Professional Documents
Culture Documents
Probability
f ( x) F '( x)
F ( x) f ( x)
F ( x) 1 S ( x)
S ( x) 1 F ( x)
S ( x) e H ( x )
H ( x) ln( S ( x))
h( x) H '( x)
H ( x) h( x)
f ( x) S '( x)
f ( x)
h( x )
S ( x)
E X x p ( x) or
k k
x
k
f ( x)dx
x -
Percentiles
A 100p th percentile of a random variable X is a number p satisfying:
Pr( X p ) p
Pr( X p ) p
Conditional Probability
Pr( A B)
Pr( A | B)
Pr( B)
Pr( B | A) Pr( A)
Pr( A | B)
Pr( B)
Scaling
To scale a lognormal:
X ~ lognormal ,
cX ~ lognormal ln c,
Let Y cX
y y
Then FY ( y ) Pr Y y Pr cX y Pr X FX
c c
Variance
E XY E X E Y if Independent
E aX bY aE X bE Y
Var aX b a 2Var ( X )
Var aX bY a 2Var ( X ) 2abCov X ,Y b 2Var (Y )
Var ( X ) E X 2 E ( X )2
VaR X VaR 1 X
Cov( X , Y ) E XY E ( X ) E (Y )
Cov A B , C D Cov ( A, C ) Cov ( A, D ) Cov ( B, C ) Cov ( B, D )
Cov( A, A) Var ( A)
Cov( A, B) 0 if Independent
Cov( X , Y )
XY (Correlation Coefficient)
X Y
Bernoulli Shortcut: For any RV with only 2 values a and b:
Var ( X ) a - b pq
2
Parametric Distributions
Let h x | a( x)
If a(x) is a constant then the hazard rate is exponential, otherwise Weibull
can be Gamma or Inverse Gaussian
x
A( x) a(t )dt
0
H x | A( x)
S x | e A( x )
Conditional Variance
Splices
Shifting
If f 3e
3 1
The mean = 1/3 + 1 (the mean of the unshifted exponential plus the shift)
Policy Limits
X ^ d x
d
f ( x )dx d k S (d )
k k
E
0
d
kx k 1S ( x)dx
0
d
If Y (1 r ) X then E Y ^ d 1 r E X ^
1 r
Deductibles
Franchise Deductible of d – pays nothing if Loss is less than d, and full amount if Loss > d
Ordinary Deductible
0 X d
Payment from Ins. Co. E X d
X d X d
E X d ( x d ) f ( x)dx
d
S (x)dx
d
E X d
k
( x d )k f ( x)dx E X k E X ^ d
k
d
Payment per Payment with Deductible
Ordinary Deductible
Y P X d | X d
FX ( x d ) FX (d )
FY P ( x)
1 FX (d )
SX (x d )
SY P ( x)
S X (d )
e( d ) E X d | X d
E X d E X E X ^ d
e(d ) E Y P
S (d ) S (d )
x d f ( x)dx S ( x)dx
e( d ) d d
S (d ) S (d )
e(d ) Mean Excess Loss
E X E X ^ d E X d
pmt from customer pmt from ins. co.
E X ^ d E X ^ d | X d Pr X d E X ^ d | X d Pr X d
Average loss < d Pr X d d Pr X d
Franchise Deductibles
Distribution e(d)
Exponential
d
Uniform 0, 2
d
2 Parameter Pareto 1
d
1 d
1 Parameter Pareto d d
d
1
E X ^ d
LER(d )
E X
LER(d ) Expected % of loss not included in payment
Special Cases for LER(d)
Distribution LER(d)
d
Exponential 1 e
1
2 Parameter Pareto 1
d
1
1 Parameter Pareto 1 d
Properties of Risk Measures
Translation Invariance: X c X c
Positive Homogeneity: cX c X
Subadditivity: X Y X Y
Monotinicity: X Y if X Y
FX1 ( p )
xf ( x)dx
1 p
1
VaR ( X )dy
p
y
1 p
VaR p ( X ) e X VaR p ( X )
z p where ( x)
Normal 1 p 2
z p z p
e E X
1 p
Lognormal
S ( x)dx
d
E X d E X u
Coinsurance of 80% means that the insurance pays 80% of the costs
Inflation
u d
E Payment per Loss (1 r ) E X ^
1 r
E X ^
1 r
X Loss RV
Y L payment per loss RV
E Y L 0 Pr X d E Y P Pr X d
Var Y L E Var Y L | Case Var E Y L | Case
0 Pr X d Var Y P Pr X d E Y P 0 Pr X d Pr X d
2
Bonus
Discrete Distributions
n
Pr N n Geometric Distribution (memoryless)
1
A sum of 'n' independent Negative Binomial random variables having the same
and parameters r1,..., rn has a Negative Binomial distribution with parameters
n
and r
i 1
i
and m
i 1
i
pk b
a k 1,2,...
pk 1 k
ab
EN
1 a
ab
Var N
1 a 2
x x( x 1) ( x n 1)
n
n!
