Beruflich Dokumente
Kultur Dokumente
Distribution function of X:
FX (x) = Pr(X x)
(x) =
Z
s(x) = exp (y)dy
d
t qx
dt
d
t px
dt
d
Tx
dt
d
Lx
dt
d
ex
dt
= Pr[T (x) t]
t+u qx
t qx
t px
t+u px
t qx
(x + t) = fT (x) (t)
= t px (x + t)
= lx
= dx
= (x)
ex 1
X
ex =
k px
u qx+t
t px
t px
= 1 t qx
(y)dy
Derivatives:
t|u qx
FX (z) FX (x)
1 FX (x)
s(x) s(z)
s(x)
= exp
n px
Notations:
t qx
0
x+n
Z
k=1
s(x + t)
s(x)
s(x + t)
= 1
s(x)
V ar[T (x)] = 2
V ar[K(x)] =
k px
k px qx+k
k| qx
c
Exam M - Life Contingencies - LGD
(2k 1) k px e2x
k=1
t t px dt
e2x
k+1 px
Z
Tx =
lx+t dt
0
Lx = Tx Tx+1
Z1
Z1
=
lx+t dt = lx t px dt
0
n mx
n Lx = Tx Tx+n =
n1
X
n Lx
Lx+k
R1
k=0
Zn
=
a(x) =
lx+t dt
=
t px
(x + t) dt
Lx lx+1
lx lx+1
E[K] = ex = px (1 + ex+1 )
Tx
lx
E[T ] =
ex = px (1 +
ex+1 ) + qx a(x)
ex = ex: n + n px ex+n
ex =
ex: n + n px
ex+n
Lx
=
lx
c
Exam M - Life Contingencies - LGD
R1
Recursion formulas:
t t px (x + t) dt
ex: 1
ex =
lx lx+1
Lx
lx lx+n
n Lx
mx =
= ex: m +
m px ex+m: n
E[T (m + n)] =
ex: m+n
1
s(x + m(x))
=
s(x)
2
=
ex: m +
ex+m: n
m px
v t t px x (t)dt
E[Z] = Ax =
V ar[Z] =
v k+1 k px qx+k
k=0
Ax (Ax )2
v k+1
k| qx
k=0
V ar[Z] =
E[Z] = Ax: n =
1
Zn
v t t px x (t)dt
V ar[Z] =
Ax (Ax )2
A1x: n (A1x: n )2
E[Z] = A1x: n =
E[Z] = m| Ax =
V ar[Z] =
E[Z] = Ax: n =
n1
X
v k+1 k px qx+k + v n n px
k=0
E[Z] = Ax: n = v n px n Ex
V ar[Z] = 2Ax:1n (Ax:1n )2 = v 2n n px n qx
n
Ax: n (A1x: n )2
v t t px x (t)dt
Zn
v k+1 k px qx+k
k=0
m| Ax
1
m| Ax = Ax Ax: m
n1
X
v t t px x (t)dt + v n n px
Ax = A1x: n + n| Ax
2
Ax = 2A1x: n + n|2Ax
n| Ax = n Ex Ax+n
1
Ax = A
+ n Ex Ax+n
x: n
m|n Ax
m Ex
Ax = v 2 qx + v 2 px 2Ax
m|n Ax
A1x: n
1
Ax+m:
n
1
= A1x: m+n
= m| Ax
c
Exam M - Life Contingencies - LGD
m|n Ax
x: m
m+n| Ax
1
= vqx + vpx Ax+1:
n1
= vpx
(m1)|n Ax+1
Ax: 1
= v
Ax: 2
= vqx + v 2 px
Z
=
bt + 1cv t t px x (t)dt
an
a
n
(Ia) n
(Da) n
(Ia)
(n + 1)a n
s 1
1
(I A)
x: n
Zn
=
bt + 1cv t t px x (t)dt
t v t px x (t)dt
(IA)x =
1 vn
i
1 vn
i
= an
1
1
1
, a = , a
=
i
d
a
n nv n
i
na
n
1
2
(Ia) n + (Da) n
i
iv
1
1+i
= 2
id
i
0
1
(IA)x: n
Zn
=
t v t t px x (t)dt
1
(DA)
x: n
Zn
=
(n btc)v t t px x (t)dt
d =
(DA)1x: n
Zn
=
(Ia)
(n t)v t t px x (t)dt
1 + i (1 + i)2
1
= nvqx + vpx (DA)x+1:
n1
(IA)1x: n
1
(I A)
x: n
+ (DA)1x: n = nA1x: n
1 = (n + 1)A1
+ (DA)
x: n
x: n
(IA)1x: n
+ (DA)1x: n = (n + 1)A1x: n
v v2
