Providing Optimal Model to Predict Financial Distress Using Simulated Annealing Algorithm

Providing Optimal Model to Predict Financial Distress Using Simulated Annealing Algorithm

Gholamhossein Mahdavi1 Sara Ziyari2

1) Professor of Accounting
2) M.A. in Accounting

IOI Code : XYGC-BCDBK
Publication : 3rd International Conference on Management and Humanities(3icmhconf.com)
Abstract :
    This study measures the applicability of simulated annealing (SA) algorithm in financial issues and present an optimal model to predict the financial distress of the companies listed on the Tehran Stock Exchange. The efficiency of SA algorithm is evaluated based on two main dimensions. First, the accuracy of simulated annealing algorithm in predicting financial distress correctly compared with Altman’s Z model and second, the ability of this algorithm in providing optimal model in an acceptable time was evaluated. The result indicates that it is possible to provide a flexible optimal model by using SA algorithm in a reasonable computational time. As total accuracy of financial distress models based on SA and Altman’s Z model in the event year, respectively, equals to 98.1% and 79.6%, in the year prior to the event year, respectively, equals to 84.3% and 71.3%, and in two years prior to the event year, respectively, equals to 69.4% and 60.2%. Although the accuracy of the metaheuristic model decreases as we get far from the event year, yet it would have better predictable power compared with the Altman’s Z traditional model.

Keywords : prediction financial distress simulated annealing algorithm altman’s z model