This paper investigates the possible nexus between the 'risk-return paradox' and corporate-governance of firms in a cross-country cross-cultural setup. We use corporate governance as well as accounting risk and return data for a large dataset of 45,322 firm-years from 27 countries and show that the firm-level risk-return association may be a non-linear one, contingent on the firm performance. Firms which are below the industry median in terms of operating performance, exhibit an inverse relation in line with Bowman's (1980) ‘paradox' while those above-median exhibit a positive risk-return association. Further, we establish empirically that such risk-return association could be due to the rent-seeking actions of managers and that strong corporate-governance in a firm substantially moderates and reverses these effects. Our results are robust and hold strong through a number of robustness tests.
Risk-return association in financial markets is typically expected to be positive, i.e., bearing higher risk is justified when expected returns are higher (Sharpe, 1964; Ghysels et al., 2005). This risk-return trade-off should also apply to firm-level strategic decisions that involve risks such as entering new markets, launching new products, and exploring new ways to prune costs. Therefore, to maximize shareholders’ wealth, managers may possibly undertake high-risk investments if they are expected to yield commensurate higher returns for shareholders (Chari et al., 2019). Thus, prior knowledge of such a risk-return association is critical from the perspective of firms and their stakeholders, especially managers. Our study draws its primary motivation from this premise.
Bowman (1980) reports an interesting risk-return ‘paradox,’ or a negative correlation between accounting risk and return. This observation is counterintuitive to the generally accepted theory in traditional finance, as highlighted above (Sharpe, 1964; Ghysels et al., 2005). Since then, a series of empirical studies have presented possible theoretical and empirical explanations for this. ‘Prospect theory’ (Kahneman and Tversky, 1979), ‘behavioural theory’ (Cyert and March, 1963; Bromiley, 1991), ‘agency theory’ including ‘value-reducing risk-taking’ (Andersen et al., 2007; Chari et al., 2019), and ‘career concerns’ of managers (Dewatripont et al., 1999), to name a few. However, most of these studies draw their inferences based on US data. Whether Bowman’s ‘paradox’ is generalizable across different countries with wide inherent variability, to the best of our knowledge, has yet to be explored extensively. A few recent studies, such as Patel et al. (2018) and Dasgupta and Deb (2020), have used data from non-US countries to check the validity of the paradox. However, many potential dimensions of inherent variability, such as stages of economic growth, market cycles, legal systems, investor-protection rights, national culture, and various other firm-level heterogeneity are not controlled for in those studies. The current study draws its secondary motivation from this premise
Summary and conclusion
This study explores the previously reported ‘risk-return paradox’ and examines whether the risk return relationship is contingent on measures of risk and return used, firm- or country-level heterogeneity, and market cycles. The study draws its primary motivation from the premise that prior knowledge of an established risk-return association is critical from the perspective of international firms. Such knowledge could help in shaping strategic decision-making processes involving risk or uncertainty, such as entering new markets, launching new products, and exploring new ways to prune costs. Managers need to undertake high-risk investments only if they yield commensurately higher returns for shareholders (Chari et al., 2019). We use accounting risk and return data from a sample of 45,322 firm-years from 27 countries, representing emerging and developed economies. We control for a large number of potential factors that may influence firm-level risk and return association. Using several univariate, multivariate, and robustness tests, we find evidence of a partial paradox in the risk-return association, which is contingent on the current level of firm performance. Firms that are below the median, in terms of performance, exhibit a negative (i.e., paradoxical) risk-return association primarily, while the firms that are above the median exhibit mostly a regular or positive risk-return relationship. These findings are in line with some other studies, a majority of which were in the context of US markets (Lehner, 2000; Becerra and Markarian, 2013; Holder et al., 2016).