to stock-market-related signals by adapting their marketing activities, with mixed consequences for their product-market performance.

An important question is how stock-market listings per se affect firms' marketing behavior.

The empirical evidence on this question is limited to firms' innovation behavior. Moorman et al. (2012) find that publicly listed firms time their innovation activities so that they appear to have more innovations from year to year than private firms. Wies and Moorman (2015) and Bernstein (2015) show that not only is the public-market effect limited to such a "ratchet strategy" but firms also change their overall innovation behavior when turning from private to public. They increase the volume and variance of innovations but sacrifice radicalness and originality.

Finding 11. Firms that are (or become) publicly listed alter their innovation behavior substantially.

Our review emphasizes the role of investors in the formulation and implementation of marketing tactics and strategies. Revisiting the research agenda in SH, we find that several questions posed have been answered since 2009 while a few remain unaddressed. One such underresearched topic pertains to biases in investor decision making. For example, when are investor reactions accurate, how do biases occur, and how can they be corrected? Furthermore, as 90% of the studies in this review are from US settings, we call for future research to extend the findings J o u r n a l P r e -p r o o f to other international markets. We next provide a rich research agenda for marketing scholars in this domain organized into substantive and methodological topics.

First, we call for marketing scholars to collaborate with finance in the area of fintech in which investment decisions incorporate insights from marketing models. Relatedly, the adoption of artificial intelligence (AI) can enhance firm value in two ways. AI adoption can increase firm value by increasing productivity through improving workflow efficiencies, and AI technologies can potentially increase the value of existing investments within the firm (Cockburn, Henderson, & Stern, 2017) . Combined, these two features of AI technologies suggest that adopting them will substantially increase firm value. A key empirical challenge in many studies on technology adoption is identifying investments in new technologies. Measures of R&D will not suffice, as these capture the firm's total investment, not just in new technologies. This calls for new measures based on textual analysis of a firm's disclosure of AI activities, which might be replicable for a large sample of publicly listed firms. Digitization presents new and ongoing research issues on online reviews, crypto currency products, consumer privacy, and data breaches, and their impact on investor behavior should be investigated.

Second, we call for research on how marketers can leverage financial decisions for more efficient and effective marketing. A focus on advertising efficiency rather than effectiveness, driven by a fiduciary responsibility to shareholders, may have led firms to over-invest in digital advertising at the expense of long-term brand building, a decision that top management teams at Adidas and Gap have begun to question (Bayer et al., 2020; Graham, 2019) . Relatedly, the question of how financial outcomes feed back to marketing decisions needs to be addressed more systematically. For example, research could address how CMOs/marketing managers view the effect of capital market feedback on marketing plans and strategies, including product innovations (e.g., J o u r n a l P r e -p r o o f Park et al., 2019) , service offerings, pricing, and pricing responses of competitors, and selection of distribution channels. A related topic is the effect of "going private" and how a delisting from the stock market affects marketing actions and assets.

Third, we call for research on the effect of economic and social transformations on marketing

decisions. An important macroeconomic trend of the past few years has been the global decrease in interest rates to historic lows, which positively affects individuals' risk-taking behavior (Lian, Ma, & Wang, 2018) . Future studies could address the question of how "cheap money" influences consumers' and marketing managers' decisions. Other highly relevant topics are recessions (e.g., the one caused by COVID-19) and demographics (e.g., the aging global population).

Fourth, the marketing-finance literature has long emphasized returns versus risk, with the former aligning with metrics traditionally studied in marketing, such as market share, revenues, or profits. As a consequence, theories and insights are better developed for returns, even though risk-related topics have received a great deal of coverage in accounting and finance. Increasingly salient are socio-economic and political issues that have the potential to affect firm value (see Bhagwat et al., 2020; Fournier, Srinivasan, & Marrinan, 2020; Josephson et al. 2019) . Given rising stakeholder expectations of risk-laden issues such as immigration, gender and race equality, #MeToo, political ideology, income inequality, climate change, and gun control, firms often find themselves in situations-whether by intended action or by unintended association-in which they confront bad publicity, consumer protests, value-damaging boycotts, and even legal prosecution (Cohen & Gurun, 2018) . Further research is necessary on emerging risks that capture stakeholder attention and lead to brand devaluation, cash flow volatility and firm risk, and attendant firm-value drawdowns.

Finally, future research should consider the topic of human capital and the role of CMOs/top management teams, a frequently mentioned theme among our survey respondents. Ya, Sriniva-J o u r n a l P r e -p r o o f san, Joshi, and Pauwels's (2020) study provides a systematic review and agenda on this topic.

