Arthur CHARPENTIER, Distortion in actuarial sciences Distorting ... in actuarial sciences Arthur Charpentier Université Rennes 1 [email protected]http ://freakonometrics.blog.free.fr/ Montréal Seminar of Actuarial and Financial Mathematics, McGill, November 2010. 1
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
11pt
11pt
Note Example Example
11pt
Proof
Arthur CHARPENTIER, Distortion in actuarial sciences
Montréal Seminar of Actuarial and Financial Mathematics, McGill, November2010.
1
Arthur CHARPENTIER, Distortion in actuarial sciences 1 DECISION THEORY AND DISTORTED RISK MEASURES
1 Decision theory and distorted risk measures
Consider a preference ordering among risks, � such that1. � is distribution based, i.e. if X � Y , ∀X L= X Y
L= Y , then X � Y ; hence,we can write FX � FY
2. � is total, reflexive and transitive,3. � is continuous, i.e. ∀FX , FY and FZ such that FX � FY � FZ , ∃λ, µ ∈ (0, 1)
such thatλFX + (1− λ)FZ � FY � µFX + (1− µ)FZ .
4. � satisfies an independence axiom, i.e. ∀FX , FY and FZ , and ∀λ ∈ (0, 1),
FX � FY =⇒ λFX + (1− λ)FZ � λFY + (1− λ)FZ .
5. � satisfies an ordering axiom, ∀X and Y constant (i.e.P(X = x) = P(Y = y) = 1, FX � FY =⇒ x ≤ y.
2
Arthur CHARPENTIER, Distortion in actuarial sciences 1 DECISION THEORY AND DISTORTED RISK MEASURES
Théorème1Ordering � satisfies axioms 1-2-3-4-5 if and only if ∃u : R→ R, continuous, strictlyincreasing and unique (up to an increasing affine transformation) such that ∀FX and FY :
FX � FY ⇔∫Ru(x)dFX(x) ≤
∫Ru(x)dFY (x)
⇔ E[u(X)] ≤ E[u(Y )].
But if we consider an alternative to the independence axiom4’. � satisfies an dual independence axiom, i.e. ∀FX , FY and FZ , and ∀λ ∈ (0, 1),
FX � FY =⇒ [λF−1X + (1− λ)F−1
Z ]−1 � [λF−1Y + (1− λ)F−1
Z ]−1.
we (Yaari (1987)) obtain a dual representation theorem,Théorème2Ordering � satisfies axioms 1-2-3-4’-5 if and only if ∃g : [0, 1]→ R, continuous, strictlyincreasing such that ∀FX and FY :
FX � FY ⇔∫Rg(FX(x))dx ≤
∫Rg(FY (x))dx
3
Arthur CHARPENTIER, Distortion in actuarial sciences 1 DECISION THEORY AND DISTORTED RISK MEASURES
Standard axioms required on risque measures R : X → R,– law invariance, X L= Y =⇒ R(X) = R(Y )– increasing X ≥ Y =⇒ R(X) ≥ R(Y ),– translation invariance ∀k ∈ R, =⇒ R(X + k) = R(X) + k,– homogeneity ∀λ ∈ R+, R(λX) = λ · R(X),– subadditivity R(X + Y ) ≤ R(X) +R(Y ),– convexity ∀β ∈ [0, 1], R(βλX + [1− β]Y ) ≤ β · R(X) + [1− β] · R(Y ).
– additivity for comonotonic risks ∀X and Y comonotonic,R(X + Y ) = R(X) +R(Y ),
– maximal correlation (w.r.t. measure µ) ∀X,
R(X) = sup {E(X · U) where U ∼ µ}
– strong coherence ∀X and Y , sup{R(X + Y )} = R(X) +R(Y ), where X L= X
and Y L= Y .
4
Arthur CHARPENTIER, Distortion in actuarial sciences 1 DECISION THEORY AND DISTORTED RISK MEASURES
Proposition1IfR is a monetary convex fonction, then the three statements are equivalent,– R is strongly coherent,– R is additive for comonotonic risks,– R is a maximal correlation measure.