i 2i + i2
d 2d d2
i
2i + i2
2
Limit of interest rate i = 0:
i=0
Ax 1
A1x: n
n kx =
n Ex
Share of the survivor:
n| Ax
Ax: n
m|n Ax
1
(1 + i)n
accumulation factor =
=
n Ex
n px
i=0
A1x: n
i=0
n qx
n px
i=0
1
i=0
m|n qx
i=0
(IA)x 1 + ex
i=0
(IA)x
ex
c
Exam M - Life Contingencies - LGD
Recursion relations
Z
x+1
a
x = a
x: 1 + vpx a
a
t t px (x + t)dt
a
x = E[
aT ] =
a
x = a
x: n +
v t px dt =
a
x = 1 + vpx a
x+1
a
x = 1 + v 2 px 2a
x+1
t Ex dt
a
x: n
Ax (Ax )2
2
V ar[
aT ] =
a
x: n
= 1 + vpx a
x+1: n1
(m)
(m)
= a
(m)
n Ex a
x+n
x
a
x: n
x
n| a
v t t px dt =
(I
a)x = 1 + vpx [(I
a)x+1 + a
x+1 ]
Zn
= a
x + vpx (I
a)x+1
t Ex dt
0
Ax: n (Ax: n )2
2
V ar[Y ] =
v t t px dt =
a
x = E[
a K+1 ] =
x
n| a
V ar[
a K+1 ] =
k=0
2
)
t Ex dt
n
= v n px a
x+n = n Ex a
x+n
2 2n
V ar[Y ] =
v n px (
ax+n 2a
x+n ) (n| a
x )2
x
n| a
Ax (Ax
d2
a
x: n
= E[Y ] =
n1
X
v k k px
k=0
Ax: n (Ax: n )2
d2
V ar[Y ] =
n-yr certain and life annuity: a
x: n
a
x: n
= a
x + a
n a
x: n
= a
n +
x
n| a
= E[Y ] =
v k k px
k=n
= a
x a
x: n
1 =
ax + Ax
1 Ax
a
x =
ax: n
Ax: n = 1
2
x: n
Ax: n = 1 (2) 2a
n Ex
a
x+n
1 Ax
d
1 Ax: n
a
x: n =
d
1 = d
ax: n + Ax: n
a
x: n
a
x =
c
Exam M - Life Contingencies - LGD
x
n| a
=a
n + n Ex a
x+n
= a
x + a
n a
x: n
X
= a
n +
v k k px
k=n
= a
n +n| a
x
v k k px
X
ax = E[
aK ] =
v k k px
Accumulation function:
k=1
ax =
sx: n
1 (1 + i)Ax
i
a(m)
x
(m)
ax: n
1
nt Ex+t
(m)
Ax
d(m)
(m)
i=0
ax ex
(m)
(Ax )2
(d(m) )2
1
= a
(m)
x
m
1
(m)
= a
x: n (1 n Ex )
m
2
V ar[Y ] =
Zn
0
m-thly annuities
a
(m)
=
x
a
x: n
=
=
n Ex
Ax
c
Exam M - Life Contingencies - LGD
i=0
a
x 1 + ex
i=0
ex
a
x
i=0
ax: n
ex: n
a
x: n
1 + ex: n1
a
x: n
ex: n
i=0
i=0
P (Ax ) =
P (Ax ) =
P (Ax ) =
V ar[L] =
V ar[L] =
V ar[L] =
h P (Ax )
Ax
a
x
Ax
1 Ax
1
a
x
2
2
P
Ax (Ax )2
1+
2
Ax (Ax )2
(
ax )2
2
Ax (Ax )2
(1 Ax )2
P (Ax ) =
P (Ax ) =
V ar[L] =
V ar[L] =
V ar[L] =
Ax
= Px
a
x
dAx
1 Ax
1
d
a
x
P 22
1+
Ax (Ax )2
d
2
Ax (Ax )2
(d
ax )2
2
Ax (Ax )2
(1 Ax )2
h Px: n
1
( n Px Px:
sx: n
n )
(Px: n n Px )
sx: n
1
(Px: n Px:
sx: n
n )
Ax: n
a
x: n
a
x: n
A#
(m)
a
#
c
Exam M - Life Contingencies - LGD
(1)
1
Px:
n
Px: n1
Px: n n Px
= 1 Ax+n =
Px: n1
1
Px: n Px:
n
= 1=
Px: n1
= Ax+n =
Miscellaneous identities:
P (Ax: n )
Ax: n =
P (Ax: n ) +
h Px
a
x: h
Ax
a
x: h
Ax: n
a
x: h
Ax
P (Ax ) =
a
x
(m)
Px: n1 sx: n = 1
P#
h P (Ax: n )
