ECONOMETRIC DETERMINANTS OF THE COST LEVEL IN GRAIN-PROCESSING ENTERPRISES: PANEL EVIDENCE FROM UZBEKISTAN

Abstract

This study identifies the financial and economic determinants of the cost-to-revenue ratio in industrial enterprises using a novel firm-level panel dataset from two grain-processing joint-stock companies in Uzbekistan: Galla Alteg JSC and Uchqorgon Don Mahsulotlari JSC (2021–2025, N=10). Based on official balance sheets and income statements, five candidate explanatory variables were constructed: revenue scale, material intensity, inventory turnover, the share of period expenses, and the share of fixed assets in total assets. Before estimation, a pooled Levin–Lin–Chu-type panel unitroot test was applied. Four alternative specifications—pooled ordinary least squares (OLS), fixed effects (LSDV), partial least squares (PLS) regression, and a parsimonious panel ARDL(1,0) model—were subsequently compared using standard goodness-of-fit and diagnostic criteria (VIF, Breusch–Pagan, Breusch–Godfrey, Jarque–Bera, and CUSUM). The fixed-effects specification incorporating inventory turnover and the fixed-asset share as regressors was selected as the final model. The adjusted \(R^2\) for the full specification was 0.759, while the \(R^2\) for the parsimonious final model was 0.498. The model passed all diagnostic tests and showed no evidence of structural instability. The results indicate that a 1% increase in inventory turnover reduces the cost level by approximately 0.022 percentage points, whereas a one-percentage-point increase in the share of fixed assets in total assets raises it by approximately 0.413 percentage points. These findings provide two quantitative risk indicators that can be incorporated into cost-accounting audit methodology. A one-year-ahead forecast is also presented for both enterprises.