Abstract
Objective – to substantiate and test methodological tools for validating the economic efficiency of grain agricultural enterprises, combining coefficient monitoring of indicators with non-parametric frontier methods, allowing for the identification of unrealized production reserves. Methods – comparative analysis was applied to systematize mechanisms for comparison according to uniform criteria; Spearman’s rank correlation – when comparing index parameters with a threshold characteristic; Data Envelopment Analysis with an output-oriented BCC model was used on contemporary and intertemporal frontiers to determine production resources; scale efficiency decomposition – in establishing optimal size; the Malmquist index – to separate technological and managerial components; robustness checks were performed on a reduced sample. Results – diagnostics of official data for 12 grain-growing regions for 2021–2025 showed an average technical efficiency of 0.870 on the contemporary frontier and 0.654 on the intertemporal frontier. In 2025, average productivity was 0.809, meaning output could be increased by 23.6 % without additional funds. The scale effect was 0.824: nine out of twelve districts perform tasks outside the optimal range. The average annual value of the Malmquist index reached 1.190, with a predominance of the technological component (1.163 versus 1.023); however, it requires cautious interpretation due to weather volatility. The exception is the Kyzylorda Region, which raises the rating to 0.935 without altering the outcome. Rank correlation of the frontier assessment with yield is at 0.561, resource return – 0.381, cost price – 0.336, and profitability – 0.331. Conclusions – coefficient indicators reproduce approximately 11–31 % of the joint rank variation of boundary measurements, proving their complementary rather than alternative application.