Ax
a
x: h
Ax: n
P (Ax ) =
Px: n
Px: n + d
1
P (Ax: n ) +
1
Px: n + d
n Px
2
Ax+t (Ax+t )2
V ar[ t L] =
assuming EP
(1 Ax )2
2 h
i
P
2
V ar[ t L] =
1+
Ax+t: nt (Ax+t: nt )2
2
Ax+t: nt (Ax+t: nt )2
V ar[ t L] =
assuming EP
(1 Ax: n )2
Benefit reserve t V :
The expected value of the prospective loss at
time t.
Continuous reserve formulas:
Prospective:
Retrospective:
Premium diff.:
Paid-up Ins.:
Annuity res.:
Death ben.:
Premium res.:
A1x: n
V
(
A
)
=
P
(
A
)
s
t
x
x x: t
t Ex
x+t
Ax+t
t V (Ax ) = 1
P (Ax+t )
a
x+t
t V (Ax ) = 1
a
x
x+t Ax
A
t V (Ax ) =
1 Ax
P (Ax+t ) P (Ax )
t V (Ax ) =
P (Ax+t ) +
tV
Cost of insurance: funding of the accumulated costs of the death claims incurred between
age x and x + t by the living at t, e.g.
4 kx
=
=
1 kx
= Ax+k Pk a
x+k
i
h
x+k: nk
= Px+k: nk Px: n a
#
"
Px: n
Ax+k: nk
= 1
Px+k: nk
dx
qx
=
lx+1
px
= Px: n sx: n t kx
a
x+k
= 1
a
x
Px+k Px
=
Px+k + d
Ax+k Ax
=
1 Ax
h
k Vx: n
h
kV
( n Px Px )
sx: n
(Px: n Px )
sx: n
( m Px: n
n
n Vx
n Vx
n Vx: n
n Vx
sx: m
m Px )
m
m Vx: n
m
m Vx
= Ax+m: nm Ax+m
Relation between various terminal reserves (whole life/endowment only):
= Ax+k: nk h Px: n a
x+k: hk
c
Exam M - Life Contingencies - LGD
= 1 n Vx
= Ax+t: nt h P (Ax: n )
ax+t: ht
= Ax+k: nk
n Vx:1 n )
= Px a
x+n
n
n Vx
= Ax+n 0 = Ax+n
h-payment reserves:
h
tV
m+n+p Vx
(m) (m)
h Px: n a
x+k: hk
= 1
(1
m Vx )(1
tV
bh+j+1 v
j+1
j px+h qx+h+j
j=0
j=0
h+j v j j px+h
t+u v u u px+t du
Recursion relations:
hV
h+1 V
+ qx+h (bh+1
h+1 V
h+1 V
h+1 V
Terminology:
policy year h+1 the policy year from time t = h to time t = h + 1
h V + h initial benefit reserve for policy year h + 1
h V terminal benefit reserve for policy year h
h+1 V terminal benefit reserve for policy year h + 1
Net amount at Risk for policy year h + 1
Net Amount Risk bh+1
h+1 V
n1
X
[P vqx+h (bh+1
h+1 V
)] (1 + i)nh
h=0
= P s n
n1
X
vqx+h (bh+1
h+1 V
)(1 + i)nh
h=0
If the death benefit is equal to the benefit reserve for the first n policy years
nV
= P s n
If the death benefit is equal to $1 plus the benefit reserve for the first n policy years
nV
= P s n
n1
X
vqx+h (1 + i)nh
h=0
c
Exam M - Life Contingencies - LGD
If the death benefit is equal to $1 plus the benefit reserve for the first n policy years and qx+h q
constant
nV
= P s n vq
s n = (P vq)
sn
s px+h
h+s V
+ s qx+h v 1s
UDD
i.e.
