1 Introduction

Recent research notes the importance of the interplay between firms’ strategies and capital markets, as companies internationalize (e.g., Filatotchev et al., 2016; Müllner et al., 2023; Oxelheim et al., 2001; Puck & Filatotchev, 2020; Purkayastha & Kumar, 2021). Multinational corporations’ (MNCs’) international expansion decisions and performance are not only influenced by well-known ownership- and location-specific advantages and the transaction costs affecting global operations (Buckley & Casson, 1976; Dunning, 1979; Rugman & Verbeke, 1992), but also by their financing choices, such as initial public offerings (IPOs) (Jiang et al., 2022) and cross-listings (Forssbæck & Oxelheim, 2008; Peng & Su, 2014). Likewise, the extent and patterns of international expansion by firms can have an impact on stock market reactions, firms’ cost of capital, and their participation in capital markets (Denis et al., 2002; Doukas & Kan, 2006; Lindorfer et al., 2016).

While research on internationalization often holds implicit assumptions on capital availability and firms’ abilities to overcome problems due to information costs in capital markets (Bell et al., 2012), research has not systematically investigated MNCs’ specific activities in capital markets to address information asymmetries surrounding their internationalization initiatives (e.g., Puck & Filatotchev, 2020). For instance, although information asymmetry is believed to contribute to investors’ initial negative reaction to cross-border acquisitions (Aybar & Ficici, 2009; Cho & Ahn, 2017), there is little research as to how the MNC might subsequently counter this effect. In this study, we aim to contribute to the literature on firms’ cross-border strategies and their financing and governance practices through a detailed analysis of how executives engage with analysts and disclose information in earnings calls.

A focus on the firm’s tactics to alleviate information asymmetry between the MNC and the investing public is different from, yet complementary to, the current focus of the international business (IB) literature on the information costs associated with international expansion. For instance, many studies examine company- or host country-specific information costs firms encounter in dealing with foreign partners and institutions in product markets and focus on the ways in which firms might address the information asymmetries, such as through their market entry strategies (e.g., Barbopoulos et al., 2018; Cuypers & Martin, 2010; Grøgaard et al., 2019). What is lacking, however, is a conceptual and empirical analysis of how MNCs engage with capital markets and its stakeholders, to reduce information asymmetry between firms, their foreign investments, and investors. There have been a few recent studies investigating this level of information asymmetry, but their focus has been notably on the domestic setting, and away from the challenges and complexity of the international context (e.g., Ragozzino & Reuer, 2024; Yakis-Douglas et al., 2017). Thus, while sorting out the adverse effects of information asymmetry with international acquisition or alliance partners is a tall order in and of itself, our goal is not to address this topic. Instead, we focus on a distinct issue, which pertains strictly to information disclosure between firms, their international corporate strategy, and the investing public.

In this sense, our work is complementary to extant research in IB and finance, which has examined how investors evaluate internationalized firms and foreign companies (e.g., Denis et al., 2002; Ke et al., 2010), from the perspective of how equity investors react to MNCs’ internationalization decisions (Aybar & Ficici, 2009; Li et al., 2021; Moeller & Schlingemann, 2005). We build upon and extend these streams of research in a new dimension by examining how MNCs engage with capital markets by disclosing information regarding their internationalization transactions. We argue that while public firms have regulatory obligations to disclose key developments in their corporate strategies and voluntary disclosures have also generally been found to improve access to capital markets (Bushee et al., 2018; Frankel et al., 1999), the relationship between firms’ strategic actions and the information they relay to stakeholders in earnings calls is theoretically ambiguous, for several reasons.

First, insiders may hold incentives for engaging in corporate strategies that depart from the best interest of the shareholders, and this agency issue may also well curb information disclosure (Stiglitz, 2002). Second, even in the absence of agency concerns, leveraging managers’ proprietary foresight in the formulation and implementation of corporate strategy transactions is central to creating and sustaining firms’ competitive advantages (Angwin, 2007). Disclosing the details of these insights to the public may enable competitors to imitate the company’s actions, thereby imposing so-called proprietary costs on the company (Berger & Hann, 2007) and preventing the managers from bringing their vision to fruition (Litov et al., 2012).Footnote 1 Third, as the complexity of companies’ corporate actions increases, the need to relay more and more complex information also increases. This directly affects the burden of information disclosure borne by executives, who are faced with the task of covering several potentially relevant items during the time-bound limits of earnings calls. Similarly, the information processing cost borne by analysts and shareholders also increases as they try to develop an informed valuation of the company (Benner & Zenger, 2016). In turn, the heightened complexity of corporate communication can lead to a greater discount applied by shareholders to the company’s value, and finally to a greater cost of capital faced by the company (Guay et al., 2016). Therefore, besides the agency issue, the concerns over the cost of processing and interpreting firms’ disclosures and the proprietary costs stand in the way of full and complete disclosures of information regarding firms’ strategies.

To improve our understanding of the relationship between internationalization strategy and corporate disclosures, we investigate earnings calls hosted by newly public firms with analysts and examine the degree to which executives disclose information related to their international activities when engaging in international alliances and M&A. We maintain that newly public firms provide an attractive context to examine the research question under investigation. Despite the information provided in IPO prospectuses concerning the firm’s past operations and future strategies, outside investors have far less information about the firm’s capabilities and intentions than the managers themselves (Certo et al., 2009). Furthermore, these investors may discount information supplied at the time of an IPO because managers have an incentive to provide overly-optimistic projections (Lehmann & Vismara, 2020). These information challenges tend to be exacerbated when newly public firms expand overseas since capital market participants not only have imperfect information on firms’ international strategies but also on the resources and prospects of international alliance partners and acquisition targets in the deals that firms pursue following their IPOs. Under such circumstances where firms’ liabilities of foreignness and their newness to capital markets combine to create substantial information asymmetries between firms and capital markets, the firms’ approach to information disclosure in earnings calls becomes particularly important for shareholders to properly determine the firm’s market value.