P ( n ) (0)
pn
n!
p0 P(0)
( n ) P ( n ) (1)
P '(1) E X
P ''(1) E X ( X 1)
P '''(1) E X ( X 1)( X 2)
pk b
a k 2,3,4,...
pk 1 k
Zero-Truncated Distributions
p0T 0
pk
pkT
1 p0
Zero-Modified Distributions
p0M 0
pkM 1 p0M 1 p p
k
pkM 1 p0M pkT
1 p0M
E M
1 p0
E Orig 1 p0M E Zero Truncated
E N cm
Var ( N ) c 1 c m 2 cv
c 1 p0M
m is the mean of the corresponding zero-truncated distribution
v is the variance of the corresponding zero-truncated distribution
Sibuya
ETNB with 1 r 0 and take lim as
a 1
b r 1
p1T r
Poisson/Gamma
Gamma
Mean
Variance 2
Negative Binomial
Mean r
Variance r 1
Coverage Modifications
1 p0*
1 p0M *
1 p0M
where * indicates revised parameters
1 p0
Aggregate Loss Models
Compound Variance
S= aggregate losses
N = frequency RV
X =severity RV
E S E N E X
Var S E N Var X Var N E X
2
Convolution Method
pn Pr( N n) f N (n)
f n Pr( X n) f X (n)
g n Pr( S n) f S (n)
FS ( x) g
n x
n
i1 ...ik n
fi1 fi2 fik
Method 1 - Definition of E S ^ d
u
E S ^ d
hjg
j 0
hj d Pr S d
Method 2
E S ^ 2.8 2 Pr S 0 0.8 Pr S 2
dist between values of x dist b/w highest value of x below d and d
Aggregate Coverage Modifications
Two cases for which the sum of independent random variables has a simple distribution
1) Normal Distribution. If X i are Normal with mean and variance , their sum is normal.
2
2) Exponential or Gamma Distribution. If X i are exponential or gamma, their sum has a gamma
distribution
Normal Distributions
If n random variables Xi are independent and normally distributed with parameters and 2 ,
their sum is normally distributed with parameters n and n 2 .
Then multiply each these probabilities by their respective p1, p2, etc.
x
x
n
e
The probability of exactly n events occurring before time x is
n!
Gamma CDF
x
x
j
n 1
e
FX ( x) 1
j!
j 0
x
-
If 1, F ( x) 1- e
x x
-
x -
If 2, F ( x) 1- e e
Empirical Models
Bias
Definition: Estimator is consistent if lim Pr n 1 for all 0
n
and Var n 0 as n
2) The MLE is always consistent
3) If MSE 0 then is consistent
2
MSE E |
MSE Var bias
2
Grouped Data
nj
f n ( x)
n c j c j 1
where x is in c j , c j 1
n j = # points in the interval
n = total points
f n ( x) Histogram
Fn ( x) Ogive
If there is a policy limit (say 8000), then E X^8000 would have as its last 2 terms:
2
8000 10,000
5000
f n ( x) x dx
2
8000
f n ( x) 80002 dx
Binomial:
Variance = mq(1 q)
X
If Y= (Binomial Proportion),
m
q (1 q )
Variance =
m
Multinomial:
Variance mqi (1 qi ) i = category
Covariance -mqi q j
X q (1 qi )
If Y= , Variance = i
m m
-q q
Covariance i j
m
Individual Data
S x 1 S x
Var Sn x if S is known
n
Sn x 1 Sn x
Var Sn x
n
nx n nx
Var Sn x 3
where nx is the # of survivors past time x
n
nx n y n y
Var y x p x | nx Var y x q x | nx
nx3
The empirical estimators of S ( x) and f ( x) are unbiased
Individual Data
j 1
si
Sn (t ) 1
i 1 ri
S (t ) S t0
t
t0 where t 0 is the end of the study Exponential Extrapolation
ri risk set
si death
di entry time
ui withdrawal time
xi death time
S x - Pr X x
1 1 2
Shortcut:
n n 1 n 0.5
j 1
Hˆ (t )
i 1
si
ri
ˆ
Sˆ ( x) e H ( x )
Lives that leave at the same time as a death are in the risk set.