We see several opportunities for future research in the methodological area. First, we call for further research using machine learning to parse marketing-related topics in financial and other data sources. Given the interdisciplinary nature of the research field, marketing-finance re- 

This article provides a comprehensive review of the substantial body of research at the mar- Our review also reveals that the relationship between marketing-related activities (industry-, firm-, and customer-level) and firm value will continue to be one of the most important topics in the next decade, which presents key opportunities. First, the overlap of topics we identified in marketing, finance, and accounting shows that the time is ripe for marketing-finance researchers to enhance interdisciplinary collaborations to translate insights into research targeted at the top finance and accounting journals. Second, there is an opportunity to gain more traction (e.g., on brand and customer disclosures) with regulatory agencies such as the SEC in the US and ESMA in Europe to benefit firms and their stakeholders. We encourage the Marketing Accountability Standards Board to play a more active role in this regard. Third, there is a growing emergence of the crossover of marketing-finance insights from academia into actual practice (e.g., with customer satisfaction and customer-based firm valuation). To accelerate this, a transformation of our J o u r n a l P r e -p r o o f methods into tools that are practically implementable by professionals will be necessary. 8 As a step in that direction, we hope that the collective findings in this article will generate a muchneeded discussion among academics, practitioners, and finance and marketing executives and foster further research on the ever-expanding role of marketing in determining firm value.

Funding: This work was supported by a postdoc fellowship from the German Academic Exchange Service (DAAD) for the first author, who was a visiting researcher at Boston University during the time of writing. Stronger theoretical discussion than in finance/accounting articles (e.g., stakeholder theory in Wies et al. [2019] ; associative-network theory in Borah and Tellis [2016] ; theory of news value in Stäbler and Fischer [2020]) Often assumed that reader is familiar with key finance theories such as the efficient market hypothesis (Fama 1991) ; focus is rather on empirical phenomena (e.g., Focke et al., 2020; Huang, 2018) Direction of investigation Rather from marketing to financial variables Larger percentage of studies that examine the finance  marketing direction Focus of attention marketing variables Rather broad with focus on innovation, advertising, customer satisfaction, and brand equity Rather narrow with focus on innovation, advertising, online metrics, and, particularly, CSR

The "big 4" firm-value variables stock return, Tobin's q, idiosyncratic and systematic risk Stronger focus on analyst and, particularly, investor behavior Single-vs. multi-method approach Traditionally single-method approach, slowly changing to multi-method Larger percentage of multi-method studies (e.g., both directions of investigations in the same study as in Larkin [2013] ) Strategies to establish causality Panel-data and instrumental-variable approaches (e.g., Germann, Ebbes, & Grewal, 2015) ; vectorautoregressive models (e.g., Colicev et al., 2018) Panel-data and instrumental variable approaches (e.g., Chen, Dong, & Lin, 2020); (Quasi-)natural experiments using differences-indifferences (e.g., He & Tian, 2013) or regression discontinuity (e.g., Manchiraju & Rajgopal, 2017) ; field experiments (Lawrence et al., 2018) ; lab experiments (Martin & Moser, 2016)  Five-factor asset pricing model that adds profitability and investment to the original 3-factor Fama-French model (Fama & French, 2015) [ Fornell et al., 2016] ; Hou et al., (2015) also use investment and profitability factors to explain asset pricing anomalies  Behavioral factor model based on investor psychology by Daniel et al. (2020)  New significance hurdles (t ≥ 3.0) for models that try to explain the cross section of stock returns (Harvey et al., 2016 )  Machine-learning approaches to identify relevant asset pricing factors outperform traditional four-factor models in return prediction (Gu et al., 2020 ) Event study  Superiority of retaining confounded events in events (Sorescu et al., 2017 )  Decomposition of abnormal return into manager-predicted and unpredicted abnormal returns (Park et al., 2019 )  Using the cumulative abnormal return on the operative business (CAR OB ) as the dependent  Event-study regression as an alternative to standard analysis of cumulative abnormal returns (Beber & Pagano, 2013; Boehmer et al., 2013) [Hock & Raithel, 2020 ]  New t-statistic that takes into account crosssectional correlation among abnormal returns (Kolari & Pynnönen, 2010) [Feng et al., 2020] J o u r n a l P r e -p r o o f Journal Pre-proof variable  Event studies to analyze change in stock price risk (Carlson et al., 2010) [Thomaz & Swaminathan, 2015] Stock return response models Analysis of the total long-term financial consequences of marketing assets and decomposition into immediate and future effects (Mizik, 2014) Developments from the "Factor models" (see above) apply here Calendar time portfolio Developments from the "Factor models" (see above) apply here Persistence modeling  Structural panel VAR model by Kang, Germann, and Grewal (2016)  Interactions in panel VAR model by Huang and Trusov (2020) Tobin's q models  Formal derivation of the inferiority of the Tobin's q measure for market-based assets studies (Bendle & Butt, 2018) Development of a new "Total q" measure that overcomes many of the limitations of the traditional Tobin's q metric (Peters & Taylor, 2017 ) [Du & Osmonbekov, 2020] Marketing-based valuation 

An experimental investigation of the impact of information on competitive decision making

Does customer satisfaction matter to investors? Findings from the bond market

Mostly harmless econometrics: An empiricist's companion. Princeton

The impact of high-quality firm achievements on shareholder value: Focus on Malcom Baldrige and JD power and Associates awards

Can Twitter help predict firm-level earnings and stock returns?

