Tržni udeleženci lahko investirajo v kratkoročne ali dolgoročne obveznice. Slednje so bolj izpostavljene različnim vrstam tržnih tveganj, za kar tveganju nenaklonjeni investitorji zahtevajo nadomestilo v obliki premije za tveganje. Nominalno donosnost obveznice lahko torej razčlenimo na dve komponenti: do tveganja nevtralno donosnost in terminsko premijo. Komponenti nista neposredno opazljivi, zato ju moramo oceniti z uporabo ekonometričnih modelov. V magistrskem delu natančno predstavimo matematično izpeljavo ACM modela in ga uporabimo za modeliranje terminske strukture nemških državnih obveznic. Ovrednotimo in primerjamo uspešnost različnih specifikacij osnovnega modela znotraj in izven vzorca podatkov. Analiza znotraj vzorca podatkov zajema aproksimacijo in dekompozicijo terminske strukture. Gibanje posameznih komponent skozi čas razložimo z makroekonomskega vidika. V okviru analize izven vzorca podatkov primerjamo napovedi kratkoročnih obrestnih mer za tri leta vnaprej.
Language: | English |
---|
Title: | The modelling of the term structure of German Government Bonds by using the ACM Model |
---|
Abstract: |
---|
Market agents can invest in short- or long-term bonds. The latter are more exposed to various types of market risks. This is why the risk-averse investors demand a compensation in a form of risk premium.
A nominal yield can be decomposed into two components: risk-neutral yield and term premium. None of the components is directly observable, therefore they need to be estimated by using econometric models. The present thesis covers a detailed mathematical derivation of the ACM model and its application to the modelling of the term structure of German government bonds. It assesses and compares the in- and out-of-sample performance of the model's variations. In-sample analysis includes approximation and decomposition of the term structure of interest rates. The dynamics of both components over time is interpreted from the macroeconomic point of view. Out-of-sample analysis generates and compares short-rate predictions up to three years ahead.
|
Keywords: | yield curve, risk-free interest rate, expectation theory, term structure, term premium, affine term structure model, pricing kernel, market price of risk, pricing factor, linear regression, principal component analysis |
---|