1s
h+s V
+ s qx+h (v 1s
1s qx+h+s
bh+1 + v
1s
h+s V
1s px+h+s
h+s V
= (1 s)( h V + h ) + s(h+1 V )
h+s V
= (1 s)( h V ) + (s)(h+1 V ) +
h+1 V
(1 s)(h )
{z
}
|
unearned premium
E[h ] = 0
V ar[h ] = v 2 (bh+1
h+1 V
)2 px+h qx+h
h+1 V
V ar[ h L] = v 2 (bh+1
h+1 V
)2 px+h qx+h
+v 4 (bh+2
h+2 V
+v 6 (bh+3
h+3 V
c
Exam M - Life Contingencies - LGD
10
+ t pxy = t px + t py
Axy + Axy = Ax + Ay
a
xy + a
xy = a
x + a
y
t qxy
ex +
ey
exy +
exy =
= 1 t pxy
exy + exy = ex + ey
n| qxy
n| qx
n| qy
n| qxy
Independant lives
t pxy
t px
t qxy
t qx
exy =
t pxy dt
t py
+ t qy t qx t qy
exy =
k pxy
t pxy dt
Variances:
t pxy
V ar[T (u)] = 2
xy (t)
fT (xy) (t)
fT (xy) (t)
=
1 FT (xy) (t)
t pxy
V ar[T (xy)] = 2
V ar[T (xy)] = 2
xy (t) = (x + t) + (y + t)
Notes:
For joint-life status, work with ps:
k qxy =
k px
n pxy
k py [IL]
= n px n py
k qx + k qy k qx k qy [IL]
P r[K = k] =
k pxy
k pxy
qx+k:y+k
k pxy
qx+k:y+k =
n qxy
= n qx n qy
k+1 pxy
[1]
n pxy
X
exy = E[K(xy)] =
k pxy
[1]
a
xy
c
Exam M - Life Contingencies - LGD
exy )2
t t pxy dt (
t py [(x + t) + (y + t)]
t px
t t pxy dt (
exy )2
Independant lives
fT (xy) (t) =
t t pu dt (
eu )2
11
n pxy
n pxy
n px
+ n py 2 n p x n p y
n qx
+ n qy 2 n qx n qy
= a
x + a
y 2
axy
Insurances:
Ax = 1
ax
Axy = 1
axy
Axy = 1
axy
= sT (x) (t) et
sT (y) (t) = sT (y) (t) sz (t)
= sT (y) (t) et
sT (x)T (y) (t) = sT (x) (t) T (y) (t) sz (t)
Premiums:
Px =
Pxy =
Insurance functions:
X
Au =
v k+1 k pu qu+k
=
Axy =
Axy =
k=0
X
k=0
X
k=0
Pxy =
v k+1 P r[K = k]
1
d
a
x
1
d
a
xy
1
d
a
xy
Annuity functions:
Z
a
u =
v t t pu dt
k+1
( k px qx+k + k py qy+k
Au (Au )2
2
var[Y ] =
k=0
k pxy qx+k:y+k )
Reversionary annuities:
A reversioanry annuity is payable during the existence of one status u only if another status v
has failed. E.g. an annuity of 1 per year payable
continuously to (y) after the death of (x).
Au (Au )2
Axy (Axy )2
Cov[v T (xy) , v T (xy) ] = (Ax Axy )(Ay Axy )
V ar[Z] =
a
x|y = a
y a
xy
c
Exam M - Life Contingencies - LGD
12
(j)
t qx
)
(j)
fT,J (t, j) = t p(
x x (t)
Joint PDF:
Marginal PDF of J:
fJ (j) =
)
( )
fT (t) = t p(
x x (t)
m
X
=
fT,J (t, j)
Marginal PDF of T :
j=1
(j)
x
(j)
qx
Z
j=1
to decrement j
(j)
( )
x (t)
Survivorship group:
( )
Group of la people at some age a at time t = 0.