The impact of online display and paid search advertising relative to offline advertising on firm performance and firm value

Do disclosures of customer metrics lower investors' and analysts' uncertainty but hurt firm performance

Short-selling bans around the world: Evidence from the 2007-09 crisis

Share repurchases and myopia: Implications on the stock and consumer markets

The misuse of accounting-based approximations of Tobin's q in a world of market-based assets

Uniting the tribes: Using text for marketing insight

Does going public affect innovation

Corporate sociopolitical activism and firm value

Corporate brand value and cash holdings

International Journal vice innovations on firm value and firm risk: An empirical analysis

Direct effect of advertising spending on firm value: Moderating role of financial analyst coverage

Marketing's impact on firm value: Generalizations from a meta-analysis

Does the stock market fully value intangibles? Employee satisfaction and equity prices

Efficient capital markets: II

Common risk factors in the returns on stocks and bonds

A five-factor asset pricing model

Is proprietary trading detrimental to retail investors

The impact of unprofitable customer management strategies on shareholder value

Larry Fink's 2019 letter to CEOs

Advertising, attention and financial markets. Review of Financial Studies

Customer satisfaction and stock prices: High returns, low risk

Stock returns on customer satisfaction do beat the market: gauging the effect of a marketing intangible

Re-envisioning marketing for an age of risky business. working paper

Value appropriation and firm shareholder value: Role of advertising and receivables management

Handbook of marketing and finance

The chief marketing officer matters

Large-scale sentiment analysis for news and blogs

Causal inference in accounting research

Old navy will shift back to brand-building in a growing trend among advertisers. cnbc

Empirical asset pricing via machine learning

Valuing customers

Don't believe the hype: Local media slant, local advertising, and firm value

Are online job postings informative to investors? Management Science

Relative strategic emphasis and firm-idiosyncratic risk: The moderating role of relative performance and demand instability

The value of empirical generalizations in marketing

Marketing strategy and Wall Street: Nailing down marketing's impact

Comparing automated text classification methods

… and the cross-section of expected returns

The dark side of analyst coverage: The case of innovation

Drivers of the cost of capital: The joint role of non-financial metrics

Text-based network industries and endogenous product differentiation

Product market threats, payouts, and financial flexibility

Managing negative celebrity endorser publicity: How announcements of firm (non)responses affect stock returns

Digesting anomalies: An investment approach

Brand architecture strategy and firm value: how leveraging, separating, and distancing the corporate brand affects risk and returns

The role of social media and brand equity during a product recall crisis: A shareholder value perspective

Evidence on the information content of text in analyst reports

The customer knows best: The investment value of consumer opinions

Customer satisfaction underappreciation: The relation of customer satisfaction to CEO compensation

Customer satisfaction-based mispricing: Issues and misconceptions. Marketing Science

Uncle Sam rising: Performance implications of business-to-government relationships

Washing away your sins? Corporate social responsibility, corporate social irresponsibility, and firm performance

Birds of a feather: Intra-industry spillover of the Target customer data breach and the shielding role of IT, marketing, and CSR

Assessing performance outcomes in marketing

Event study testing with cross-sectional correlation of abnormal returns

Managing for the moment: The role of earnings management via real activities versus accruals in SEO valuation

Celebrity endorsements, firm value, and reputation risk: Evidence from the Tiger Woods scandal

A thematic exploration of digital, social media, and mobile marketing: Research evolution from 2000 to 2015 and an agenda for future inquiry

Modeling heterogeneity in the satisfaction, loyalty intention and shareholder value linkage: A cross-industry analysis at the customer and firm level

Brand perception, cash flow volatility, and financial policy

Earnings announcement promotions: A Yahoo Finance field experiment

Metrics for making marketing matter

Low interest rates and risk-taking: Evidence from individual investment decisions

Investors' evaluations of price-increase preannouncements

When is a liability not a liability? Textual analysis, dictionaries, and 10-Ks

Textual analysis in accounting and finance: A survey

Does advertising spending really work? The intermediate role of analysts in the impact of advertising on firm value

Customer satisfaction, analyst stock recommendations, and firm value

Social media and firm equity value. Information Systems Research

Introduction to the Journal of Marketing Research special interdisciplinary issue on consumer financial decision making

From finance to marketing: The impact of financial leverage on customer satisfaction

Does corporate social responsibility (CSR) create shareholder value? Evidence from the Indian companies act 2013