Proposition2A coherente risk measureR is additive for comonotonic risks if and only if there exists adecreasing positive function φ on [0, 1] such that
R(X) =∫ 1
0φ(t)F−1(1− t)dt
where F (x) = F(X ≤ x).
see Kusuoka (2001), i.e. R is a spectral risk measure.
5
Arthur CHARPENTIER, Distortion in actuarial sciences 1 DECISION THEORY AND DISTORTED RISK MEASURES
Definition1A distortion function is a function g : [0, 1]→ [0, 1] such that g(0) = 0 and g(1) = 1.
For positive risks,Definition1Given distortion function g, Wang’s risk measure, denotedRg , is
Rg (X) =∫ ∞
0g (1− FX(x)) dx =
∫ ∞0
g(FX(x)
)dx (1)
Proposition1Wang’s risk measure can be defined as
Rg (X) =∫ 1
0F−1X (1− α) dg(α) =
∫ 1
0VaR[X; 1− α] dg(α). (2)
6
Arthur CHARPENTIER, Distortion in actuarial sciences 1 DECISION THEORY AND DISTORTED RISK MEASURES
More generally (risks taking value in R)Definition2We call distorted risk measure
R(X) =∫ 1
0F−1(1− u)dg(u)
where g is some distortion function.
Proposition3R(X) can be written
R(X) =∫ +∞
0g(1− F (x))dx−
∫ 0
−∞[1− g(1− F (x))]dx.
7
Arthur CHARPENTIER, Distortion in actuarial sciences 1 DECISION THEORY AND DISTORTED RISK MEASURES
risk measures R distortion function g
VaR g (x) = I[x ≥ p]Tail-VaR g (x) = min {x/p, 1}PH g (x) = xp
Dual Power g (x) = 1− (1− x)1/p
Gini g (x) = (1 + p)x− px2
exponential transform g (x) = (1− px) / (1− p)
Table 1 – Standard risk measures, p ∈ (0, 1).
8
Arthur CHARPENTIER, Distortion in actuarial sciences 1 DECISION THEORY AND DISTORTED RISK MEASURES
Here, it looks like risk measures can be seen as R(X) = Eg◦P(X).Remark1Let Q denote the distorted measure induced by g on P, denoted g ◦ P i.e.
Q([a,+∞)) = g(P([a,+∞))).
Since g is increasing on [0, 1] Q is a capacity.
Example1Consider function g(x) = xk. The PH - proportional hazard - risk measure is
R(X; k) =∫ 1
0F−1(1− u)kuk−1du =
∫ ∞0
[F (x)]kdx
If k is an integer [F (x)]k is the survival distribution of the minimum over k values.
Definition2The Esscher risk measure with parameter h > 0 is Es[X;h], defined as
Es[X;h] = E[X exp(hX)]MX(h) = d
dhlnMX(h).
9
Arthur CHARPENTIER, Distortion in actuarial sciences 2 ARCHIMEDEAN COPULAS
2 Archimedean copulas
Definition3Let φ denote a decreasing function (0, 1]→ [0,∞] such that φ(1) = 0, and such thatφ−1 is d-monotone, i.e. for all k = 0, 1, · · · , d, (−1)k[φ−1](k)(t) ≥ 0 for all t. Definethe inverse (or quasi-inverse if φ(0) <∞) as
is a copula, called an Archimedean copula, with generator φ.
Let Φd denote the set of generators in dimension d.Example2The independent copula C⊥ is an Archimedean copula, with generator φ(t) = − log t.
10
Arthur CHARPENTIER, Distortion in actuarial sciences 2 ARCHIMEDEAN COPULAS
The upper Fréchet-Hoeffding copula, defined as the minimum componentwise,M(u) = min{u1, · · · , ud}, is not Archimedean (but can be obtained as the limit ofsome Archimedean copulas).
Set λ(t) = exp[−φ(t)] (the multiplicative generator), then
Note that it is possible to get an interpretation of that distortion of theindependence.