Each member of the group has a joint pdf for
time until decrement and cause of decrement.
j=1
( )
t px
x (t)
= e
Rt
(j)
n dx
= la( )
( )
x (s)ds
= la( )
( )
xa pa
xa+n
Z
n qx(j)
( )
t pa
(j)
a (t)dt
xa
( )
n dx
Derivative:
la( ) =
=
Zt
=
0 (j)
t qx
( )
(j)
s px x (s)ds
0 (j)
t px
( )
s px
dx
( )
lx
0
( )
t qx
la(j)
(j)
qx(j)
Integral forms of t qx :
(j)
t qx
j=1
m
X
(j)
n dx
j=1
d ( )
d ( )
) ( )
=
= t p(
t px
t qx
x x
dt
dt
Zt
m
X
)
(
x (s)ds
= 1 t qx(j)
c
Exam M - Life Contingencies - LGD
Zt
= exp (j)
x (s)ds
13
Basic relationships:
Zt h
i
( )
(m)
= exp
(1)
t px
x (s) + + x (s) ds
Z
m
X
(j) t
) (j)
Bx+t
A =
v t p(
x x (t)dt
( )
t px
0 (j)
t qx
0 (j)
t px
m
Y
j=1
0 (i)
t px
i=1
(j)
t qx
( )
t px
Instead of summing the benefits for each possible cause of death, it is often easier to write
the benefit as one benefit given regardless of the
cause of death and add/subtract other benefits
according to the cause of death.
(j)
x (t)
0 (j)
t px
Premiums:
= t qx(j)
= t qx( )
=
=
( )
t px
(j)
qx
( )
t px
(j)
x (t)
( )
t px
Px( ) =
k=0
(j)
qx
( )
qx(j)
Px(j) =
qxt
qx(1)
qx(2)
qx(1)
= t qx(j)
0 (1)
= qx
1
0 (2)
= qx
1
0
= qx(1) 1
1 0 (2)
q
2 x
1 0 (1)
q
2 x
1 0 (2) 1 0 (3) 1 0 (2) 0 (3)
q
qx + qx qx
2 x
2
3
c
Exam M - Life Contingencies - LGD
( )
( )
v k k px
k=0
( )
1 t qx
k=0
(j)
Bk+1 v k+1
P
k=0
0 (j)
t qx
( )
14
( )
(j)
k px qx+k
( )
v k k px
k CV
(w)
k+1 AS
(d)
qx+k (bk+1
Direct formula:
n AS =
(d)
(w)
n1
X
h=0
( )
nh Ex+h
c
Exam M - Life Contingencies - LGD
15
k+1 CV
( )
+ px+k
k+1 AS)
k+1 AS
(w)
qx+k ( k+1 CV
k+1 AS)
x = Ax a
(I A)
x
(x) = > 0, x
x
s(x) = e
lx = l0 ex
(IA)x = Ax a
x
= en = (px )n
1
ex =
= E[T ] = E[X]
ex: n =
ex (1 n px )
1
V ar[T ] = V ar[x] = 2
mx =
ln2
Median[T ] =
= Median[X]
px
ex =
= E[K]
qx
px
V ar[K] =
(qx )2
n px
ax )2
(Ia)x = (
(I
a)x = (
ax )2
Chapter 6
1
Px = vqx = Px:
n
1
P (Ax ) = = P (Ax: n )
Chapter 4
Chapter 7
=
+ 2
= Ax (1 n Ex )
Ax =
Ax
A1x: n
n Ex
k Vx
Ax =
=
( + )2
q
q+i
q
q + 2i + i2
Ax (1 n Ex )
Chapter 5
a
x =
2
a
x =
a
x =
a
x =
Axy =
1
+
1
+ 2
1+i
q+i
(1 + i)2
q + 2i + i2
a
x (1 n Ex )
a
x: n
a
x: n
= a
x (1 n Ex )
c
Exam M - Life Contingencies - LGD
= 0, k = 0, 1, 2, . . .