Data privacy: Effects on customer and firm performance

The role of data privacy in marketing

Managers' green investment disclosures and investors' reaction

Customer-based corporate valuation for publicly traded noncontractual firms

Valuing subscription-based businesses using publicly disclosed customer data

Investor sentiment and advertising expenditure

What drives managerial use of marketing and financial metrics and does metric use affect performance of marketing activities

The theory and practice of myopic management

Assessing the total financial performance impact of brand equity with limited timeseries data

Financial markets research in marketing

Organizing for marketing excellence

Assessing marketing strategy performance

JM as a marketplace of ideas

Firm innovation and the ratchet effect among consumer-packaged goods firms

Chief marketing officer presence and firm performance: assessing conditions under which the presence of other C-level functional executives matters

Sales force downsizing and firm-idiosyncratic risk: The contingent role of investors' screening and firm's signaling processes

A guide to using event study methods in multi-country settings

Capital market returns to new product development success: Informational effects on product market advertising

Intangible capital and the investment-q relation

The marketing firm as a metacontingency: Revealing the mutual relationships between marketing and finance

Linking customer and financial metrics to shareholder value: The leverage effect in customer-based valuation

The impact of corporate social responsibility on firm value: The role of customer awareness

What should be the dependent variable in marketing-related event studies

Selective publicity and stock prices

Two centuries of innovations and stock market bubbles

Event study methodology in the marketing literature: An overview

Relating online, regional, and national advertising to firm value

With power comes responsibility: How powerful marketing departments can help prevent myopic management

Marketing and firm value: Metrics, methods, findings, and future directions

Product innovations, advertising, and stock returns

When does corporate social irresponsibility become news? Evidence from more than 1,000 brand transgressions across five countries

What goes around comes around: The impact of marketing alliances on firm risk and the moderating role of network density

Does chatter really matter? Dynamics of user-generated content and stock performance

Talented people and strong brands: The contribution of human capital and brand equity to firm value

The impact of psychological identification with home-name stocks on investor behavior: an empirical and experimental investigation

Interpreting the stock returns to new product announcements: how the past shapes investors' expectations of the future

When 1 + 1 > 2: How investors react to new product releases announced concurrently with other corporate news

Can advertising investments counter the negative impact of shareholder complaints on firm value

Going public: How stock market listing changes firm innovation behavior

Stock market response to regulatory reports of deceptive advertising: the moderating effect of omission bias and firm reputation

The effect of cause-related marketing on firm value: a look at Fortune's most admired all-stars

Asymmetric roles of advertising and marketing capability in financial returns to news: Turning bad into good and good into great

How CEO/CMO characteristics affect innovation and stock returns: Findings and future directions

Managing the future: CEO attention and innovation outcomes

Organizational learning and technological innovation: the distinct dimensions of novelty and meaningfulness that impact firm performance

Long-term event study (n = 11)

 Fig. 2 . Metrics studied at the marketing-finance interface before and after 2009.

Equity-related metrics (374)

• Idiosyncratic risk (29) • Cash-flow volatility (6) Metrics for young businesses (7) • Going public (3) • IPO value (3) Debt-related metrics (8) • Credit spread (4) • Credit rating (3) Equity-and debt-related metrics (9) • Leverage (9) (no new metrics) Product-market performance (e.g., sales, market share) and Accounting performance (e.g., revenues, profitability) [important mediators of the marketing-finance interface, but not focus of this review article]

Analysts (25) • Analyst coverage (8) • Earnings forecast error (7) Investors ( • Marketing capability (4) • Absorptive capacity (1)

• Firm's cultural orientation (2) • Organizational service climate (1) Notes: Metric (sub)categories and variables in black ("old" are those that had been investigated at least once before 2009. Metric categories and variables in red ("new") have been investigated since 2009. The first variable in each (sub)category is the most-investigated variable (can be "old" or new"). The second variable is either the "new" variable with the highest coverage (if "old" variable has highest coverage) or the second-highest "new" variable (if "new" variable has highest coverage). The number of studies is in parentheses. comp stands for competitor studies.

• Alliances (17) • Outsourcing (2) Human capital (20)Workforce-related (6) • Employee satisfaction (3) • Human capital (1) C-suite related (14 ) • Top-executive compensation (5) • CMO appointment (2) J o u r n a l P r e -p r o o f (consumers) a Total number of studies that investigate the impact of marketing on financial variables. For example, 32 studies examine the effect of customer satisfaction on any financial variables (not just the ones listed here). Similarly, four studies assess the effect of any marketing variables on earnings forecast error. Notes: The table includes main effects based on model-based results (regression coefficients). One study can contribute more than one effect (e.g., due to different methodological approaches, different samples). Gray fields indicate that the relationship has not been studied.J o u r n a l P r e -p r o o f Journal Pre-proof