Properties of Future Lifetime Distributions and Estimation


 Vincent Phelps
 3 years ago
 Views:
Transcription
1 Properties of Future Lifetime Distributions and Estimation Harmanpreet Singh Kapoor and Kanchan Jain Abstract Distributional properties of continuous future lifetime of an individual aged x have been studied. The expressions for moments, coefficients of skewness and kurtosis have been derived when newborn s lifetime follows different distributions. The maximum likelihood (ML estimation of the parameters of future lifetime s distributions has been explored for whole life and term assurance contracts. Simulations have been carried out to find the ML estimates and the corresponding root mean square errors. Keywords: Future Lifetime, Whole life assurance, Term assurance, Maximum Likelihood Estimation, Censoring. 2 AMS Classification: 62P5 1. Introduction Financial and actuarial models are built around the future lifetime random variable of individuals. The distribution of future lifetime is required for statistical measurements, analysis and inference. There are two approaches for modeling this distribution. In the first approach, a theoretical statistical distribution is assumed for the future lifetime associated with a stochastic status. In the second, empirical or observed survival and mortality data are used for constructing life mortality tables [1, 3, 4, 12]. In Actuarial Science, the future lifetime of a life aged x plays a significant role. The importance of its distribution is highlighted when dealing with continuous whole life and term assurance contracts where the death benefit is paid at the moment of death. It also plays a role while studying continuous nyear temporary and nyear deferred whole life annuities. The distribution of future lifetime and its properties are utilized for finding the expected present values (Actuarial values and the variances of the present value random variables for benefits or payments in case of continuous life annuities/assurances. The force of mortality, alternatively known as failure rate/hazard rate function in reliability is defined through its density function [1, 2, 3]. Let X be the new born s age at death and f(x, F (x and S(x be the corresponding probability density function (pdf, cumulative distribution function (cdf and survival function (sf respectively. (x symbolizes an Department of Statistics, Panjab University, Chandigarh1614 (India Department of Statistics, Panjab University, Chandigarh1614 (India Corresponding Author Kanchan Jain
2 2 individual aged x and T (x denotes the future lifetime of (x at the inception of an insurance contract. The survival function of T (x is written as (1.1 tp x = P [T (x > t] = P (X > x + t X > x = S(x + t. S(x tp x gives the survival probability that (x survives for more than t years. For t, t q x = 1 t p x gives the distribution function of T (x and represents the probability that (x does not survive beyond age t. The force of mortality is an important and fundamental concept in modelling future lifetime and is defined as 1 f(x + t (1.2 µ x+t = lim dt + P [t < T (x t + dt T (x > t] = dt S(x + t. In reliability theory, µ x is known as failure rate or hazard rate function and T (x is termed as residual life. One can refer to [2, 5] for the notations given above. If the policies are not surrendered or do not lapse as in case of whole life assurance, then the observed future lifetimes are uncensored. The censored observations arise if the policyholders surrender the policy before the expiry of the contract. In some cases, the policy lapses due to nonpayment of regular premium amount. In term assurance contract or nyear temporary policy, the death benefit is paid only if the insured dies within the term of the contract. In this case, the future lifetime of those individuals who survive the term of the policy, will not be known to the company. The same is true for an nyear temporary life annuity where annuity payments stop when an individual dies within or after n years, whichever is earlier. In literature, many authors have explored the ageing properties of T (x [11]. The distributional and ageing properties of curtate future lifetime K(x, the greatest integer less than or equal to T (x have