= e
Ax =
Ax: n
(Ax ) = 0, t 0
tV
n(+)
(IA)x =
( + )2
(1 + i)
=
( + )(q + i)
q(1 + i)
=
(q + i)2
1
=
( + )2
1+i 2
=
q+i
(IA)x = Ax a
x =
Chapter 3
a
xy =
exy =
Axy =
a
xy =
exy =
16
M + F
M + F +
1
M
+ F +
1
M
+ F
qxy
qxy + i
1+i
qxy + i
pxy
qxy
De Moivres Law
Chapter 3
Ax =
x
= l0
( x)
1
= (x) =
x
m
=
x
xn
=
x
= qx = (x) = fT (x), 0 t < x
1
=
(lx + lx+1 )
2
x
=
= E[T ] = Median[T ]
2
x 1
= E[K]
=
2
2
2
( x)
=
12
( x)2 1
=
12
qx
2dx
=
=
1
lx + lx+1
1 2 qx
1
= E[S] =
2
n
= n n px + n q x
2
n
= ex: n + n qx
2
s(x) = 1
lx
qx
n|m qx
n px
t px (x
+ t)
Lx
ex
ex
V ar[T ]
V ar[K]
mx
a(x)
ex: n
ex: n
Ax =
A1x: n
(IA)x =
(IA)x =
A1x: n
(IA)1x: n
(IA)1x: n
2( x)
a x
x
an
x
(Ia) x
x
(Ia) x
x
(Ia) n
x
(Ia) n
x
Chapter 5
No useful formulas: use a
x =
chapter 4 formulas.
exx =
exx =
exy =
x
( MDML with = 2/( x))
3
2( x)
3
eyy + yx qx
ey
yx px
ex =
V ar[T ] =
x c
1
c
x
l0
( x)c
c
x
xn c
x
x
= E[T ]
c+1
c
Exam M - Life Contingencies - LGD
and the
a
x
x
a
n
x
1Ax
d
Chapter 9
Chapter 4
Ax =
a
2(x)
( x)2 c
(c + 1)2 (c + 2)
Chapter 9
exx =
x
2c + 1
ex with =
17
2c
x
t qx
(x + t)
lx+t
s qx+t
V ar[T ]
mx
Lx
a(x)
ex: 1
t px (x
Chapter 6
+ t)
P (A1x: n ) =
P (Ax: n ) =
Px(m) =
hP
Ax(m) =
A1x: n
1
(I A)
x: n
Ax: n
n| Ax
Ax =
(m)
Px: n
(m)
(m)
n Px
(A1x: n ) =
Chapter 7
Chapter 4
Ax =
i
Px
i 1
P
x: n
i 1
+ Px: n1
P
x: n
Px
(m) (m)(Px + d)
Px: n
1
(m) (m)(Px:
n + d)
n Px
1
(m) (m)(Px:
n + d)
i
1(m)
hP
x: n
P (Ax ) =
i
Ax
i
Ax
(m)
i
i 1
A
x: n
i
(IA)1x: n
i 1
i
Ax: n + Ax:1n = A1x: n + n Ex
i
Ax
n|
2i + i2 2
Ax
2
h (m)
k Vx: n
h
k Vx: n
(m)
h (m)
(Ax: n )
kV
h
kV
k V( Ax: n )
i h 1
V
+ hk Vx: h1
k x: h
Chapter 10
UDDMDT
(j)
t qx
qx(j)
qx( )
0 (j)
t px
= t qx(j)
= (j)
x (0)
)
= (
x (0)
qx(j)
( ) qx( )
=
t px
(j)
(j)
x
qx
( )
1 t qx
( )
Chapter 5
)
(
x
a
(m)
x
'
a
(m)
x
(m)
a
x: n
with:
(m)
a
x: n
m1
a
x
2m
(m)
ax (m)
( )
1 t qx
UDDASDT
0 (j)
t qx
(m)
ax: n (m)(1 n Ex )
id
(m) = (m) (m) 1
i d
i i(m)
m1
and (m) = (m) (m)
2m
i d
(m)
(m)
x+n
= a
x n Ex a
=
c
Exam M - Life Contingencies - LGD
qx
qx(1)
qx(2)
qx(1)
18
= t qx(j)
0 (1)
= qx
1
0
= qx(2) 1
0 (1)
= qx
1
1 0 (2)
q
2 x
1 0 (1)
q
2 x
1 0 (2) 1 0 (3) 1 0 (2) 0 (3)
q
qx + qx qx
2 x
2
3
(t)k
k!