A large subclass of Archimedean copula in dimension d is the class ofArchimedean copulas obtained using the frailty approach.
Consider random variables X1, · · · , Xd conditionally independent, given a latentfactor Θ, a positive random variable, such that P (Xi ≤ xi|Θ) = Gi (x)Θ whereGi denotes a baseline distribution function.
11
Arthur CHARPENTIER, Distortion in actuarial sciences 2 ARCHIMEDEAN COPULAS
The joint distribution function of X is given by
FX (x1, · · · , xd) = E (P (X1 ≤ x1, · · · , Xd ≤ Xd|Θ))
= E
(d∏i=1
P (Xi ≤ xi|Θ))
= E
(d∏i=1
Gi (xi)Θ
)
= E
(d∏i=1
exp [−Θ (− logGi (xi))])
= ψ
(−
d∑i=1
logGi (xi)),
where ψ is the Laplace transform of the distribution of Θ, i.e.ψ (t) = E (exp (−tΘ)) . Because the marginal distributions are given respectivelyby
Fi(xi) = P(Xi ≤ xi) = ψ (− logGi (xi)) ,the copula of X is
C (u) = FX
(F−1
1 (u1) , · · · , F−1d (ud)
)= ψ
(ψ−1 (u) + · · ·+ ψ−1 (ud)
)This copula is an Archimedean copula with generator φ = ψ−1 (see e.g. Clayton(1978), Oakes (1989), Bandeen-Roche & Liang (1996) for more details).
12
Arthur CHARPENTIER, Distortion in actuarial sciences 3 HIERARCHICAL ARCHIMEDEAN COPULAS
3 Hierarchical Archimedean copulas
It is possible to look at C(u1, · · · , ud) defined as
Arthur CHARPENTIER, Distortion in actuarial sciences 3 HIERARCHICAL ARCHIMEDEAN COPULAS
It is also possible to consider
φ−13 [φ3(φ−1
1 [φ1(u1) + φ1(u2) + φ1(u3)])+ φ3(φ−1
2 [φ2(u4) + φ2(u5)])].
if φ3 ◦ φ−11 andφ3 ◦ φ−1
2 are inverses of Laplace transform. Or
φ−13 [φ3(φ−1
1 [φ1(u1) + φ1(u2)] + φ3(u3)+ φ3(φ−1
2 [φ2(u4) + φ2(u5)])].
16
Arthur CHARPENTIER, Distortion in actuarial sciences 3 HIERARCHICAL ARCHIMEDEAN COPULAS
U1 U2 U3 U4 U5
φ1
φ3
φ2
U1 U2 U3 U4 U5
φ1
φ3
φ2
Figure 2 – Copules Archimédiennes hiérarchiques avec deux constructions dif-férentes.
17
Arthur CHARPENTIER, Distortion in actuarial sciences 3 HIERARCHICAL ARCHIMEDEAN COPULAS
Example3If φi’s are Gumbel’s generators, with parameter θi, a sufficient condition for C to be aFNA copula is that θi’s increasing. Similarly if φi’s are Clayton’s generators.
Again, an heuristic interpretation can be derived.
18
Arthur CHARPENTIER, Distortion in actuarial sciences 4 DISTORTING COPULAS
4 Distorting copulas
Genest & Rivest (2001) extended the concept of Archimedean copulasintroducing the multivariate probability integral transformation (Wang, Nelsen &Valdez (2005) called this the distorted copula, while Klement, Mesiar & Pap(2005) or Durante & Sempi (2005) called this the transformed copula). Considera copula C. Let h be a continuous strictly concave increasing function[0, 1]→ [0, 1] satisfying h (0) = 0 and h (1) = 1, such that
is a copula. Those functions will be called distortion functions.Example4A classical example is obtained when h is a power function, and when the power is theinverse of an integer, hn(x) = x1/n, i.e.
Dhn(C) (u, v) = Cn(u1/n, v1/n), 0 ≤ u, v ≤ 1 and n ∈ N.