Lives that arrive at the same time as a death are not in the risk set.
Censored lives are in the risk set but are not counted as deaths
Confidence Intervals
si
1) Enter in column 1
ri
2) Enter the formula ln(1- L1) for column 2
3) Select L1 as first variable and L2 as second
4) Calculate e x - e y
Nelson Aalen Kaplan Meier
If using Kaplan-Meier or Nelson-Aalen methods, E(X^d) = area under the curve of S(x). Multiply
the base x height of each of the rectangles.
Bayes Theorem
P E | A1 P A1
P A1 | E
P E | A1 P A1 P E | A2 P A2 ... P E | An P An
Greenwood's Approximation of Variance (Kaplan-Meier)
r r s
sj
Var S (t ) S (t ) 2
j j j
yj t
Sn x 1 Sn x
Var S (t )
n
if data is complete (no censoring or truncation)
sj
Var H (t )
rj 2
yj t
S (t) z
n 0.5(1 p ) Var Sn (t ) , Sn (t ) z0.5(1 p ) Var S n (t )
Log-transformed confidence interval for S(t)
z Var Sn (t )
S (t ) 1U , S (t )U where U exp
0.5(1 p )
n n Sn (t ) ln Sn (t )
H (t )
, H (t ) U
z
0.5(1 p ) Var H (t )
where U exp
U H (t )
Kernel Smoothing
Uniform
1
xi b x xi b
k xi ( x) 2b
0
Otherwise
0 x xi b
x xi b
K xi ( x) xi b x xi b
2b
1 x xi b
n
fˆ ( x)
i 1
1
n
k xi ( x)
f n xi probability
n
Fˆ ( x)
i 1
1
n
K xi ( x)
f n xi probability
xi is a sample point
x is the estimation point
Kernel distribution is 1 for observation points more than one bandwidth to the left
Kernel distribution is 0 for observation points more than one bandwidth to the right
K 6 (13) K10 (13) K 25 (13)
ex) To find K12 (11), linearly interpolate between K12 (7) and K12 (17)
Triangular
1
Height of triangle is
b
Base of triangle is 2b
Expected Values
EX |Y Y (Y is the original random variable)
The mean of the smoothed distribution is the same as the original mean
E X E Y
Uniform Kernel
b2
Var ( X ) Var (Y )
3
Triangular Kernel
b2
Var ( X ) Var (Y )
6
Approximations for Large Data Sets
x j is the # of losses in c j , c j 1
r j is the risk set c j , c j 1
q 'j is the decrement rate in c j , c j 1
xj
q 'j
rj
Pj 1 Pj d j u j x j
v j # withdrawals
w j # survivors
v j wj u j
r j Pj 0.5 d j v j
Multiple Decrements
'( x ) xt xt 1 xt 2
p
t 3 t 1 1 1
t t 1 rt 2
r r
'( x )
t 3 qt 1 t 3 pt'( x)
Parametric Models
Method of Moments
n n
i 1
xi
i 1
xi 2
m t
n n
m2 t m 2
Gamma ˆ ˆ
t m2 m
2t 2m2 ˆ
mt
Pareto ˆ
t 2m 2 t 2m 2
Lognormal ˆ 2ln(m) 0.5ln(t ) ˆ 2 2ln(m) ln(t )
Uniform on 0, ˆ 2m
When they don’t specify which moment to use, use the first ‘k’ moments, where ‘k’ is the number
of parameters you’re fitting.
