USING MULTIPLE DISCRIMINANT ANALYSIS IN THE BANKRUPTCY PREDICTION IN ALBANIA – A STUDY WITH THE STATE-OWNED ENTERPRISES

This paper applies the Multiple Discriminant Analysis technique to study the bankruptcy of the state-owned enterprises in Albania. The discriminant function derived by this technique had an overall accuracy rate by 94.6 percent when tested on the initial sample and 92.9 percent if tested using the cross-validation method. The variables that best discriminated between bankrupt and non-bankrupt firms were the level of operating profitability and size of investments. Liquidity or cash flow variables often cited as important predictors of bankruptcy in other previous studies, did not result important. We conclude that the economic profitability and good investment opportunities are the main factors that affect the success of the state-owned enterprises in Albania. We argue that often liquidity problems arise quite shortly before the bankruptcy filing, thus it does not allow for inclusion among the early predictors of bankruptcy.