C5 - Econometric ModelingReturn

The study brings a new approach to the problem of the currently low use of financial rehabilitation methods in Czech economic reality. We examine several aspects of this problem, but consider the main problem to be the fact that entrepreneurial subjects enter into insolvency proceedings with a very small amount of assets compared to their liabilities. In such a situation, however, willingness on the part of creditors to enable reorganisation cannot be assumed - especially given that attempts to revive such a business are in the majority of cases clearly mistaken. Insufficient debtor property is also reflected in the fact that creditors receive satisfaction only to an insufficient degree; in comparison to OECD countries, the Czech Republic most of the time significantly behind the OECD country average when comparing insolvency proceedings. One of the possible ways of improving the prognosis for using the financial rehabilitation principle and the prognosis for investors to achieve adequate fulfilment during insolvency proceedings would be to tighten present regulations regarding the possibilities of indebting a business. This is presently done by the definition of over-indebtedness, according to which the liabilities of an entrepreneurial subject may not exceed the value of his property.

Financial risk modeling, measuring, and managing are an inherent part of management in financial institutions. It is also an important step within the setting of optimal level of capital eligible to cover risk exposures. A significant portion of capital is usually assigned to cover the risk of unexpected changes in FX rates. FX rates (the returns) commonly exhibit significant skewness and relatively huge kurtosis. In this paper, we apply subordinated Lévy models coupled together by ordinary elliptical copula functions in order to estimate the FX rate risk of normalized portfolio. Selected models are applied in order to estimate the risk ex-post, as well as ex-ante. The models are also compared to the more standard assumption of the joint normal distribution. Although the results for both types of modeling are quite different and Lévy measure is ignored, suggested models deliver us improved risk estimation.