Moreover, in addition to examining information disclosures in cross-border transactions, we investigate the degree to which executives provide information specifically related to their corporate strategies. By collecting detailed information on the exchanges between analysts and executives in earnings calls, we are able to track the questions that analysts pose to executives regarding their firms’ internationalization and corporate activities. We then deploy matched samples of active and inactive firms, in order to address the potential for unobserved heterogeneity that tends to confound findings in this sort of empirical settings. Our findings indicate that international alliances and acquisitions lead executives to disclose information associated with these transactions in general. Partially guiding this disclosure, we find that analysts raise direct questions about firms’ corporate strategy, but the questioning is nearly entirely surrounding the firm’s acquisition activity. In contrast, alliances are not subjected to a comparable line of inquiry by analysts. In supplementary analysis we examine a number of additional contingencies, such as whether certain types of internationalization affect information disclosure more than others, by separating non-equity alliances from equity joint ventures, as well industry related from unrelated transactions. Taken together, the results we bring to the fore offer a novel and nuanced perspective of how firms exposed to the hazards of information asymmetry, and engaged in cross-border corporate growth activities tackle the important task of information disclosure to shareholders.

This study contributes to the IB and strategy literature in several ways. At the broadest level, we devote attention to capital markets as an important albeit neglected factor market that is related to firms’ internationalization, and we do so by joining together research on firms’ financing and governance activities with research on their international strategies. We also build upon and extend research that has considered information asymmetries affecting shareholders scrutinizing their firms’ corporate activity in general, and international alliances and acquisitions in particular. Future research might follow this study in drawing connections between economic transactions in certain product markets (e.g., international alliances and acquisitions) and the firm’s activities in other factor markets (e.g., information disclosures in capital markets) to be able to fully account for the information costs that surround internationalization. Lastly, while text mining is increasingly used in organizational research (Banks & Aguinis, 2023; Kobayashi, Mol, Berkers, Kismihok, & Den Hartog, 2018; Ragozzino & Reuer, 2024), its application specifically in the IB literature has been limited. Our study demonstrates the usefulness and potential of computer aided text analysis in developing and testing novel theoretical predictions in IB research.

2 Theory and Hypotheses

2.1 Background Theory

When firms internationalize through acquisitions or alliances, they face significant exchange hazards that accompany their international transactions. Information economics provides a useful theoretical lens for understanding some of these hazards: When there are information asymmetries across exchange partners regarding their resources and market prospects, negotiations can break down before a transaction is completed, or become inefficient, resulting in overpayment by the buyer or in the selection of suboptimal partners (e.g., Akerlof, 1970; Spence, 1973; Stiglitz, 2002). For instance, in Akerlof’s (1970) classic model of the market for used cars, buyers face a risk of overpayment, or adverse selection, because they cannot distinguish cars of higher versus lower quality. If this risk of adverse selection is left unaddressed by social or asocial remedies (e.g., warranties, reputation, etc.), the market can fail to a point where only lower quality cars are transacted. Similarly, in Spence’s (1973) classic model of labor markets, prospective employees face challenges obtaining positions at high wages if they are unable to credibly convey their unobserved productivity to employers via educational attainment.

Research in many fields has noted that such adverse selection concerns surface in alliances and acquisitions, particularly when transactions involve firms’ intangible assets. Firms face costs in evaluating each other’s resources and market opportunities prior to completing transactions and therefore they put at risk the financial and other resources they bring to such transactions (e.g., Chi, 1994; Merchant & Schendel, 2000). As a consequence, firms’ alliance prospects improve with their network connections and the signals conveyed on their resources and commercial prospects (e.g., Hsu, 2006; Nicholson et al., 2005). Likewise, research in other fields has noted how firms cope with information asymmetries by staging investments, selecting targets presenting lower information asymmetries, avoiding sunk investments, or using contractual provisions and contingent payments to mitigate risks that acquisitions entail (Balakrishnan & Koza, 1993; Capron & Shen, 2007; Gilson & Schwartz, 2005; Higgins & Rodriguez, 2006; Macias & Moeller, 2016; Schildt & Laamanen, 2006; Villalonga & McGahan, 2005).

Existing research notes that the problems of asymmetric information and adverse selection surrounding these external corporate activities can be expected to magnify in the international context (Dikova et al., 2010; Ke et al., 2010). For instance, the liability of foreignness suggests that foreign firms will have worse information on local market opportunities, resources in firms situated overseas, and the affiliates of these organizations (e.g., Hymer, 1976; Zaheer, 1995). Information asymmetries also tend to increase naturally with geographic distance, given that they often concern soft information and knowledge embedded in localized relationships and interactions. Moreover, adverse selection risks can be significant in international transactions, because institutions whose purpose is to gather and diffuse information to the public may be lacking in many countries (Ahuja & Yayavaram, 2011; Jandik & Kali, 2009; Li et al., 2019). Research suggests that if information asymmetries are not addressed through staged investment, reliance on signals or contracts, promising deals can fail to materialize or break down (e.g., Cuypers & Martin, 2010; Dikova et al., 2010; Reuer & Ragozzino, 2014).

Internalization and governance theories lend additional credence to the relevance of information asymmetry (Grøgaard et al., 2019): As an illustration, the classic internalization framework suggests that information asymmetries between buyers and sellers, along with other factors such as the lack of futures markets for knowledge generation and imperfect knowledge pricing, are the sources of market imperfections that lead MNCs to internalize markets for intermediate goods across national borders and bring economic activities under common ownership and control (Buckley & Casson, 1976; Hennart, 1982; Rugman, 1981; Rugman et al., 2011). Forsgren (2017) further suggested that market failure in internalization theory can be seen as a result of both bounded rationality, resulting from information asymmetries between parties in cross-border transactions, and opportunism, resulting from transacting parties taking advantage of information symmetries.

While the large body of research briefly outline above has offered significant insight into the role of information asymmetry between transacting companies, this study departs from this perspective and focuses instead on the information asymmetry separating a public firm and its own shareholders in capital markets. The main observation guiding our research is that when firms internationalize via alliances and acquisitions, they not only face the challenges of adverse selection when conceiving of, structuring, and carrying out transactions. They also need to be able to explain their international strategies and transactions to the investing public, in order to allow the latter to develop an informed expectation of value and risk of these firms. Thus, international activities give rise to information asymmetries in both product and capital markets, and a firm’s ability to address these challenges in one market is separate from its ability to address the same in the other market. Notably, the topic of information disclosure in capital markets carries important competitive consequences for firms, as the failure to convey information about a company’s strategic actions to shareholders will affect firms’ cost of capital, thereby hindering their ability to engage in corporate deal making in the first place. Despite the relevance of the topic, there is little work in this area, and our goal is to begin to remedy this shortfall and submit that research needs to attend to how firms interact with capital markets, and how this interaction is shaped by their international alliance and acquisition activities.