E[N (t)] = t
V ar[N (t)] = t
1
V ar[T ] =
2
i (t)] =
E[N
Sn GammaRV[ = n, =
P r[Sn t] =
et
j=n
fSn (t) = et
E[Sn ] =
V ar[Sn ] =
(s)ds
0
Zt
1
]
(t)j
j!
[m(t)]k
k!
(t)n1
(n i)!
N (t)
X(t) =
n
2
Yi
i=1
E[X(t)] = t E[Y ]
V ar[X(t)] = t E[Y 2 ]
c
Exam M - Loss Models - LGD
k = E[X k ]
= (X d)+ =
0
X<d
X d X d
Variance:
V ar[x] = 2 = E[X 2 ] E[X]2
E[(X
= 2 = 2 2
Z
=
(x d)k f (x)dx
d)k+ ]
Standard deviation:
p
= V ar[X]
Coefficient of variation:
= e (d)[1 F (d)]
(x d)k p(x)
x>d
k
Limited loss:
Skewness:
Y = (X u) =
E[(X )3 ]
3
1 =
= 3
3
X X<u
u Xu
Kurtosis:
1 =
E[(X )4 ]
4
= 4
4
Z0
E[X u] =
Zu
F (x)dx +
S(x)dx
0
Zu
[1 F (x)]dx if X is always positive
Y P = X d|X > d
E[(X u) ] =
xk f (x)dx + uk [1 F (u)]
xk p(x) + uk [1 F (u)]
xu
c
Exam M - Loss Models - LGD
Zu
mX (t) = E[etX ]
Sum of random variables Sk = X1 + +Xk :
mSk (t) =
k
Y
j=1
mXj (t)
ai = 1):
aj = 1):
a1 f1 (x)
..
f (x) =
.
ak fk (x)
Tail weight:
existence of moments light tail
hazard rate increases light tail
c0 < x < c1
..
.
ck1 < x < ck
y
1
y
Y = X FY (y) = FX ( ) and fY (y) = fX ( )
Raising to a power:
> 0 : FY (y) = FX (y )
1
fY (y) = y 1 fX (y )
Y = X
< 0 : FY (y) = 1 FX (y )
fY (y) = y 1 fX (y )
PN (z) = E[z N ] = p0 + p1 z + p2 z 2 +
P 0 (1) = E[N ]
P 00 (1) = E[N (N 1)]
Poisson distribution:
pk = e
Exponentiation:
Y = eX
P (z) = e(z1)
FY (y) = FX (ln(y))
fY (y) = y1 fX (ln(y))
E[N ] =
V ar[N ] =
Mixing:
The random variable X depends upon a parameter , itself a random variable . For a given
value = , the individual pdf is fX| (x|).
Z
fX (x) = fX| (x|)f ()d
k+r1
k
1+
k
1
1+
r
P (z) = [1 (z 1)]r
E[N ] = r
V ar[N ] = r(1 + )
k
k!
