Ketepatan Pemeringkatan Obligasi Antara Model Prediksi dan Agen Pemeringkat

Abstract

Bond rating which is generally released by rating agency has to be paid attention by enthusiastic investors who buy the obligation, because the rating gives the informative statement and gives the signal about probability failure owned by company. Research target is to evaluate and give the empirical evidence so meet the accuracy of bond rating by agent PEFINDO, by testing ability prediction model using accounting information that is finance ratio. By using sample as much 36 companies of non finance or 168 observation since year 2000 until 2002, the result from analysis of Multiple Discriminant Analysis (MDA) by statistic give chapter and verse that fifth of finance ratio raised by a researcher that is ratio leverage (LEVLTLTA), liquidity (LIKCAICL), solvency (SOLCFOTL),profitability (PROFOIS) and productivity (PRODSFA) own the ability in forming model for the prediction of bond rating. Meanwhile the results of analysis of Chi Square gives chapter and verse that bond rating tye prediction model of based on finance ratio by statistic differ and better than bond rating conducted by agent PEFINDO. Thereby prediction model conducted by a the researcher can replace the rating conducted by agent PEFINDO.