2.2 Earnings Calls

It is useful to explain our motivation to use earnings calls as the setting for our investigation. Regulatory authorities such as the Securities and Exchange Commission (SEC) in the US mandate disclosures by firms to capital markets as a condition for firms to trade on stock exchanges. These information disclosures are critical for capital market participants to have adequate information on firms’ performance, activities, and plans in order to make investment decisions, and information disclosures also promote market efficiency more generally. Beyond providing information via required audited financial reports, companies also take part in voluntary quarterly earnings calls. These calls afford executives opportunities to disclose information on the firm’s strategy, investments, and operations as well as provide opportunities for analysts to ask spontaneous, probing questions on the firm’s various activities (e.g., Bushee et al., 2004; Frankel et al., 1999; Harrison et al., 2019; Pan et al., 2018).

As a consequence, these calls are important contexts in which information asymmetries exist and might be reduced, and they can ultimately influence analysts’ stock recommendations and earnings forecasts (Bushee et al., 2018; Henry & Leone, 2015; Mayew & Venkatachalam, 2012). The richness of information provided by these calls has enabled recent research from many different theoretical traditions across disciplines (for a review of research in management, see Gao et al., 2016). In sum, this developing literature suggests that earnings calls are an important means by which executives provide information on a firm’s strategy and activities to capital market participants, just as analysts use these calls as part of their information intermediary role in equity markets.

There is also substantial evidence that information asymmetry contributes to bias among portfolio investors against holding foreign securities (Ke et al., 2010; Lundholm et al., 2018) and less precise earnings forecasts by financial analysts for foreign firms (Bae et al., 2008). Analogously, IB scholars have noted that asymmetric information problems can lead financiers to apply unfavorable evaluation methods and hence exhibit a bias against proposed foreign investment projects, resulting in underinvestment by aspiring MNCs (De Maeseneire & Claeys, 2012). Recent research suggests that firms can mitigate information asymmetry facing the participants in global capital markets by improving the quality of corporate disclosures by adopting international accounting standards or by simply adopting English as an external reporting language (Baik et al., 2013; Bell et al., 2012; Jeanjean et al., 2015). However, although these studies identify governance mechanisms through which MNEs can interact with the capital markets to enhance the quality of disclosure, there is limited research on whether MNCs may address analysts’ or investors’ concerns over cross-border transactions and mitigate transaction-specific information asymmetry. We are therefore able to join this developing literature with research on international alliances and acquisitions to determine how these deals might also bear upon firms’ information exchanges with capital markets.

2.3 Research Hypotheses

The starting premise of this research is that executives and analysts will be concerned with unearthing information with respect to both the international aspect of their alliances and acquisitions, as well as the specifics of these transactions and how they may help to advance firms’ strategies. For their part, keeping in mind that earnings calls are voluntary information disclosure mechanisms, and that the information conveyed during these calls is entirely discretionary, executives may be hesitant to fully disclose the details of their strategies during earnings calls, as doing so may cause them to inadvertently deliver sensitive information to competitors, who might then imitate their activities, or otherwise organize themselves to fend off new threats (Hsu et al., 2010). Moreover, executives may withhold information about certain international transactions from the investing community because these transactions may enhance executives’ personal welfare at the expense of the shareholders (e.g., lack of due diligence for an acquisition in a politically unstable host country) (Stiglitz, 2002). Lastly, because international activities typically represent multitudes of risks even for experienced MNCs, disclosures can thus expose weaknesses or even risks and failures to investors, customers, and others as the firm internationalizes.

Despite these potential downsides of disclosing information on the firms’ international activities and plans, firms will generally have some incentive to disclose information on international activities on earnings calls for several reasons. Obtaining capital ultimately requires transparency (Chemmanur & Fulghieri, 1999). Were firms not willing to provide information on their international activities and expansion plans, equity market participants would interpret such withholding of information with skepticism, with adverse effects on their valuations and ability to finance the pursuit of growth opportunities in the future. Consistent with this observation, while engaging in information disclosures during earnings calls is voluntary, in fact 97 percent of public companies in the US hold such calls, according to the National Investor Relations Institute (https://www.niri.org/).

Previous research has also found that the disclosure of information increases with organizational complexity, which is closely related to the firm’s growth and corporate activities (e.g., Frankel et al., 1999). Companies that have simple businesses in stable operating environments are expected to have relatively short and straightforward calls, whereas it is more difficult to make attributions on the performance of executives with complex operations and those engaged in noisier operating environments. Internationalization results in an increase in organizational complexity owing to the new demands placed on managerial resources as well as to the multiple new external environments the firm must contend with (e.g., Penrose, 1959; Sundaram & Black, 1992). The complexity of firms’ strategies as well as their operations increases as firms internationalize due to the multiplicity of entities they confront, the new interactions brought about when partnering with or acquiring overseas firms, and the dynamics of global competition and the economic systems in which the firms are embedded (e.g., Eden & Nielsen, 2020).

The foregoing implies that cross-border M&A and alliance activities are prominent corporate activities with significant performance implications, and thus are likely of great interest to the capital market. We therefore posit that executives will prioritize disclosing more about their firms’ international activities when they are engaged in international alliances and acquisitions:

Hypothesis 1a (H1a) Executives in earnings calls will provide more information on the firm’s international activities when it engages in international alliance activity.

Hypothesis 1b (H1b) Executives in earnings calls will provide more information on the firm’s international activities when it engages in international M&A activity.