been already studied [6]. To the best of our knowledge, the distributional properties of T (x have not been studied so far. Hence, we study the distributional forms of T (x for different distributions of X and explore the Maximum Likelihood Estimation (MLE for uncensored and censored cases. The distributions chosen for X are Gompertz, exponential, Weibull, Pareto Type I, exponentiated exponential and burr [7, 8]. Gompertz and Weibull are considered to be very good models for studying human lifelength. Pareto is a particular case of Benktander Gibrat distribution [9] and Burr distribution is considered since Pareto type II is its special case. The expressions for pdfs, n th order moments, coefficients of skewness and kurtosis of different distributions of T (x are derived in Section 2. In Sections 3 and 4, we discuss the maximum likelihood (ML estimation of parameters future lifetime s distributions for uncensored and censored observations in case of whole life and nyear term assurance contracts. Section 5 consists of simulation results for ML estimation using numerical techniques. The root mean square errors (RMSEs are also reported. 2. Distributional Properties of Future Lifetime In this section, we derive the n th order moment of T (x which helps in finding its expectation, variance, coefficients of skewness and kurtosis. The probability
3 3 density function of T (x is given by (2.1 f T (x (t = d dt t p x = t p x µ x+t. The cdf of T (x is t q x = t s p x µ x+s ds [2]. If we assume that the n th order moment of T (x,that is, E[(T (x n ] exists, then E[(T (x n ] = This implies that (2.2 = E[(T (x n ] t t t t n tp x µ x+t dt y n yp x µ x+y dy + y n yp x µ x+y dy + (t n t y n yp x µ x+y dy t yp x µ x+y dy y n yp x µ x+y dy (t n yp x µ x+y dy t (t n t p x As t, LHS of (2.2 tends to zero if E[(T (x n ] exists and is finite. Hence lim t t n tp x. On the other hand, t n tp x. Therefore lim t t n tp x. This gives that lim t t n tp x =. Hence (2.3 E[(T (x n ] = t n d( t p x = nt n 1 tp x dt. For n = 1, if the expected value of T (x is assumed to exist, then E[T (x] = tp x dt as the existence of E[T (x] implies that lim t t( t p x =. ([2], Chapter 3, pp 68. E[T (x] is called the complete expectation of life and is denoted by e x. In reliability theory, this is termed as the mean residual life function. We also get ( 2 (2.4 V [T (x] = 2 t( t p x dt tp x. The expectation and variance of T (x are same as derived in [2, 3]. If µ k denotes the k th order central moment of T (x, then coefficient of skewness is γ 1 = µ 3 µ 3/2 2 (2.5 µ 3 = where µ 2 = V [T (x] and 3t 2 tp x dt 3 tp x dt ( 3 2t t p x dt + 2 tp x dt.
4 4 A measure of kurtosis is given by γ 2 = µ 4 µ 2 2 where (2.6 µ 4 = ( 4t 3 tp x dt 4 ( tp x dt 3t 2 tp x dt ( ( 2 ( t t p x dt tp x dt 3 tp x dt. Using (1.1 and (2.1, we can find the probability density function (pdf and force of mortality of T (x for different distributions of X. For x, t >, the expressions for survival probability ( t p x and force of mortality (µ x+t are listed in Table 1. Table 1: Survival function and force of mortality of T (x Distribution of X tp x µ x+t Exp(λ, λ > e λt λ Gompertz (α, λ, α >, λ > 1 e (αλx (λ t 1/logλ αλ (x+t Weibull (α, λ, α >, λ > Pareto (θ, θ > Burr (γ, θ, γ >, θ > Exponentiated Exp. (α, λ, α >, λ > e (λ(x+tα e (λxα αλ α (x + t α 1 x θ (x + t θ θ (x + t (1 + x γ θ γθ(1 + x γ θ (x + t γ 1 (1 + (x + t γ θ (1 + (x + t γ 1 (1 e λ(x+t α 1 (1 e λx α α(1 e λ(x+t α 1 λe λ(x+t 1 (1 e λ(x+t α From Table 1, we can conclude that : For X Exp (λ and all λ, t p x is independent of x, the initial age; For X Gompertz (α, λ, t p x is a decreasing function of x for all α > and λ > 1; For X Weibull (α, λ, t p x is an increasing (decreasing function of x for all λ and < α < 1 (α > 1 ; For X Pareto (θ, t p x is an increasing function of x for all θ; For X Burr (γ, θ, t p x is an increasing function of x for all θ, γ; For X Exponentiated Exp (α, λ, t p x is an increasing (decreasing function of x for all α and λ > 1 ( < λ < 1. Since t p x = P [T (x > t] is the survival function of T (x, hence it is decreasing in t for the distribution under consideration. The pdfs of future lifetime distributions of X mentioned in Table 1 can be obtained by multiplying t p x and µ x+t.