Percentile Matching
ˆ p x(n1) p if (n 1) p is an integer
Otherwise multiply (n 1) p and interpolate
The smoothed empirical percentile is not defined if the product is less than 1 or greater than n
Maximum Likelihood
Distribution Result
Exponential MLE = MoM
Gamma MLE = MoM if fixed
Normal MLE = MoM
Poisson MLE = MoM
Negative Binomial MLE = MoM if r is known
Binomial MLE = MoM if m is known
If the MLE is the sample mean, the variance of the MLE is the variance of the
Var ( X )
distribution =
n
x d i i Yes
Exponential ˆ i 1
n
n
ln x
i 1
i
ˆ
n
Lognormal No
n
ln 2
xi
i 1
ˆ
2
ˆ
n
Inverse Exponential n
ˆ n No
1
xi
i 1
nc nc
n
Uniform on Grouped Data 0, ˆ c j
nj
No
OR Some observations are censored
at a single point c j Upper bound of highest finite interval
There must be at least one observation n j Number of observations below c j
above c j
ˆ Min of
Uniform on Grouped Data 0, 1) UB of highest interval with data
All groups are bounded
n
2) LB of highest interval with data *
nj
Two-parameter Pareto, fixed ˆ
n
K
nc nc Yes
K ln di ln xi
i 1 i 1
n = # of uncensored observations
c = # of censored observations
d = truncation point
x = observation if uncensored or the censoring point if censored
CT = formula can be used for left-truncated or right-censored data
Bernoulli Technique
Whenever there is one parameter and only 2 classes of observations, maximum likelihood will
assign each class the observed frequency, and you can then solve for the parameter.
1) Method of Moments and Percentile Matching only use a limited number of features from the
sample.
3) Method of moments and percentile matching cannot always handle truncation and censoring.
4) Method of moments and percentile matching require arbitrary decisions on which moments
or percentiles to use.
3) There may be local maxima in addition to the global maximum; these must be avoided.
4) It may not be possible to find the maximum by setting the partial derivatives to zero; a
numerical algorithm may be necessary.
Fisher’s Information
1 Parameter
d2
E 2 l Fisher's Information
d
Var
1
2 Parameters
2
2 l , l ,
I , E
2
l , l
,
2
Inverting a Matrix
1
a b 1 d b
c
d ad bc c a
Delta Method
2
dg
Var g( X ) Var ( X ) 1 Variable
dx
2 2
g g g g
Var g( X , Y ) Var ( X ) 2Cov( X , Y ) Var (Y ) 2 Variables
x x y y
Var g( X ) g g General
g g
where g ,..., and is the covariance matrix
1x xk
Take derivative with respect to unknown variable
Fitting Discrete Distributions
ˆ x ˆ x
Poisson
x2 ˆ ˆ 2 x
Negative Binomial rˆ 2
ˆ x
rˆ x
x
knk
2) Calculate and observe the slope as a function of k
nk -1
If ratios are increasing, then a > 0
Poisson = 0 slope
Negative Binomial = positive slope
Binomial = negative slope
n is the # of policies/observations of k
The variance of a mixture is always at least as large as the weighted average of the
variances of the components and usually greater due to:
Distribution Formula
2
Exponential
Var ˆ
n
Uniform 0, n 2
Var ˆ
n 12 (n 2)
2
Weibull fixed
ˆ
Var 2
n
Pareto fixed 2
Var ˆ
n
2
Pareto fixed 2
Var ˆ
n
2
Var ˆ
n
Lognormal
Cov , 0
2
Var ˆ
2n
Poisson
Var ˆ
n
Var X Var ( X )
1
n
Hypothesis Tests – Graphic Comparison
F ( x) F (d )
F * ( x)
1 F (d )
f ( x)
f * ( x)
1 F (d )
D( x) plots
D( x) Fn ( x) F *( x)
Empirical - Fitted
Empirical calculation uses a denominator of n
If D( x) 0, then Fn ( x) F *( x)
had more data x than predicted by model
If D( x) 0, then Fn ( x) F *( x)
had less data x than predicted by model
If data truncated at d , D( d ) 0
1
Every vertical jump has distance
n
p p plots
1
On horizontal axis, one point every multiple of
n 1
Domain and Range of Graph are 0,1
Points are Fn x j , F * x j
The first data point corresponds to the first sample value
Kolmogorov-Smirnov Test
D max Fn ( x) F * ( x)
d x u
Max occurs right before or after a jump
Having censored data lowers D and also lowers the critical value
Critical values 0 as n
Then F4 5000 F4 5000 0.5
Anderson-Darling Test
k
Sn y j ln S * y j ln S * y j 1
2
A nF * (u ) n
2
j 0
k
Fn y j ln F * y j 1 ln F * y j
2
n
j 1
2
k Oj E j
Q Ej
j 1
k O j2
Q E
n
j 1 j
Degrees of Freedom
Distribution with parameters is given
k 1 DoF
Distribution is fitted/estimated by MLE or using different data
k 1 r DoF Parameters are fitted from the data
r # parameters
k # groups
Independent Periods
O j E j
2
k
Q Vj
where V j is the variance
j 1
Degrees of Freeedom k - p (p is number of estimated parameters)
Kolmogorov- Anderson-Darling Chi-square Loglikelihood
Smirnov
If there is censored If there is censored data If there is censored data If there is censored data
data , should , , no adjustment of , no adjustment of
lower critical value should lower critical value critical value critical value
If parameters are If parameters are fitted, If parameters are fitted, If parameters are fitted,
fitted, critical value critical value should be critical value adjusts critical value adjusts
should be lowered lowered automatically automatically
If parameters are added to the model, the new model will have a loglikelihood at least as great.