As executives explain their international expansion plans, they not only are expected to explain the international facets of their activities and growth ambitions, as these hypotheses suggest, but they will also likely convey information that is specific to these investment vehicles within their corporate strategies. Acquisitions are typically more complex as they involve a series of steps ranging from obtaining regulatory approvals to detailed integration planning prior to closing, and the most complex tasks of integration follow that are related to structures, cultures, functions, and business processes. It is therefore not surprising that equity markets often respond skeptically to the announcement of acquisition transactions (e.g., Kaplan & Weisbach, 1992). Executives will therefore carry out due diligence not only for their own purposes before coming to terms with a target, but they will also disclose details of their M&A strategy, the basis for value creation (e.g., accessing new markets, obtaining key resources, etc.), and key terms of the transaction (e.g., method of payment, governance, etc.). As a result, failure to disclose such pertinent information on corporate strategy and M&A can adversely affect the firm’s access to capital on attractive terms. Thus, accounting research reports that executives convey private information on M&A surrounding announcements of these deals, and this disclosure leads to more positive shareholder reactions to M&A (i.e., Kimbrough & Louis, 2011).

For their part, international alliances can often involve lower financial outlays, but they can still expose the organization to key risks (e.g., Reuer & Leiblein, 2000). Many of these risks stem from cooperation and coordination challenges required for coping with the complexities that alliances entail. Executives therefore can discuss how alliances fit within the firm’s corporate strategy, the ways in which individual alliances are expected to create value, key alliance investment and design parameters (e.g., rationales for partner selection, form of the alliance, governance, etc.), and the ways in which the firm can address cooperation and coordination risks while operating under an incomplete contract. We therefore expect that firms engaging in international alliances and acquisitions will not only disclose information focused on the international dimension of their activities, but will also prioritize providing information on their corporate strategy and the challenges that attend alliances and acquisitions in particular. Thus, we posit:

Hypothesis 2a (H2a) Executives in earnings calls will provide more information on the firm’s corporate strategy when it engages in international alliance activity.

Hypothesis 2b (H2b) Executives in earnings calls will provide more information on the firm’s corporate strategy when it engages in international M&A activity.

In the foregoing discussion, we have associated firms’ international alliance and acquisition activities with executives’ disclosure of information on the firm’s international activities and corporate strategy. However, we also expect that these activities will lead to probing questions by analysts during earnings calls. In fact, analysts may well steer the conversation to those topics over which information asymmetries exist, by raising questions on complex cross-border corporate strategy deals in which the firm might have entered in the recent past. It is plausible that analysts will not only ask for clarification on the general aspects of these deals, but also bring more nuanced inquiries to the table, in order to obtain detailed information which they will use to make stock recommendations and develop earnings forecasts in the following quarter (e.g., Bowen et al., 2002; Heinrichs et al., 2019; Kimbrough, 2005; Lobo et al., 2017). Arguably, even though executives may remain cagy about sensitive topics and fail to answer questions comprehensively and to the full satisfaction of analysts, it is the very role of analysts to use the limited time available to them to talk to company insiders and mitigate information asymmetries (e.g., Bochkay et al., 2020; Bushee et al., 2018; Frankel et al., 1999; Rabier, 2017). We therefore propose a set of hypotheses that parallel those above and relate to analysts’ interests in asking probing questions related to firms’ international activities and the alliances and acquisitions they engage in. Thus, we posit:

Hypothesis 3a (H3a) Analysts in earnings calls will ask more questions on the firm’s international activities when it engages in international alliance activity.

Hypothesis 3b (H3b) Analysts in earnings calls will ask more questions on the firm’s international activities when it engages in international M&A activity.

Hypothesis 4a (H4a) Analysts in earnings calls will ask more questions on the firm’s corporate strategy when it engages in international alliance activity.

Hypothesis 4b (H4b) Analysts in earnings calls will ask more questions on the firm’s corporate strategy when it engages in international M&A activity.

3 Methods

3.1 Data and Sample

Our empirical analysis focuses on newly public firms. The IPO literature suggests that one of the primary motivations for going public is to facilitate merger and acquisition activities, and newly listed firms tend to engage in M&A and alliances at an elevated rate shortly after their initial offerings (Brau & Fawcett, 2006; Chod & Lyandres, 2011; Reuer & Ragozzino, 2014; Wu, 2012). In the meantime, IB research indicates that a firm’s strengthened financial position following the IPO can help alleviate managers’ concerns over the increased operational and institutional complexity resulting from international expansion and can lead them to take more risks when the additional liquidity insures that they will not jeopardize the firm’s overall financial viability (Jiang et al., 2022).

However, not only do international corporate activities create new opportunities for firms (e.g.,Bertrand & Capron, 2015), they also carry with them substantial risks (Miller, 1992), especially when foreign deals occur at an increased pace (e.g., Jiang et al., 2014; Kim et al., 2020). Accordingly, research has shown that investors and financial analysts use higher discount rates for evaluating firms’ international operations (De Maeseneire & Claeys, 2012; Denis et al., 2002; Doukas & Kan, 2006; Fauver et al., 2004), and it may be no surprise that increased acquisition activities can be a contributing factor to a commonly observed phenomenon – the long-run underperformance of IPO firms – owing to the tendency to overinvest after the IPO (Brau et al., 2012). Thus, many newly public firms face a dilemma regarding post-IPO internationalization: on the one hand, these firms may pursue international growth opportunities, such as cross-border acquisitions and alliances, by using IPO proceeds and by leveraging their improved access to capital markets (Aktas, Andres, & Ozdakak, 2018); on the other hand, heightened international activities contribute to information asymmetry between these newly public firms and the investing public, which may undermine the firm’s ability to garner investors’ support for impending transactions, or to obtain future capital on attractive terms. Therefore, there is a need for these firms to address information asymmetry when the complexity of their strategies and operations increases as they internationalize (Cahan et al., 2005; Cannizzaro & Weiner, 2015; Krause et al., 2021). In sum, IPOs bring tensions between firms’ pursuit of international growth opportunities and the information asymmetries experienced by capital market participants.

The final sample used in the analysis draws from several data sources. First, we constructed a roster of firms that went public in the United States between 2002 and 2017 across all industries, except for the broad financial sector (i.e., SIC codes 60–67), as is routinely done in the literature on IPOs (e.g., Ljungqvist & Wilhelm Jr, 2005). The resulting 1,819 IPOs were then matched against a database of quarterly earnings call transcripts, which we obtained from Seeking Alpha, a crowdfunded organization that collects and diffuses information to investors. More precisely, we matched each newly public firm with this transcript database, so as to capture every earnings call released by the firm starting in the quarter following its IPO and up to five years hence. Earnings call transcripts are text files which record the interactions that occur among executives, analysts and the operator during a call. We were able to mine these files and construct a database that isolates each interaction, along with the identity of the speaker. We further augmented these data by merging them with Compustat quarterly data, in order to obtain general accounting information on the firms comprising our sample. Requiring that data be available for all firms from these sources resulted in 4,407 observations covering 861 unique firms, or an average of just over 5 transcripts per firm.

