ASEAN-listed firms that deploy big-data analytics report stronger profitability and shareholder value, with measurable gains in forecasting and risk control; however, the evidence is correlational and may reflect selective adoption by already stronger firms.

Main Finding

The paper reports a strong positive association between corporate big data / advanced analytics adoption and financial performance among ASEAN-listed firms. High-adoption firms show materially higher profitability and earnings growth (average ROA 8.5%, ROE 15.2%, EPS growth 12%), and econometric tests indicate that increased analytics adoption is linked to statistically significant gains in ROA and ROE (≈ +0.45 pp ROA and +0.78 pp ROE per unit increase in an adoption score; p < 0.05). Qualitative interview evidence from CFOs and analysts supports the quantitative results: analytics improves forecasting, risk management, and investment allocation.

Note: the manuscript metadata at the top (title about cryptocurrency taxation) appears inconsistent with the article body, which focuses on big data analytics and corporate finance in ASEAN.

Key Points

• Scope and sample

  • Region: ASEAN capital markets, focusing on Indonesia, Malaysia, Singapore, Thailand, and the Philippines.
  • Sectors: banking, manufacturing, technology, and other listed firms.
  • Sampling: purposive selection of listed firms actively using or integrating big data.

• Quantitative results

  • High-adoption firms: mean ROA = 8.5%, ROE = 15.2%, EPS growth = 12%.
  • Regression (controls for firm size, industry, market cap): 1-unit rise in analytics adoption → +0.45% ROA and +0.78% ROE (both significant at p < 0.05).
  • Correlation between analytics adoption and EPS growth: Pearson r = 0.62 (p < 0.01).
  • Results robust across sectors and firm types; partial adoption also yields measurable benefits.

• Qualitative insights

  • Executives report improved market projection accuracy, faster decision-making, better scenario simulation, and enhanced investor transparency.
  • A Singaporean financial institution case illustrates integrated platforms combining market feeds, transaction logs, and macro indicators to refine capital allocation and risk management.

• Drivers and constraints

  • Drivers: data volume/velocity in ASEAN, ML/AI advances, localized tailoring of analytics to diverse markets.
  • Constraints: heterogeneous regulatory regimes, uneven technical infrastructure, weak data governance, skills shortages, and limited adoption among smaller firms.

• Policy/managerial recommendations

  • Firms: invest in analytics infrastructure and talent; embed analytics into core financial workflows.
  • Policymakers: incentivize data infrastructure projects, support workforce readiness in data science, and clarify regulatory frameworks for data use.

Data & Methods

• Research design: mixed-methods (quantitative firm-level analysis + qualitative semi-structured interviews).
• Quantitative data sources: audited annual reports, balance sheets, historical market data from publicly listed firms.
• Variables and outcomes: analytics adoption score (categorical/high–medium–low), ROA, ROE, EPS growth, market capitalization; controls for firm size, industry, market cap.
• Econometric approach: regression analysis with controls; correlation analysis (Pearson); robustness checks across sectors.
• Qualitative data: semi-structured interviews with CFOs, financial analysts, and data specialists; one in-depth case study of a Singaporean financial institution.
• Triangulation: synthesis of statistical results with thematic coding of interview transcripts to explain mechanisms.
• Limitations acknowledged by authors: purposive sampling (focuses on firms already adopting analytics), short-term outcome focus, need for longer-term causal inference, and unexplored barriers for smaller/less-resourced firms.

Implications for AI Economics

• Returns to AI/big-data investment: the paper provides empirical evidence that analytics/AI investments are associated with measurable firm-level gains (higher ROA/ROE and EPS growth). This supports models that treat data/AI as a productive capital input that raises firm productivity and profitability.
• Heterogeneous adoption and market structure: widespread productivity divergence can arise between high-adoption and low-adoption firms, implying increasing firm-level heterogeneity, market reallocation, and potentially greater concentration—key considerations for industrial organization models of AI.
• Labor and skills: demand for data-science talent and analytics capabilities will likely raise skill premia and reconfigure firm hiring; policymakers should consider retraining and education to reduce frictions.
• Information and asset pricing: richer firm information sets (real-time analytics, predictive models) may change volatility, expected returns, and risk premia (e.g., faster incorporation of signals into prices; altered tail-risk management). Asset-pricing and risk models should incorporate differences in firms’ information-processing capacities.
• Measurement and causal inference challenges: adoption scores may correlate with unobserved management quality or capital intensity. Future AI-economics work should use stronger identification (instruments, difference-in-differences, randomized pilots) to establish causal returns to AI/big-data investments.
• Policy design: evidence that public support for data infrastructure and workforce readiness can amplify aggregate benefits of AI suggests targeted subsidies, standards for data governance, and policies to reduce adoption barriers for smaller firms.
• Research directions: study long-run sustainability of returns, effects on competition and market concentration, distributional impacts across workers and firms, cross-country heterogeneity in regulatory effects, and rigorous causal evaluation of AI deployments.

If you want, I can (a) extract and format the regression table and quantitative results into a concise table, (b) draft possible identification strategies to estimate causal effects of analytics adoption for follow-up research, or (c) produce a short slide-ready summary for presentations. Which would be most useful?