Work

Artificial Intelligence and Corporate Financial Decision-making: Evidence on Forecasting Accuracy, Capital Allocation, and Firm Performance

Year: 2026
Type: article
Abstract: Artificial intelligence (AI) is changing corporate financial decision-making by enabling firms to process large and diverse datasets, identify nonlinear patterns, and update forecasts more frequently than conventional analytical approaches. This paper examines the ways in which AI may influence three connected dimensions of corporate finance: forecasting accuracy, capital allocation, and firm performance. It develops an integrated framework in which AI-enabled analytics improve the timeliness and informational content of financial estimates, support the comparison of competing investment opportunities, and potentially strengthen operating and financial outcomes. The framework also recognizes that these benefits are conditional rather than automatic. Data quality, model transparency, organizational capabilities, managerial judgment, and governance arrangements can shape whether analytical outputs are converted into sound decisions. In forecasting, machine-learning models may capture complex relationships among sales, costs, liquidity, demand, and macroeconomic indicators, but their apparent accuracy can weaken when economic conditions change or when training data do not represent future environments. In capital allocation, predictive systems can support project screening, scenario analysis, working-capital management, and risk-adjusted investment appraisal; however, optimization based on incomplete objectives may amplify short-term incentives or overlook strategic and ethical considerations. At the firm level, the relationship between AI adoption and performance is therefore likely to operate through intermediate mechanisms, including forecast quality, resource-allocation discipline, decision speed, and risk control. The paper proposes an empirical approach using firm-year observations, measures of AI adoption and analytical intensity, forecast-error indicators, investment-efficiency measures, and accounting-based performance outcomes.
Language: English
Open Access status: hybrid
OpenAlex ID: W7225254377
FWCI:
Cites: 14
Cited by:
Related to: 10