The next step was to track firms’ corporate domestic and international transactions before their IPOs and thereafter. To this end, we collected data from Thomson One (formerly Security Data Corporation), which is the key data source for firms’ M&A and alliance activity and provides the most comprehensive available accounts of these transactions. For M&A, we only considered deals that were in fact consummated (as opposed to merely announced), and we excluded transactions listed as acquisitions, but fulfilling purposes other than corporate growth.Footnote 2 In the case of strategic alliances, we included all equity and non-equity agreements.Footnote 3 The baseline sample was subsequently truncated to eliminate outliers, and yielded 4,091 usable observations for the implementation of the next steps of our empirical analyses.

Matching samples. Insofar as we are attempting to find associations between firms’ post-IPO international corporate activities and the information the firms relay to the investing public, it is important that this corporate activity be the only difference separating more and less active firms. In other words, we needed to eliminate (or at least minimize) the possibility that unobservable heterogeneity might simultaneously affect the propensity to engage in international corporate activity and the information disclosed by these firms. In the absence of an experimental setting, in which an exogenous shock provides a clear-cut separation between treatment and control groups, we resorted to propensity score matching and coarsened exact matching techniques. Specifically, we first ran logistic regressions in which we predicted the occurrence of post-IPO international alliances or acquisitions for up to six months leading up to the date of the focal earnings call.Footnote 4 These models account for the firm’s size at IPO (defined as the log-transformed value of its IPO proceeds), the exchange in which it is listed, time and industry fixed effects, as well as whether or not the firm was backed by a venture capital (VC) firm when it went public. The estimation of these models resulted in the creation of propensity scores, which are effectively the probabilities assigned to each firm to engage in international alliances or acquisitions. We ranked all firms based on their propensity score, breaking the sample down into five distinct groups.

Next, we separated active and inactive firms and matched them based not only on their propensity score group, but also on their industry and the time at which the observation was taken. More precisely, we sought to identify as many as three inactive firms in the same industry and in the same quarter and year as their active benchmark. These matching criteria helped us to eliminate industry- and time-specific effects that may drive firms’ propensity to engage in international alliances or acquisitions. The implementation of these steps eventually resulted in the creation of two separate samples: one sample contains firms active and inactive in strategic alliances, while the other sample separates otherwise highly comparable firms based on their engagement in M&A activity. The alliance sample is made up of a total of 1,652 observations, of which 568 involve active firms. The M&A sample contains 1,854 observations, of which 441 entered in at least one acquisition transaction in the period leading up to their earnings conference call.Footnote 5

3.2 Variables and Specifications

Dependent variables. The dependent variables in our models reflect the utilization of key words relating to either international activity or corporate activity by executives and analysts during quarterly earnings calls. We build custom dictionaries of words, as follows: For the international dictionary, we tracked the use of the following words, as spoken by executives and analysts, respectively: INTERNAT*, CROSS-BORDER (with or without the hyphen), FOREIGN*, ABROAD, OVERSEA*, TRADE, EXPORT*, IMPORT*, TARIFF* and GLOBAL*. For the corporate activity, we created separate dictionaries for M&A and alliance transactions. The M&A dictionary contains these words: MERG*, ACQUIR*, ACQUIS*, M&A, CONSOLIDAT* and DIVERSIF*. The alliance dictionary contains these words: ALLIANCE, ALLY, JV, JOINT, COOPERAT*, and CO-OPERAT*. We counted the number of instances in which these words were spoken and then scaled these counts by the total number of words spoken in the relevant segment of the call.Footnote 6 While the specifications of the dictionaries above is arbitrary, we experimented with several variations and found that the results were not sensitive to incremental changes to our lists of keywords. Also, for clarity, below we present excerpts from two separate transcripts, which provide examples of the words we extracted from the text (underlined):

Example 1: Oracle Corporation, fourth quarter 2011, question by an analyst for the executives in attendance:

Right on, guys. I guess, I'm curious [ph] a little bit I'm curious with, I think, you said $29 billion a lot of cash on the balance sheet. And I'm just curious how you guys are thinking about the acquisition strategy? I mean, valuations for a lot of private and small and mid-capped companies aren't necessarily at trough levels. There's also some folks out there who seem to think that Oracle has to buy to keep growing. Just what's your sort of philosophy right now? How would you sort of couch your guidance to think about the use of cash in the M&A front?

Example 2: General Motors Company, second quarter 2016, answer from an executive in attendance:

We have certainly factored in our footprint and manufacturing footprint. The fact is we import a heck of a lot more units into the UK than we export - manufacture and export out, so we're long pound sterling. There is very little pricing mitigation assumed in that $400 million for the balance of the year specifically because, number one, it's early days. Number two, before you can take pricing action, you have to sell through inventory and the overall inventory levels in the UK are reasonably high across all of the OEMs. And that will certainly be one of the items as we work through this and see how it develops that we will be looking at it, as I'm sure other competitors will as well, as well as cost and other mitigating actions. So primarily, as I said earlier, the $400 million, looking at our overall position, primarily weakness in the pound sterling and some industry headwind.

Covariates. The theoretical variables of interest are dummy indicators separating firms active in international alliances or M&A from firms that were inactive in these transactions after going public. As previously discussed, we tracked firms’ engagement in alliances and acquisitions over the course of six months leading up to the firm’s earnings call and then assigned a value of one if the firm had been active during this period, and zero otherwise.

