Purpose: The purpose of this paper is to identify the determinants of dividend policy in an emerging and developing market.Design/methodology/approach: The study employs a quantitative approach using 191 Sri Lankan firms and 1,337 firm-year observations as the sample. The authors apply a Binary Logistic Regression model to uncover the determinants of the propensity to pay dividends, and a Fixed Effect Panel Regression to investigate the determinants of dividend payout.Findings: The authors identify past dividend decision, earnings, investment opportunities, profitability, free cash flow (FCF), corporate governance, state ownership, firm size and industry influence as the key determinants of propensity to pay dividends. In addition past dividends, investment opportunities, profitability and dividend premium are identified as the determinants of dividend payout. Moreover, there is a feedback between dividend yield and profitability in one lag and between dividend yield and dividend premium in two lags, as short-term relationships. Hence, past dividend decision or payout, profitability and investment opportunities are a common set of determinants with implications for both propensity to pay dividends and its payout. The findings support theories of dividends such as signaling, outcome, catering, life cycle, FCF and pecking order.Practical implications: The findings are important for investors, managers and future research. Investors should focus on the determinants identified by our study when making investment decisions whereas managers should practice the same when formulating appropriate dividend policies for their firms. Future research should rely on propensity to pay dividends and its payout simultaneously to promote a theoretical consensus on the dividend determinant puzzle.Originality/value: This is the first study that investigates determinants of propensity to pay dividends and dividend payout along with short-term relationships in a single study.

This paper uses machine learning to improve the prediction of corporate emissions so that financial regulators and investors can make better decisions about climate transition risk. The need for predictions arises because only a subset of global companies report emissions. The novelty is to use machine learning rather than the conventional regression approaches and naïve models implemented by data providers. Our best-performing model is a two-step framework that applies a Meta-Elastic Net learner to combine predictions from multiple base-learners. It results in an accuracy gain based on mean absolute error of up to 30% as compared with the existing models. We find that prediction accuracy can be further improved by incorporating additional predictors (energy data) and additional firm disclosures in particular sectors (utilities and real estate) and regions (Asia and Latin America). This provides an indication of where policymakers should concentrate their efforts for greater disclosure.

Using bank-issued green bonds as a proxy for banks’ green financing activity, this study examines how green bond issuance affects bank performance through the CAMEL framework. Drawing on an international panel of more than 1,100 banks across 99 countries over 2010–2023, we estimate bank fixed-effects models, two-stage least squares specifications, and entropy-balanced regressions to assess whether green bonds matter for bank performance beyond their signalling value. We show that, among bond-issuing banks, green bond issuance is associated with improvements in selected CAMEL dimensions, and that these effects are stronger when issuance intensity is higher. Specifically, banks with greater green bond issuance exhibit stronger asset-based profitability, lower risk, and a shift in income structure away from traditional interest-based intermediation toward non-interest income. We further show that green bonds operate through two economically meaningful channels. First, green bond issuance is associated with greater participation in green loan deals, consistent with a balance-sheet greening effect that improves asset quality and capital resilience. Second, green bond issuance reduces banks’ cost of debt, indicating that green bonds can improve funding conditions and thereby support financial performance. Overall, the evidence suggests that bank-issued green bonds are not merely sustainability signals; they are balance-sheet instruments that help banks reallocate assets, strengthen resilience, and adapt their intermediation model in the climate transition.

We re-examine the market efficiency of commodity futures using a new approach that accounts for both time-varying risk premium and conditional heteroscedasticity of spot prices. The conventional market efficiency tests so far in the literature are based on either risk neutral or constant risk premium assumptions as such they are biased towards the rejection of the market efficiency hypothesis especially for commodity futures. The time varying risk premium is estimated using a state space model with a modified Kalman filter. Using a Monte Carlo simulation, we show that the proposed test produces robust and superior results under varying market conditions compared to the conventional approaches. By employing the proposed test we analyse the efficiency of crude oil, corn, copper and gold futures and find that gold futures is inefficient throughout the sample period 2000-2011 while others are efficient especially after the global financial crisis (GFC) in 2008. We also find significant changes in the underlying risk premiums due to the GFC. We extend the analysis to a comprehensive sample of 85 commodities traded on 16 exchanges worldwide and find that efficiency and risk premiums vary across the four market sectors while GFC has caused to increase both efficiency and risk premiums in all markets other than precious metals.