Going digital: do actions speak louder than words?

ABSTRACTThis study investigates whether investors assign more value to actions or words when evaluating firms’ digitalization efforts, with a focus on publicly listed companies in emerging Chinese stock markets over the sample period from 2007 to 2021. We employ a text-based proxy – the frequency of digital-related keywords in annual reports – and a direct measure of digital investments. Our findings indicate that while the frequency of digital-related keywords (words) does not significantly impact firm value, digital investments (actions) are positively and significantly associated with Tobin’s Q. Robustness checks, using the two-stage least squares (2SLS) instrumental variable regression, Heckman two-step method, propensity score matching (PSM), firm-fixed effects, and difference-in-differences (DiD) analysis, strongly support the causal effects of digital investments on firm value. These results suggest that investors place higher value on firms’ actual digital investments over mere disclosures of digital efforts, reinforcing the principle that actions speak louder than words. Consequently, we recommend that corporate managers prioritize investments in digital intangible assets to enhance their firms’ market valuations.KEYWORDS: Firm digitalizationfirm valuetextual analysisdigital investmentsChinaJEL CLASSIFICATION: D22G32O32 Disclosure statementThe authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.Notes1 According to the 2022 ‘White Paper on Global Digital Economy’ published by the China Academy of Information and Communications Technology (CAICT), China’s digital economy was valued at US$7.1 trillion in 2021, placing it in second place globally. The United States topped the list with a digital economy worth US$15.3 trillion. Source: https://research.hktdc.com/en/article/MTI4OTE5MTYwMg (retrieved on 28 February 2023).2 Empirical studies in the fields of business, management, and information systems generally construct measures of digitalization using a Likert scale for digital-related questions in survey questionnaires (see Kraus et al. Citation2022). However, primary data collected from surveys or interviews are often limited by small sample sizes and cross-sectional data, which make them susceptible to sampling and self-report biases.3 We construct the composite digital economy development index at the province level using the entropy weight method. This index incorporates five indicators, namely financial inclusion, internet users, computer services and software employees, telecom business development level, and mobile phone users.4 Our selected instrument satisfies the relevance and exclusion conditions. The relevance condition is met as the availability of digital infrastructure in a province incentivizes firms to invest in digital technologies to remain competitive. In addition, the exclusion condition is satisfied as there is no empirical evidence to suggest that the development level of digital economy directly affects firm value. In Table 2, the Kleibergen−Paap rk LM statistic and Cragg−Donald Wald F statistic support the validity of the province-level digital economy development index as an instrument.5 For the analysis, TREAT is a dummy variable that takes a value of one for firms located in pilot cities (treatment group), and zero for those located elsewhere (control group). Similarly, PYEAR takes a value of one for the corresponding year in which a firm is located in a ‘Broadband China’ pilot city, and zero otherwise.Additional informationFundingThe publication is partially funded by the Faculty of Business & Economics, Universiti Malaya Special Publication Fund.