In addition to the theoretical variables just described, we introduce a host of controls to account for other factors that might also explain the propensity to use the words in our custom dictionaries during quarterly earnings calls. First, we control for the firm’s pre-IPO corporate activity (i.e., Pre-IPO activity) over a period of 5 years leading up to the IPO, because firms that are historically active might also have a natural tendency to discuss their international and corporate agendas. These variables are counts of the firm’s total alliances (M&A) and they are logged in the models as a remedy for positive skewness. In the models predicting the use of words contained in our corporate dictionaries, we also control for the firm’s domestic corporate activity (i.e., Domestic activity) over the same six-month window used for the theoretical variables. As before, we track alliance and M&A activities separately, depending on the sample used for the analyses and we assign a value of one and zero to active and inactive firms, respectively. Second, we incorporate a control variable – Months from the IPO – which counts the number of months elapsed from the focal firm’s IPO to the date of the earnings call. The main reason for including this control is that as the firm become more mature after its IPO, it will become more adapt at conducting earnings calls, which might affect what information will be disclosed during these calls. Third, we control for whether the firm is listed on the NYSE to account for possible differences in analyst scrutiny driven by exchange-specific considerations. The variable Firm listed in the NYSE is a dummy that takes the value of one if the firm was listed on the New York Stock Exchange, and zero otherwise.Footnote 7 The next set of controls are basic accounting and financial variables often used in IPO and acquisition studies: Firm leverage, Firm market cap and Firm ROA track indebtedness, size and profitability, respectively. Finally, we control for year and industry fixed effects, separating industry across manufacturing, services and other sectors, and also by whether the firm operated in a high-technology industry, regardless of main industry.Footnote 8

Models. The models used to test our formal hypotheses on information disclosures are Tobit regressions, as the dependent variables are continuous and truncated in the lower bound at zero.Footnote 9 Specifically, zero values may be observed if no words in the custom dictionary are spoken, while values of one represent unrealized theoretical scenarios in which every word spoken belonged in the custom dictionary. We estimate a total of 8 Tobit models, based on a 2 × 2x2 structure, as follows: First, we track the use of custom words across company executives versus analysts. Within each of these groups, we separate the samples based on firms’ activities related to alliances and acquisitions, as previously discussed. Lastly, we use two separate custom dictionaries – one for international words, and the other for the specific mode of corporate activity (alliances or M&A, depending on the models).

4 Results

Table 1 shows the results of the logistic regressions estimated for the purpose of obtaining the propensity scores needed for matching active and inactive firms. The first model predicts the use of international alliances by the firm, while the second predicts the use of M&A. While the interpretation of the results of these models is outside the scope of our formal predictions, it is interesting to note that international corporate activity in general is associated with firms’ size and industry. In contrast, whether the firm is listed on the NYSE (as opposed to NASDAQ) and the IPO timing only explain M&A activity, whereas the backing of VCs appears to drive alliance formation, but not acquisitions. Both models are highly significant overall, and the majority of the covariates are clearly significantly associated with their respective dependent variables, which indicates appropriate model specifications vis-à-vis our need to account for endogeneity concerns through the estimation of propensity scores.

Table 1 Logistic regression results for estimation of propensity scores a
Full size table

Tables 2 and 3 offer basic descriptive statistics and correlations across the covariates we use in the subsequent analyses. In the alliance data, the mean firm market capitalization was around $1.5 billion, and roughly 37 percent of the firms were in the broad high technology sector. These statistics are somewhat similar in the M&A dataset, although the average level of pre-IPO activity is lower (i.e., 48 percent), as acquisitions tend to be more resource-intensive and might be out of reach for some IPO firms. Moreover, market capitalization was slightly higher (i.e., $1.8 billion), as was participation in the broad high-tech sector (i.e., 54 percent). There are also noteworthy differences between the two samples: For instance, only 31 percent of the alliance-active firms were listed on the NYSE, as opposed to 65 percent of the M&A-active firms.

Table 2 Descriptive statistics and correlation matrix for alliance matched sample a
Full size table
Table 3 Descriptive statistics and correlation matrix for M&A matched sample a
Full size table

Turning to the correlation coefficients, in the case of alliances the use of the international and corporate dictionaries by executives is, perhaps surprisingly, only modestly correlated with the use of the same by analysts. This stands in contrast with the M&A table, in which firm activity in this area tended to result in greater inquires by analysts, as well as discussion by executives. Together with the far greater use of M&A-specific corporate dictionaries (as opposed to alliance-specific dictionaries) by both analysts and executives, this difference highlights the greater prominence of M&A as a topic for information disclosure during conference calls. While we do not test this idea explicitly, it is likely that the more consequential nature of M&A vis-à-vis the more flexible and reversible nature of strategic alliances compels analysts to inquire and executives to explain the rationale for M&A activity much more so than they do for alliances.

We now turn our attention to the multivariate models. Tables 4 and 5 focus on the textual analyses of the words spoken by executives, where the former tracks the use of the international dictionary words, and the latter tracks the use of the corporate words. We predicted that international corporate activities would result in greater information disclosure by executives during earnings conference calls. We find strong support for this prediction both in the case of strategic alliances and M&A. Specifically, firms active in international alliances see an increase of the words used in the international dictionary, as compared to inactive firms (i.e., p < 0.0001). Effectively, this means that of the total words spoken by executives during an earnings call, well over one percent more belong in the narrow and specific dictionary capturing international topics by the active firms. It is important to stress that even though the percentage increase we discuss may appear to be small, it captures the usage of a very narrow and specialized set of keywords relative to all words, so even a small increase represents a significant shift in the direction of the conversation undertaken during earnings calls. This difference is therefore highly significant not only statistically, but also economically. Moreover, cross-border M&A activity also sees an increase of words in the international dictionary (i.e., p = 0.0012).

Table 4 Tobit results for executives’ international dictionary a
Full size table
Table 5 Tobit results for executives’ corporate dictionary a
Full size table

In the models presented in Table 5, we split corporate activity between domestic and international deals. Regardless, it is apparent that whether international or domestic, M&A activity plays a significant role in information disclosure (i.e., p = 0.0002 and p < 0.0001, respectively), as do international alliances (i.e., p = 0.0003). In sum, we find strong support for H1a, H1b, H2a, and H2b.

Paralleling the analytical approach we undertook for executives, Tables 6 and 7 focus on analysts’ use of international and corporate dictionaries, respectively. With respect to the former, our results show a remarkable increase in the use of international keywords when the focal firm has engaged in cross-border alliances and acquisitions alike. Consistent with our predictions, alliances drive the use of these keywords by nearly 5 percent, while M&A by over 2 percent (i.e., p < 0.0001 and p = 0.0008, respectively). As before, the use of the corporate dictionary appears to be less affected by the activity level of firms engaged in alliances (Table 7). In fact, while M&A is still highly positively associated with analysts’ inquiries into the firm’s corporate strategy (i.e., p = 0.0001 for international activity and p < 0.0001 for domestic activity), we find no significant support for either international or domestic alliances. In sum, there is strong support for H3a, H3b, and H4b, but no support for H4a.

