Font Size: a A A

Study On Loan Pricing Based On Credit Rating And Default Probability

Posted on:2006-06-03Degree:DoctorType:Dissertation
Country:ChinaCandidate:D M JiangFull Text:PDF
GTID:1119360182975518Subject:Financial engineering and financial management
Abstract/Summary:PDF Full Text Request
Credit is one of the key businesses of commercial banks. How to develop anaccurate pricing model and to price the loans properly is an important problem thatcommercial banks have been facing for many years in their business management.Based on some conditions, in this paper, we develop some models on loan'sbreak-even in terms of no-arbitrage pricing theory and the reduced-form pricingapproach of credit risk. The main ideas and conclusions are summarized as following:Since credit risk is the most dominant factor that affect the price of loan, we firstly,in chapter 2nd, analysis the characteristics of credit risk, and conclude that credit riskis a non-system risk, so the models of CAPM and APT which are both based on thetheory portfolio are not suitable for loan pricing. And then, we review the twoapproaches of credit risk pricing: structural approach and reduced-form approach. Inchapter 3rd, we introduce the preliminaries of this paper: the theory of no-arbitragepricing and the Markov model of credit rating. In chapter 3rd, also, some conditionsand hypothesizes which will be utilized in the next two chapters are put forward to inadvance.In chapter 4th, based on the risk-neutral default probability or the default intensityof the firms, we develop some loan pricing models, in which the cash flows of theloan may be single-stage or multi-stages. Meanwhile, the models reveal therelationship among the cash flows which occur in different time. It implicate that themodels overthrow the ideas that a loan with multi-stage cash flows can be regard asmulti independent loans with single-stage cash flow. Moreover, the relationshipsamong default distribution function and distribution density and default intensity havebeen demonstrated. In this chapter, we also discuss loan pricing models in two kind ofcases that the default process is a time-homogeneous or a time-nohomogeneousPoisson process.In chapter 5th, by utilizing the Markov models of credit transition probability matrix,we study the problem of forward loan pricing. At first, we convert thetime-homogeneous credit matrix into an time-nohomogeneous one, therein, the formeris under the natural probability, and the later is under the equivalent martingalemeasure. And then, some models of forward loan pricing and a formula of expectationprice are developed.In the last chapter, we demonstrate 4 examples to verify the models mentioned inthe above chapters. It is shown that all the conclusions of the examples are consistentwith the practice. Namely, it implicate that the models we developed are logical.
Keywords/Search Tags:Credit risk, Loan pricing, Default probability, Default intensity, Credit transition matrix, Structural approach, Reduced-form approach.
PDF Full Text Request
Related items