A Firm Level Robust Credit Rating Migration Modeling Framework : A Small Sample Methodology

Abstract

Crisis taught us that the capital markets are incomplete, with hidden information, which puts limits to arbitrage and thus controversial aspect of “efficient prices” prompts the valuation scientists to investigate deeper with more robust methodological investigations. In context, did our financial statements reveal something extraordinary, which allowed our “credit raters” to apply for robust measure of rating calibration ? The present paper proposed a “destabilizing” framework which can bring some benefit (if jointly there are lot of disturbances) in the flexible dynamic system. The dynamic system is a representation of the unknown latent factor. The human capital price as factor price was considered as an unknown factor. The model used here, utilized in a disintegrated fashion, two separate latent factor pricing models. The two latent pricing models (namely market-based and internal-information based) with respect to conditional volatility of latent prices (prices derived from single-variate market-based and internal-information based) showed marked differences, and it was observed that for sample companies under robust setup, the internal information based model yielded better credit rating annual forecast compared to the market-price based model.