Smart farming as a market stabilization strategy: heterogeneous policy effects and business implications from South Korea

This study evaluates whether government-led smart farming initiatives can serve as an effective market stabilization strategy for agricultural businesses, examining South Korea’s comprehensive 2014 smart farming policy that invested in controlled-environment agriculture. We employ a novel dual-method framework combining ARIMA-TGARCH models with machine learning-enhanced propensity score matching (ML-PSM) to analyze weekly price data (2005–2024) for three high-value greenhouse crops: strawberries, cucumbers, and lettuce. This approach captures both structural breaks in volatility patterns and heterogeneous treatment effects across different subsidy intensities, seasons, and crops. Smart farming subsidies reduced price volatility by up to 28.7% for strawberries and 18.0% for cucumbers, while lettuce showed minimal response. For agribusiness managers, findings indicate that concentrated investments in responsive crops yield superior returns compared to diversification strategies. The identified optimal investment thresholds and seasonal patterns provide concrete guidance for market entry timing, facility sizing, and capital allocation decisions. For policymakers, crop-specific support programs would enhance efficiency by 40–50% compared to uniform subsidies. This research provides the first rigorous evidence of smart farming’s heterogeneous effects on agricultural market stability, offering a replicable framework for policy evaluation. The integration of machine learning with traditional econometrics reveals hidden patterns critical for strategic decision-making, while the findings challenge conventional approaches to agricultural risk management and technology adoption.