On the Predictability of Kibbutz Financial Distress: A Principal Component Analysis with Bootstrap Confidence Intervals

Abstract

This paper proposes an econometric model for conditional forecasting of kibbutz financial distress for a range of one to three years. Our sample covers financial and demographic data for about 32 kibbutzim over a seven year period (1990-1996). Data on an additional six kibbutzim over the same period serve as a validation sample. The primary interest in examination of the economic performance of the kibbutz is the fact that its driving characteristic is collective rather than individual profit-maximizing. This study is also highly motivated by the kibbutz debt crisis exposed during the mid-1980's. A major task was the adjustment of information found in kibbutz financial statements into economically meaningful variables. Our data set contains 26 explanatory variables defined over 8 different dimensions: liquidity, profitability, cash flow, operating efficiency and growth, capital structure, saving rate, demography, and size. We refer to the annual change in productive equity, scaled by either productive assets or annual sales, as representing developments in kibbutz financial confidence. Multicollinearity is coped with by the employment of a principal component regression. Goodness of fit is evaluated by examination of various dispersion statistics and by comparing the model's performance with that of a 'naive forecast', maintaining that performance measures follow a random walk. The asymptotic behavior of the results is assessed via application of a pair bootstrap simulation. The model as a whole exhibits high statistical significance, efficiency of forecast, and a good fit. It outperforms the naive forecast model in all calculated measures of dispersion. This outcome is reaffirmed by the bootstrap simulation. Our findings suggest that financially distressed kibbutzim are less profitable, tend to be highly leveraged, and are characterized by low and even negative saving rates, and a higher social burden. We also find empirical support for the emergency cash reserve, the optimal capital structure, and the heterogeneity hypotheses.

JEL Classification: G33, P13, C53