The # DoF for the likelihood ratio test is the number of free parameters in the alternative model
minus the number of free parameters in the base model (null hypothesis).
Compare 2( LogL1 LogL2 ) to critical value at selected chi-square percentile and DOF
LogL1 Alternative Model Loglikelihood (which will be higher)
LogL2 Base Model
If 2( LogL1 LogL2 ) > critical value, accept alternative hypothesis
Schwarz-Bayesian Criterion
r
LogL - ln n
2
where r is the # parameters
where n is the # of data points
The distribution with the highest resulting LogL is selected
Credibility
eF n0CV 2
2
yp
where n0
k
1 p
y p coefficient from the standard normal = -1
2
Given y p , P% = 100 2 percent corresponding to y p 1
k maximum fluctuation you will accept (i.e. within 5%)
Credibility for
Experience Number of Claims Claim Size (Severity) Aggregate Losses/Pure
expressed in Premium
Exposure Units n0 n0
CVs2 n0
1 CV s
2
eF
Number of Claims
nF
n0
n0 CVs2
n0 1 CVs2
Aggregate Losses n0 s n0 s CVs2
n0 s 1 CVs2
sF
Pure Premium is the expected aggregate loss per policyholder per time period.
Limited Fluctuation Credibility: Non-Poisson Frequency
eF n0CVs2
nF eF f
Credibility for
Experience Number of Claims Claim Size (Severity) Aggregate Losses/Pure Premium
expressed in
Exposure Units 2f s2 2f s2
n0 2
n0 2
n0 2 2
f s f
eF
f s f
Number of Claims 2f s2 2f s2
n0 n0 2 n0
nF f
s f 2
s
2f s2 2f s2
Aggregate Losses
n0 s n0 s 2 n0 s
sF f
s f 2
s
# of Insureds is Exposure
Partial Credibility
PC ZX 1 Z M
PC M Z X M
PC Credibility Premium
M Manual Premium
Z Credibility
X Observed Mean
n
Z
nF
n Expected Claims
nF Number of Expected Claims needed for Full Credibility
Bayesian Credibility
Class 1 Class 2
1) Prior Probabilities
2) Likelihood of Experience
3) Joint Probabilities Product of rows above Product of rows above
4) Posterior Probabilities Quotients of row 3 over row 3 sum Quotients of row 3 over row 3 sum
5) Hypothetical Means
6) Bayesian Premium Product of rows 4 and 5 Product of rows 4 and 5
f x1 ,..., xn |
| x1 ,..., xn Posterior Density
f x1,..., xn | d
Limits of Integration are according to prior distribution
N ~ Poisson
1
~ Gamma ,
* + claims
* exposures
Pc *
*
1
Posterior: Gamma * ,*
*
Posterior mean is the avg. # claims/policy
X ~ Normal , v
~ Normal , a
x Observed Average
n Exposure
v anx
* Posterior Mean
v an
va
a* Posterior Variance
v an
Predictive Mean *
Predictive Variance v a*
Bayesian Credibility: Lognormal/Normal
X Lognormal , v
Normal , a
Find
ln xi x
n
v anx
* Posterior Mean
v an
va
a* Posterior Variance
v an
Probability of a claim = q
q Unif 0,1
a* a claims
Plug into Posterior Distribution
b* b exposures - claims
a
E | x *
a* b*
( x 1) x( x)
Bayesian Credibility: Exponential/Inverse Gamma
1 x
f x | e Exponential
e
Inverse Gamma
1
* n
* nx
*
E (next loss)
* 1
Loss Functions
2
l ˆ, ˆ
Bayesian Point Estimate is the mean of the posterior distribution
l ˆ, ˆ
Bayesian Point Estimate is the median of the posterior distribution
Buhlmann's k
v
k
a
Buhlmann's Credibility Z
n na
Z
n k na v
,v v anx v a
* *
a v an
va 2 a* v
a*
v an
Exponential Inverse Gamma Inverse Gamma Pareto
2 2
* n * ( 1)( 2) ( 1)2 ( 2)
* nx *
Exact Credibility
If you have conjugate pairs and they ask for a Buhlmann estimate, use the Bayesian estimate.
