Etude des marchés d'assurance non-vie à l'aide d'équilibres de ...
Etude des marchés d'assurance non-vie à l'aide d'équilibres de ...
Etude des marchés d'assurance non-vie à l'aide d'équilibres de ...
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
tel-00703797, version 2 - 7 Jun 2012<br />
Equation (4.9)<br />
H(x, y) = C(FX(x), FY (y)),<br />
4.4. Focus on the <strong>de</strong>pen<strong>de</strong>nce structure<br />
when X, Y take discrete values. The answer is given by the Carley bounds C −<br />
H , C+<br />
H . For all<br />
copulas CH verifying Equation (4.9), we have<br />
C −<br />
H ≤ CH ≤ C +<br />
H ,<br />
cf. Proposition 2 of Genest and Neslehova (2007).<br />
The <strong>non</strong>-i<strong>de</strong>ntifiability issue matters, since with discrete copulas the <strong>de</strong>pen<strong>de</strong>nce measure,<br />
such as the tau of Kendall or the rho of Spearman, are no longer unique. Furthermore, if X<br />
and Y are in<strong>de</strong>pen<strong>de</strong>nt, it does not imply that the copula of (X, Y ) is the in<strong>de</strong>pen<strong>de</strong>nt copula.<br />
In other words, the copula alone does not characterize the <strong>de</strong>pen<strong>de</strong>nce: we need assumptions<br />
on margins.<br />
In the copula literature, efforts have been done to tackle the <strong>non</strong>-i<strong>de</strong>ntifiability issue of<br />
<strong>de</strong>pen<strong>de</strong>nce measure. The current answer is the interpolated copula C X,Y<br />
, which is a bilinear<br />
interpolation of the distribution function D of the pair (F (X), F (Y )) on the discrete set<br />
Im(FX) × Im(FY ).<br />
Let (ui, vj) ∈ Im(FX) × Im(FY ), i.e. ui = FX(i) and uj = FX(j). For all (u, v) ∈<br />
[ui, ui+1] × [vj, vj+1], the interpolated copula is <strong>de</strong>fined as<br />
C X,Y (u, v) = ¯ λu ¯ λvD(ui, vj)+λu ¯ λvD(ui+1, vj)+ ¯ λuλvD(ui, vj+1)+λuλvD(ui+1, vj+1), (4.12)<br />
where λu = (u−ui)/(ui+1 −ui), λv = (v −vj)/(vj+1 −vj). This copula was already mentioned<br />
in Lemma 2.3.5 of Nelsen (2006) to prove the Sklar theorem. The interpolated copula can<br />
also be interpreted as the copula of (X + U, Y + V ) where U, V are two in<strong>de</strong>pen<strong>de</strong>nt uniform<br />
random variables, see, e.g., Section 4 of Denuit and Lambert (2005). This formulation is useful<br />
when doing random generation.<br />
The properties of the interpolated copula of the pair (X, Y ) are: (i) Kendall’s tau τ(X, Y ) =<br />
τ(C X,Y ) and Spearman’s rho ρ(X, Y ) = ρ(C X,Y ), (ii) C X,Y is absolutely continuous and (iii)<br />
<strong>de</strong>pends on marginals<br />
X ⊥ Y ⇔ C X,Y = Π, the in<strong>de</strong>pen<strong>de</strong>nt copula. Unfortunately, C X,Y<br />
and when FX = FY does not imply C X,Y (u, v) = min(u, v), as well as FX = 1 − FY <br />
C X,Y (u, v) = (u + v − 1)+, see, e.g., Genest and Neslehova (2007).<br />
As the interpolated copula is absolutely continuous, a <strong>de</strong>nsity can be <strong>de</strong>rived. In<strong>de</strong>ed, by<br />
differentiating Equation (4.12) with respect to u and v, we get<br />
c X,Y (u, v) = D(ui, vj) − D(ui+1, vj) − D(ui, vj+1) + D(ui+1, vj+1)<br />
, (4.13)<br />
(ui+1 − ui)(vj+1 − vj)<br />
for (u, v) ∈ [ui, ui+1] × [vj, vj+1].<br />
To better see the differences between the copula CY1,Y2 and the interpolated copula C W1,W2 ,<br />
we choose to plot the <strong>de</strong>nsities on Figure 4.3 and not the distribution functions. On Figure 4.3,<br />
we plot <strong>de</strong>nsities of the continuous copula of the pair (Y1, Y2) and the interpolated copula of<br />
the pair (W1, W2), in the same setting as the previous subsection, i.e. Θ is gamma distributed<br />
Ga(2, 4).<br />
4.4.4 Zero-modified distributions<br />
In the previous Subsections, we worked only with a simple geometric distribution. Following<br />
the mo<strong>de</strong>l presented in Subsection 4.2.2, we use hereafter a zero-modified distribution<br />
201