Mathematical and Statistical Methods for Actuarial Sciences and Finance

(Nora) #1

286 F. Quittard-Pinon and R. Randrianarivony


price processSis supposed to behaveaccording to the following SDEunder the
chosen equivalent pricing measureQ:


dSt
St−

=

(

rt−

)

dt+ρσdWt+σ


1 −ρ^2 dZt+(Y− 1 )dN ̃t. (9)

Again,rtis the instantaneous interest rate,represents the fixed proportional insur-
ance risk charge,σis the asset’s volatility,ρis the correlation between the asset and
the interest rate,WandZare two independent standard Brownian motions, and the
last part takes into account the jumps.N ̃is a compensated Poisson process with in-
tensityλ, whileY, a r.v. independent from the former processes, represents the price
change after a jump. The jump size is defined byJ=ln(Y).
Let us emphasise here that the non-drift partM,definedbydMt=ρσdWt+
σ



1 −ρ^2 dZt+(Y− 1 )dN ̃t, is a martingale in the considered risk-neutral universe.

3.1 Modelling stochastic interest rates and subaccount jumps


Denoting byNtthe Poisson process with intensityλand applying It ̄o’s lemma, the
dynamics ofSwrites as:


St=S 0 e

∫t
0 rsds−(+
1
2 σ^2 +λκ)t+ρσWt+σ


1 −ρ^2 Zt+
∑Nt
i= 1

ln

(

(Y)i

)

, (10)

whereκ=E(Y− 1 ). The zero-coupon bond price obeys the following equation:


P(t,T)=P( 0 ,T)e

∫t
0 σP(s,T)dWs−^12

∫t
0 σ^2 P(s,T)ds+

∫t
0 rsds.

The subaccount dynamics can be written as:


St=

S 0

P( 0 ,t)

e

−(+^12 σ^2 +λκ)t+ 21
∫t
0 σP^2 (s,t)ds+

∫t
0 [ρσ−σP(s,t)]dWs+σ


1 −ρ^2 Zt+
∑Nt
i= 1

ln

(

(Y)i

)

.

Let us introduce theT-forward measureQTdefined by
dQT
dQ



Ft=

δtP(t,T)
P( 0 ,T)

=e

∫t
0 σP(s,T)dWs−^12

∫t
0 σ^2 P(s,T)ds, (11)

whereδt is the discount factor defined in (1). Girsanov’s theorem states that the
stochastic processWT,definedbyWtT=Wt−


∫t
0 σP(s,T)ds, is a standard Brownian
motion underQT. Hence, the subaccount price process can be derived under theT-
forward measure:


St=

S 0

P( 0 ,t)
eXt, (12)

whereXis the process defined by


Xt=−(+^12 σ^2 +λκ)t+

∫t

0

(

σP(s,T)

(

ρσ−σP(s,t)

)

+^12 σP^2 (s,t)

)

ds

+

∫t

0

(

ρσ−σP(s,t)

)

dWsT+σ


1 −ρ^2 Zt+

∑Nt

i= 1

ln

(

(Y)i

)

.

(13)
Free download pdf