Studies on Modeling for of Credit Risk Valuation

点击翻译按钮获取中文摘要

Nishina has investigated the theoretical structure of credit risk analysis by reexamination of existing models. Along with the theoretical approach, he has explored the mechanism of bond rating as a typical application of credit risk analysis. Recognizing the strong necessity of improvement in risk models, he expects a potential power or factor structure approach in the determination of bond prices.Tabata investigated the topics on index funds and their properties in the framework of an investment decision making problem under uncertainty. The efficient genetic algorithm to design the index fund is developed and the new procedure to revise parameters included in the algorithm is suggested from the view point of a Bayesian approach when the stock prices are available sequentially.Ohnishi mainly examined an optimal stopping problem for a geometric Brownian motion with Poissonian jumps. Although it has been argued that so called smooth pasting technique (see Dixit (1993), and Dixit and Pi … More ndyck (1994)) is useful for such stochastic optimization problems, it seems that its mathematical validity is not sufficiently discussed so far. In this project, by taking a martingale approach, he showed that it is indeed mathematically valid under a set of some mild conditions on the parameters of the problem.Tanigawa (2000a) considered implications of defaults on the ex-post moral hazard problem of corporate managers, noting that different types of debt contracts place different disciplinary burdens on the managers. Tanigawa and Koie (1999) studied exercises of corporate convertible bonds may be attributed to investors' liquidity demand, which is typical in the case of financial difficulties. Securities transaction itself involves risk of not executed. Omura, Uno, and Tanigawa (2000) estimated execution probabilities of limit orders of stock in the Tokyo Stock Exchange, to see how large such execution risk is incurred in limit orders with a more favorable price the limit order provides.The data that is available at financial market is not an ideal one. The data is supposed to be well-behaved by theories established at fields of statistical and mathematical finance, In this project, Oya has derived some properties of statistics with missing data and without any information about missing data. It is reported that the usual statistics for panel data analysis have an upward bias in the case mentioned above. Less