Table 6 Tobit results for analysts’ international dictionary a
Full size table
Table 7 Tobit results for analysts’ corporate dictionary a
Full size table

4.1 Additional Analyses

Although we needed to make decisions with respect to the measurement of variables and other aspects of the research design, we have examined the robustness of our findings through several supplemental analyses (results available from the authors). First, we initially set a maximum window of 5 years between firms’ IPOs and the last usable recorded earnings call and one might object that this window might be too long, as confounding events occurred after firms’ initial public offerings might affect corporate communication. Our goal was to have a short-enough window to minimize this concern, while at the same time have a sufficiently large number of data points to make our analysis meaningful. The implementation of our matching algorithm in particular is taxing in this sense. For example, limiting the time window to a single year post IPO results in a mere 39 observations available for the multivariate models for alliances (13 active and 26 matching inactive firms). The M&A data fares slightly better (91 total observations, with 32 active and 59 inactive firms), but it remains unsuited for model estimations.

Reducing the time window to 4 years was more sensible and it left us with 1,299 (334 active and 965 inactive) and 1,446 (370 active and 1,076 inactive) observations for the alliance and the M&A datasets, respectively. This alternate window resulted in very comparable results to the ones we reported previously. For example, the use of the international dictionary by executives was positively associated with a firm’s international alliance activity (i.e., p < 0.0001), and the use of the corporate dictionary was positively associated with both a firm’s domestic and international alliance activities (i.e., p = 0.0477 and p = 0.0008, respectively). Extremely comparable results apply to the M&A sample, as well, which shows that our findings are not extremely sensitive to the length of time we apply to track firms’ corporate activity.

We also performed a battery of supplementary tests attempting to gauge how sensitive our results are to the dictionary definitions we implemented. There can likely be no consensus on what may fully capture whether and how much a firm discusses its international or corporate agendas, and in the same way there is no consensus on any linguistic construct used in textual analysis in the literature. Our dictionaries represent our best effort to be on point in the list of keywords reflecting these items, and we saw little value in removing words from these parsimonious lists. In contrast, we experimented with additional words that may potentially proxy for international or corporate strategy concepts. Indeed, we ran a large number of alternate specifications and found that the results were uniformly strengthened by adding more terms in the respective dictionaries. Ultimately, we opted for more conservative specifications in the tables appearing in the previous section.

As a last portion of analysis, we wanted to gauge whether certain types of internationalization actions affect information disclosure more than others in a more fine-grained fashion than we do in the main analysis. Namely, we already separate alliances/JVs from M&A transactions, precisely because they are distinct in terms of resource deployment, governance mechanisms, strategic goals, etc. However, our data allowed us to break down corporate transactions more, and provide a more nuanced insight into our research question. Specifically, we separated non-equity alliances from equity JVs (instead of pooling them together), as well related from unrelated transactions (in both the alliance and the M&A space). Theory provides ample foundation for these distinctions, although, on more practical levels, we lost statistical power when constructing more refined matched samples as the quantity and quality of individual firms’ corporate actions was comparatively limited. Moreover, given our research design, the construction of multiple matched samples for every subgroup of corporate activity type would have brought about rapidly diminishing benefits, as the analyses in question caused us to estimate 16 separate new models. For this reason, we desisted from formally presenting these models, but instead provide a succinct summary of the findings.

First, it was interesting that the use of the international dictionary by executives was positively associated with non-equity alliances (i.e., p < 0.0001), unrelated alliances (regardless of type, p < 0.0001) and unrelated M&A (p = 0.0041). However, while showing a positive coefficient, equity joint ventures did not reach statistical significance in our model (p = 0.1637), although we suspect that the relative scant use of this specific corporate action might have something to do with this finding – the model was estimated from only 218 observations, as opposed to 1,488 observations for non-equity alliances, 1,058 for unrelated alliances of all types, and 942 observations for the unrelated M&A models. Turning to the use of the corporate dictionary by executives, the findings indicated that international activity triggers substantially more use of it on the alliance side (p < 0.001 for non-equity alliances and p = 0.0378 for JVs). On the other hand, unrelated M&A activity was clearly related to the use of the corporate dictionary, whether this activity had been conducted at the domestic or the international level (both p < 0.0001). Lastly, for their part analysts inquired about firms international activity across the board, while their questions involving the use of the corporate dictionary were far more associated with firms’ M&A activity than their alliance activity.

Overall, our results demonstrate significant associations between firms’ international corporate activities and information disclosures. This is true both in terms of the analysts’ probing executives on their international activities after going public, as well as the propensity by executives to elaborate at length on the same topics. We also show a marked increase in executives’ and analysts’ usage of corporate terms, when the firm engages in both international and domestic M&A, as all coefficients estimated are positive and highly significant by all standards. Lastly, we do not find strong evidence of the effects of firms alliance activity (domestic or international) on the use of corporate dictionary words during calls. We find this quite surprising, because while individual alliances may be less consequential than acquisitions, firms often build up portfolios of these cooperative agreements and their aggregate impact can be significant for firms’ performance.

5 Discussion

5.1 Contributions and Implications

This study contributes to the international strategy literature in several ways. At a broad level, we devote attention to capital markets as an important factor market that is related to firms’ internationalization. Financing considerations are integral parts of firm internationalization, as well as their corporate activities through alliances and M&A. As such, information transparency is important both for the operation of capital markets as well as firms wishing to obtain resources from these markets. Indeed, seminal research in IB has identified “capital availability and financial expertise” as an important ownership-specific advantage (Dunning, 1979: 280), yet current research adopting the OLI paradigm and other related perspectives of international expansion has emphasized the intangible resources (e.g., R&D) that firms obtain in factor markets other than financial markets. Consistent with this observation, Agmon (2006) points to a significant gap between the central role of finance in the practice of international business and the relatively minor emphasis given to capital markets in the literature. Our study shows that firms adjust their information disclosures to the capital market through their engagement with financial analysts as information intermediaries, and they do so based on the firms’ international corporate activities after going public. Our arguments and findings therefore represent one particular step in helping to narrow the gap regarding the role of capital markets in international strategy research, and we hope to encourage additional research on the links between product and capital markets as an important part of MNCs’ overall strategies.