Then this copula is the survival copula of the componentwise maxima : the copula of
19
Arthur CHARPENTIER, Distortion in actuarial sciences 4 DISTORTING COPULAS
is an i.i.d. sample, and the (Xi, Yi)’s have copula C.
A max-stable copula is a copula C such that ∀n ∈ N,
Cn(u1/n1 , · · · , u1/n
d ) = C(u1, · · · , ud).
Example5Let φ denote a convex decreasing function on (0, 1] such that φ(1) = 0, and defineC(u, v) = φ−1(φ(u) +φ(v)) = Dexp[−φ](C⊥). This function is an Archimedean copula.
In the bivariate case, h need not be differentiable, and concavity is a sufficientcondition.Let Hd denote the set of continuous strictly increasing functions [0, 1]→ [0, 1]such that h (0) = 0 and h (1) = 1, C ∈ C,
is a copula, called distorted copula.Hd-copulas will be functions Dh (C) for some distortion function h and some
20
Arthur CHARPENTIER, Distortion in actuarial sciences 4 DISTORTING COPULAS
copula C.
d-increasingness of function Dh (C) is obtained when h ∈ Hd, i.e. h is continuous,with h (0) = 0 and h (1) = 1, and such that h(k)(x) ≤ 0 for all x ∈ (0, 1) andk = 2, 3, · · · , d (see Theorem 2.6 and 4.4 in Morillas (2005)).
As a corollary, note that if φ ∈ Φd, then h(x) = exp(−φ(x)) belongs to Hd.Further, observe that for h, h′ ∈ Hd,
Arthur CHARPENTIER, Distortion in actuarial sciences 4 DISTORTING COPULAS
Again, it is possible to get an intuitive interpretation of that distortion.Consider a max-stable copula C. Let X be a random vector such that X given Θhas copula C and P (Xi ≤ xi|Θ) = Gi (xi)Θ, i = 1, · · · , d.Then, the (unconditional) joint distribution function of X is given by
F (x) = E (P (X1 ≤ x1, · · · , Xd ≤ xd|Θ)) = E (C (P (X1 ≤ xi|Θ) , · · · ,P (Xd ≤ xd|Θ)))
= E(C(G1 (x1)Θ
, · · · , Gd (xd)Θ))
= E(CΘ (G1 (x1) , · · · , Gd (xd))
)= ψ (− logC (G1 (x1) , · · · , Gd (xd))) ,
where ψ is the Laplace transform of the distribution of Θ, i.e.ψ (t) = E (exp (−tΘ)), since C is a max-stable copula, i.e.
C(G1 (x1)Θ
, · · · , Gd (xd)Θ)
= CΘ (G1 (x1) , · · · , Gd (xd)) .
The unconditional marginal distribution functions are Fi (xi) = ψ (− logGi (xi)),and therefore
CX (x1, · · · , xd) = ψ(− log
(C(exp
[−ψ−1 (x)
], exp
[−ψ−1 (y)
]))).
22
Arthur CHARPENTIER, Distortion in actuarial sciences 4 DISTORTING COPULAS
Note that since ψ−1 is completly montone, then h belongs to Hd.
23
Arthur CHARPENTIER, Distortion in actuarial sciences 5 APPLICATION TO AGING PROBLEMS
5 Application to aging problems
Let T = (T1, · · · , Td) denote remaining lifetime, at time t = 0. Consider theconditional distribution
The survival copula of the conditional distribution is the copula of
(U1, · · · , Ud) given U1 < F 1(t), · · · , Ud < F d(t)
where (U1, · · · , Ud) has distribution C , and where Fi is the distribution of TiLet C be a copula and let U be a random vector with joint distribution functionC. Let u ∈ (0, 1]d be such that C(u) > 0. The lower tail dependence copula of C
24
Arthur CHARPENTIER, Distortion in actuarial sciences 5 APPLICATION TO AGING PROBLEMS
at level u is defined as the copula, denoted Cu, of the joint distribution of U
conditionally on the event {U ≤ u} = {U1 ≤ u1, · · · , Ud ≤ ud}.