Yˆi X i
Cov X , Y
Z
Var ( X )
(1 Z ) E X
Var ( X ) pi X i2 E X
2
Cov X , Y pi X iYi E X E Y
where X are the initial outcomes and Y are the Bayesian observations
Buhlmann Predictions
Pc ( 0 ) (1 Z ) E X
first observation = 0
Pc (2) (1 Z ) E X 2Z
Pc (8) (1 Z ) E X 8Z
Graphics Questions
1) The Bayesian prediction must be within the range of the hypothetical means
-within range of the prior distribution
2) The Buhlmann predictions must lie on a straight line
3) There should be Bayesian predictions both above and below the Buhlmann line
4) The Buhlmann prediction must be between the overall mean and the observation
Cov X i X j
Cov X i X j a
Var X i v a
Empirical Bayes Non-Parametric Methods
̂ x x
Mean of all data
r n
1
r n
( xij xi )2 ij ij
m ( x xi ) 2
i 1 j 1
v̂ r (n 1) i 1 j 1
avg/cell avg/class
r
avg/cell avg/class
Mean of sample
variances of the ( ni 1)
rows i 1 per class
r
mi ( xi x ) 2 vˆ(r 1)
1 r vˆ
( xi x ) 2 i 1 exp/class avg/class overall avg
r = # groups
m = # exposures
Xi X
2
v1
n 1
Poisson Model
ˆ x
vˆ x
aˆ s 2 vˆ
Xi X
2
s2 r 1
r # Policyholders
aˆ
Z regardless of # of years (but if non-uniform exposures, use n = # exposures
aˆ vˆ
for the group you are looking at)
If non-uniform exposures, aˆ must be calculated using Non-parametric formula
For PC
1) X= total # observed claims (but if non-uniform exposures, use the average)
2) If exposure is 5 years, divide PC by 5 to get estimate for 1 year (next year)
Non-Poisson Model
1) Negative Binomial with fixed
E N | r r
Var N | r r (1 )
ˆ x
vˆ x (1 )
aˆ s 2 vˆ
2) Gamma with fixed
E X |
Var X | 2
ˆ x
vˆ x
aˆ s 2 vˆ
Simulation
Inversion Method
1) Get u F ( x)
2) Solve for x
3) Plug in 'u' to get simulated value
If F (2 ) .25
F (2) .75
Then .25 u .75 is mapped to x 2
F ( x) 1 F ( x)
Var Fˆ ( x) n
eF n0CV 2
Estimated Item
Mean
Confidence Interval:
s
x z n
n
sn is the square root of the unbiased sample variance after n runs
Number of Runs:
Calculates number of runs needed for the sample mean to be within 100k% of the true mean.
1
Remember that Var X Var X
n
F(x)
Confidence Interval:
F ( x) z VaR F ( x)
F ( x) 1 F ( x)
F ( x) z
n
Number of Runs:
Pn
n Pn 1 n
n n0 n0 P
Pn n
n
Pn # runs below x
Percentiles q
Confidence Interval:
Y , Y
a b
a nq 0.5 z1 p nq 1 q
2
b nq 0.5 z1 p nq 1 q
2
Risk Measures
TVaR p ( X ) E L | L VaR p
E L E L ^ p
VaR p ( X ) eX VaR p ( X ) p
1 p
FX1 ( p )
xf ( x)dx
1 p
1
VaR ( X )dy y
p
1 p
TVaRq ( X ) is the mean of the upper tail of the distribution
TVaRq ( X ) Conditional Tail Expectation
n
Y j k
j
TVaRq ( X )
n k 1
n
E TVaR ( X ) 2
2
sq2
E TVaR ( X )
n 1
q q
2
sq2 q TVaR q ( X ) VaR q ( X )
Var TVaR q ( X ) n k 1
Bootstrap Approximation
( F ) is the parameter
g x1 ,..., xn is an estimator based on a sample of 'n' items
n
i
x x 2
2 i 1
n
ˆ 2
i
x x 2
MSE x i 1
n n2
Sums of Distributions
Single Multiple
Bernoulli Binomial
Binomial Binomial
Poisson Poisson
Geometric Negative Binomial
Negative Binomial Negative Binomial
Normal Normal
Exponential Gamma
Gamma Gammas
Chi-Square Chi-Square