Master Thesis - Humboldt-Universität zu Berlin

E t
[
β λ t+1
λ t
]
RtP e t
C
= 1 (32)
Pt+1
C
E t
[
β λ t+1
λ t
S t
S t+1
R e∗
t Pt
C
Pt+1
C
]
= 1 (33)
R t ɛ S t
R ∗ t
= S t
S t+1
(34)
Making abstraction of the term ɛ S t we obtain the uncovered interest rate
parity. However, it is well known from empirical studies that UIP condition
is strongly rejected by the data. One of the reasons for this is the imperfect
integration of the financial markets. We therefore add the term ɛ S t to the left
hand side of the last equation.
The benchmark model de Walque et al.(2005) only includes the shock in
the numerator of real effective exchange rate on foreign bonds, while the
Dynare code suggests the placement of the shock in the denominator. The
inconsistency between the theoretical model as regards to the sign of the
stochastic component of the risk premium and the Dynare code provided by
the authors prompted us to proceed for further investigations of the nature
of this premium. Our first intuition was that, in order to increase the return
on foreign bonds it should be placed similarly to the risk premium on bond
holdings, namely at the denominator in the last equation. In order to support
to our reasoning, we consulted the theoretical set up and impulse-response
graphs in Bergin(2006) and Linde at al(2005). They model the risk premium
as a function strictly decreasing in real aggregate net foreign asset position
of the domestic econoy b ∗ t = 1
R e∗
t
B τ∗
t
P C t St :
RP (b ∗ t , ɛ S t ) = exp(−RP · b ∗ t + ɛ S t ) (35)
20