5.1 Aging with Archimedean copulas
If C is a strict Archimedean copula with generator φ (i.e. φ(0) =∞), then thelower tail dependence copula relative to C at level u is given by the strictArchimedean copula with generator φu defined by
φu(t) = φ(t · C(u))− φ(C(u)), 0 ≤ t ≤ 1, (3)
where C(u) = φ−1[φ(u1) + · · ·+ φ(ud)] (see Juri & Wüthrich (2002) or C & Juri(2007)).Example6Gumbel copulas have generator φ (t) = [− ln t]θ where θ ≥ 1. For any u ∈ (0, 1]d, thecorresponding conditional copula has generator
φu (t) =[M1/θ − ln t
]θ−M where M = [− ln u1]θ + · · ·+ [− ln ud]θ .
25
Arthur CHARPENTIER, Distortion in actuarial sciences 5 APPLICATION TO AGING PROBLEMS
Example7Clayton copulas C have generator φ (t) = t−θ − 1 where θ > 0. Hence,
hence φu (t) = C(u)−θ · φ(t). Since the generator of an Archimedean copula is uniqueup to a multiplicative constant, φu is also the generator of Clayton copula, withparameter θ.
Théorème3Consider X with Archimedean copula, having a factor representation, and let ψ denotethe Laplace transform of the heterogeneity factor Θ. Let u ∈ (0, 1]d, then X givenX ≤ F−1
X (u) (in the pointwise sense, i.e. X1 ≤ F−11 (u1), · · · ., Xd ≤ F−1
d (ud)) is anArchimedean copula with a factor representation, where the factor has Laplace transform
ψu (t) =ψ(t+ ψ−1 (C(u))
)C(u) .
26
Arthur CHARPENTIER, Distortion in actuarial sciences 5 APPLICATION TO AGING PROBLEMS
where C is a copula, and h ∈ Hd is a d-distortion function.
Assume that there exists a positive random variable Θ, such that, conditionallyon Θ, random vector X = (X1, · · · , Xd) has copula C, which does not depend onΘ. Assume moreover that C is in extreme value copula, or max-stable copula (seee.g. Joe (1997)) : C
(xh1 , · · · , xhd
)= Ch (x1, · · · , xd) for all h ≥ 0. The following
result holds,Lemma1Let Θ be a random variable with Laplace transform ψ, and consider a random vectorX = (X1, · · · , Xd) such that X given Θ has copula C, an extreme value copula.Assume that, for all i = 1, · · · , d, P (Xi ≤ xi|Θ) = Gi (xi)Θ where the Gi’s are
27
Arthur CHARPENTIER, Distortion in actuarial sciences 5 APPLICATION TO AGING PROBLEMS
distribution functions. Then X has copula
CX (x1, · · · , xd) = ψ(− log
(C(exp
[−ψ−1 (x1)
], · · · , exp
[−ψ−1 (xd)
]))),
whose copula is of the form Dh(C) with h(·) = exp[−ψ−1 (·)
].
Théorème4Let X be a random vector with anHd-copula with a factor representation, let ψ denotethe Laplace transform of the heterogeneity factor Θ, C denote the underlying copula, andGi’s the marginal distributions.
Let u ∈ (0, 1]d, then, the copula of X given X ≤ F−1X (u) is
CX,u (x) = ψu
(− log
(Cu
(exp
[−ψ−1
u (x1)], · · · , exp
[−ψ−1
u (xd)])))
= Dhu(Cu)(x),
where hu(·) = exp[−ψ−1
u (·)], and where
– ψu is the Laplace transform defined as ψu (t) = ψ (t+ α) /ψ (α) whereα = − log (C (u∗)), u∗i = exp
[−ψ−1 (ui)
]for all i = 1, · · · , d. Hence, ψu is the
Laplace transform of Θ given X ≤ F−1X (u),
28
Arthur CHARPENTIER, Distortion in actuarial sciences 5 APPLICATION TO AGING PROBLEMS