We also build upon and extend research on firms’ external corporate activities that has considered information asymmetries that executives encounter in deals such as international alliances and acquisitions by suggesting that such deals also present information asymmetries to external investors. More specifically, a growing literature has suggested how executives might address information asymmetry and the risk of adverse selection it presents by relying on signals of exchange partners’ resources and prospects (e.g., Hsu, 2006; Nicholson et al., 2005). Alternatively, they might adopt other approaches such as staging investments, selecting targets presenting lower information asymmetries, avoiding sunk investments, or using contingent contracts (Balakrishnan & Koza, 1993; Capron & Shen, 2007; Gilson & Schwartz, 2005; Higgins & Rodriguez, 2006; Macias & Moeller, 2016; Schildt & Laamanen, 2006; Vanhaverbeke et al., 2002; Villalonga & McGahan, 2005).

We would emphasize instead that even if executives can or do address the risk of adverse selection in one or more of these ways, information asymmetries can exist for financial resource providers, and executives must contend with them in capital markets. These problems can be particularly salient when firms engage in internationalization and external corporate activities, which are fraught with the liability of foreignness and heightened information asymmetries. By examining executives’ information disclosures as well as questions by analysts, we offer behavioral evidence of firms coping with information asymmetries in capital markets as they expand internationally through alliances and acquisitions.

Our findings also complement prior studies that have focused on firm valuation effects of international investments and the accuracy of analysts’ earnings forecasts to gauge the effect of information asymmetries that arise from international diversification (Denis et al., 1997; Doukas & Kan, 2006; Duru & Reeb, 2002). While our focus has been on the antecedents of information disclosures by executives and questions by analysts, it would be valuable in future research to investigate the consequences of these information exchanges. For instance, research might build off of our initial study in this area by examining the extent to which firms are successful in addressing information asymmetries, situations in which less information disclosure might be advisable, and the implications that information disclosures carry for raising capital in the future as well as for growth and performance.

5.2 Limitations and Future Research Directions

Our study has several limitations worth noting, and these limitations suggest several additional directions for future research. To begin with, given our focus on US-based IPO firms engaging in internationalization, it would be valuable to examine potential bounds on the generalizability of results. In particular, beyond determining the degree to which our findings extend to more mature companies with histories of interacting with capital markets, future research could explore firms in other countries that have different financial institutions, types of capitalism, and ownership structures. In countries emphasizing debt financing or government ownership of companies, for instance, it would be interesting to examine how firms convey information on their international strategies to lenders or governments as financial resource providers. For firms that operate in weaker institutional environments and cross-list in other countries, it would be interesting to examine their information disclosures to investors in foreign capital markets.

Our analysis has focused on international alliances and acquisitions in general, so it would be valuable to consider other means by which firms internationalize as well as interesting heterogeneity that exists in firms’ international activities. It would be valuable, for instance, to examine how executives and analysts engage each other as firms expand through exports or via greenfield operations in foreign markets. Our additional post hoc analysis provides some tentative insights in this regard. For instance, unlike non-equity alliances, executives’ usage of international words does not increase significantly when discussing equity JVs. Since equity JVs typically entail more resource commitment and greater operational control of the focal investment by the MNC than non-equity alliances, there is likely to be a greater need to disclose information concerning the financial, operational, and governance aspects of the JV, thereby diminishing the coverage of the international topic in the limited amount of time provided by conference calls. This may imply that heterogeneity in foreign entry mode could lead to differential information needs, and consequently differences in the patterns of information disclosure by the firm. It would also be interesting to consider alliances and acquisitions taking place in very different institutional contexts and markets. It is possible that a transaction occurring in a more distant country, or in a country with less developed institutions will result in different interactions between executives and analysts (e.g., Jandik & Kali, 2009). Research in aforementioned directions may provide further opportunities to integrate the insights from IB and corporate governance literature, shedding new light on not only how international activities may influence disclosure practices but how corporate governance may in turn affect internationalization strategy.

Although our textual analysis covers important strategic intents reflected in certain words, and the results are robust to alternative dictionary definitions for firms’ corporate activities, this method is potentially limited in its capacity to capture certain qualitative aspects of the information conveyed by the statements in conference calls. Our research might therefore be extended in a few directions that would be particularly fruitful. The present study has emphasized corporate strategy considerations, and it would be valuable to use our methodology to also consider communications revolving around competitive strategy issues. Researchers have suggested that word responses are important tools in the competitive engagements of rival firms (Gao et al., 2017). Thus, there are many opportunities to use data such as ours to consider how firms define their rivals, seek to follow their lead or preempt them in foreign markets, or engage in multimarket competition. Such work could also enrich our analyses by moving beyond word counts in international and corporate dictionaries by drawing on the literature on linguistic analysis of corporate disclosures to consider language metrics such as tone and obfuscation (Loughran & McDonald, 2016).

While information disclosures can reduce information asymmetries, such disclosures can also have important drawbacks that we have not considered (e.g., facilitating imitation by competitors), so it would be valuable to consider such downsides as well as examine situations under which firms might engage in obfuscation of their international strategies. For instance, the tendency to avoid proprietary costs may be more pronounced when the MNC operates in a competitive industry with rapid innovation cycles, or when it enters a host country market with an elevated level of rivalry. Thus, analyzing the balance between proprietary costs and the need for transparency under various industry or host country contingencies may add additional perspectives to our results.

We also note that earnings calls with analysts represent only one means by which firms disclose information, so it would be interesting to compare information disclosures through this channel with others available to the firm (e.g., social media, annual reports, press releases, patents, etc.). Given our focus on conference calls and executives’ engagement with financial analysts, future research might take up the broader question of how economic transactions in certain markets have a bearing with other factor markets besides the capital markets (e.g., how international alliances shape communication with employees, suppliers, or other stakeholders who have not been emphasized in prior research). Research in directions such as these holds the promise of more fully accounting for the information costs that attend internationalization and how MNCs address them.