j=1
Geometric distribution:
Special case of the negative binomial distribution with r = 1
pk =
g0 = Pprim (f0 )
k
gk =
k
(1 + )k+1
P (z) = [1 (z 1)]1 =
k
X
1
jb
fj gkj
a+
1 af0
k
X
j fj gkj
k
j=1
1
1 (z 1)
E[N ] =
V ar[N ] = (1 + )
P (z|) = e(z1)
P (z) = p (1 )e1 (z1) + + p (n )en (z1)
Binomial distribution:
Counting number of successes in m trials given
a probability of sucess q
m
q k (1 q)mk
pk =
k
P (z) = [1 + q(z 1)]m
E[N ] = mq
V ar[N ] = mq(1 q)
E[X] = E [E[X|]]
a = 0 Poisson distribution
Poisson
= nn
Negative Binomial
r,
r = nn r, =
c
Exam M - Loss Models - LGD
n
n
X amount of loss
X : exp[mean ] Y P : exp[mean ]
X : uniform[0, ] Y P : uniform[0, d]
X : 2Pareto[, ] Y P : 2Pareto[0 = , 0 = + ]
fY L (y) = fx (y + d)
FY L (y) = FX (y + d)
relations:
fx (y + d)
1 FX (d)
FX (y + d) FX (d)
1 FX (d)
= Y L |Y L > 0
E[Y L ]
E[Y P ] =
P r[Y L > 0]
E[(Y L )k ]
E[(Y P )k ] =
P r[Y L > 0]
u-coverage limit:
YL = X u
X = X
Y
YP
= (X d)+
= X d|X > d
E[Y ] = E[X u]
Zu
=
[1 Fx (x)]dx
Franchise deductible d:
X = X
Y
= 0 if X < d, X if X > d
= X|X > d,
Y L = (X u) (X d)
P
P
Yfranchise
= Yordinary
+d
YP
= (X u) (X d)|X > 0
= (X u) d|X > d
= X u|X > 0
E[X d]
E[X]
Coinsurance factor :
First calculate Y L and Y P with deductible and
limits. Then apply
Y L = Y L
Y P = Y P
V ar[N ] = (1 p)p
Modified frequency:
N (the modified frequency) has a compound distribution: The primary distribution is N , and the secondary
distribution is the Bernoulli random variable I. The probability generating function is
PN (z) = PN [1 + v(z 1)]
The frequency distribution is modified to include only non-zero claim amounts. Each claim amount probability is modified by dividing it by th eprobability of a non-zero claim. For common distributions, the
frequency distribution changes as follows (the number of positive payments is nothing else than a special
event with probability = v)
N
Parameters for N
Parameters for N
Poisson
= v
Binomial
m, q
m = m, q = vq
Negative Binomial
r,
r = r, = v
c
Exam M - Loss Models - LGD
N
X
(n)
FX
Xj
(0)
FX (k)
fS (k) =
0 , k<0
1 , k0
(n)
pn fX (k)
n=0
Notations:
fk = P r[X = k]
FS (k) =
gk = P r[S = k]
(n)
p n FX
(k)
Compound probabilities:
E[S] =
if P r[X = 0] = f0 6= 0
g0 = PN (f0 )
X
n=0
pk = P r[N = k]
k fS (k) =
k=0
p0 P r[sum of Xs = k]
Z
E [(S d)+ ] =
(x d)fS (x)dx
+ p2 P r[sum of two Xs = k]
+
If N is in the (a, b, 0) class, then
E [(S d)+ ] =
g0 = PN (f0 )
(x d)fS (x)
d+1
k
X
jb
1
fj gkj
a+
1 af0
k
j=1
(n)
E [(S d)+ ] =
(n1)
fX
[1 FS (d)]
= E[S]
(k j)fX (j)
d1
X
[1 FS (d)]
x=0
j=0
(1)
fX (k) = fX (k) = fk
1 , k=0
(0)
fX (k) =
0 , otherwise
E [(S d)+ ] =
(x d)fS (x)
= P r[sum of n Xs = k]
k
X
d1
X
x=0
fX (k) =
[1 FS (k)]
k=0
+ p1 P r[sum of one X = k]
(n)
fX (k)
(k j)fX (j)
if E[N ] is large
gk =
(n1)
FX
j=0
j=1
gk =
k
X
(k) =
da
bd
E [(S a)+ ]+
E [(S b)+ ]
ba
ba
= P r[Mn+k = j|Mn = i]
= Q(i,i)
n
= P r[Mn+1 = = Mn+k = i|Mn = i]
Theorem:
k Qn
= Qn Qn+1 Qn+k1
k Qn
= Qk
Inequality:
(i)
k Pn
(i)
(i)
(i)
k vn
(i)
h
X
i h
i
(s,i)
(i)
Q
k vn
C
k n
n+k
k=0
(i,j)
(i,j)
Qn+l
(i,j)
AP Vs@n (C
AP Vs@n (C (i,j) ) =
h
X
i h
i
(i,j)
(s,i)
(i,j)
Q
k n
n+k+1
n+k
k=0
c
Exam M - Loss Models - LGD
k+1 vn