5 5 In the sequel, the distributions of T (x shall be labeled as future lifetime distributions. 3. Estimation of the parameters for uncensored case In insurance sector, a whole life assurance policy provides cover till the death of the insured. We assume that the insured takes out life assurance by paying either a single premium or a series of premiums and does not surrender the policy till his death. Suppose factory workers buy a whole life assurance policy from the insurer, by paying a single premium. Insurer is liable to pay the sum assured at the moment of death of the individual. Hence, it is important for the insurer to predict the future lifetime of the insured for deciding the premium amount so that the assurance contract is mutually advantageous. The future lifetime T (x of the insured aged x can be modeled by assuming some continuous distribution of X and for doing this accurately, the parameters of the distribution of T (x need to be estimated and this is done through method of maximum likelihood (ML estimation. As the insurance companies are more interested in future lifetime modelling, it will be observed that the MLEs based on future lifetime can be directly computed by knowing the age of the policyholder at the time of initiation of the policy. In the following subsections, we discuss the Maximum Likelihood (ML estimation of parameters of future lifetime distributions when X follows exponential, Weibull, Pareto, burr, exponentiated exponential and Gompertz distributions. For a sample of n individuals aged x, let t i (x denote the future lifetime of the i th individual. For exponential and Pareto future lifetime distributions, the ML estimators of the parameters can be written in a closed form whereas for other distributions under study, the nonlinear equations result while maximizing the loglikelihood. As it is difficult to solve these equations analytically, the estimates are obtained in Section 5 by using numerical approximation through BFGS in R Exponential Future Lifetime Distribution. If X follows Exp (λ, then the pdf of T (x is f T (x(t(x, λ = λe λt(x, λ >. The corresponding likelihood function is L(t(x, λ = n f T (x (t i (x, λ = λ n e λ n ti(x. Maximizing LogL(t(x, λ with respect to λ gives the MLE of λ as ˆλ = 1 t(x where t(x = 1 n t i (x. Hence ML estimator of λ for exponential future lifetime distribution is of similar type as for exponential distribution.
6 Pareto Type I Future Lifetime Distribution. Let X follow Pareto (θ, then the likelihood function for T(x is n θx θ L(t(x, θ =, θ >, and x >. (x + t i (x θ+1 Taking log on both sides and differentiating with respect to θ, we get ( dlog(l(t(x, θ 1 (3.1 = n dθ θ + logx log(x + t i (x. Equating (3.1 to zero and solving, the MLE of θ is found to be n ˆθ = n log(x + t. i(x x 3.3. Weibull Future Lifetime Distribution. For X following Weibull (α, λ, the pdf of T (x is given by f T (x(t(x, λ, α = αλ α (x + t(x (α 1 e λα ((x+t(x α x α, α, λ >, t(x >. The corresponding loglikelihood can be written as log(l(t(x, λ, α = log{αλ α (x + t i (x α 1 e λα ((x+t i(x α x α }. ML estimators of λ and α can be obtained by equating the derivatives of loglikelihood with respect to λ and α to zero and solving where ( = nα n λ αλα 1 (x + t i (x α nx α ; dlog(l(t(x, λ, α dλ dlog(l(t(x, λ, α dα ( 1 = n α + logλ + (λxα log(λx + log(x + t i (x (1 (λ(x + t i (x α logλ 3.4. Burr Future Lifetime Distribution. If X follows Burr (γ, θ then f T (x (t(x, θ, γ = γθ(1 + xγ θ (x + t(x γ 1 (1 + (x + t(x γ θ+1, γ > and θ > If L(t(x, θ, γ denotes the likelihood function,then (3.2 dlog(l(t(x, θ, γ dγ = n γ + nθxγ logx (1 + x γ + log(x + t i (x ( 1 (x + t i(x γ (θ (x + t i (x γ ; (λ(x + t i (x α. (3.3 dlog(l(t(x, θ, γ dθ = n θ + nlog(1 + xγ log(1 + (x + t i (x γ.
7 7 The ML estimators of θ and γ shall be obtained by equating (3.2 and (3.3 to zero and finding numerical solutions to the nonlinear equations in Section Exponentiated Exponential Future Lifetime Distribution. Let X follow Exponentiated Exponential (α, λ, then the corresponding likelihood function is L(t(x, α, λ = n αλ(1 e λ(x+ti(x α 1 e λ(x+ti(x 1 (1 e λx α. The ML estimators of α and λ can be obtained using the nonlinear expressions given by ( dlog(l(t(x, λ, α 1 =n dλ λ + αxe λx (1 e λx α 1 1 (1 e λx α ( (α 1e λ(x+t i(x (3.4 + (x + t i (x 1 ; 1 e λ(x+ti(x (3.5 dlog(l(t(x, λ, α dα ( 1 = n α + (1 e λx α log(1 e λx 1 (1 e λx α + log(1 e λ(x+ti(x Gompertz Future Lifetime Distribution. The pdf of T (x, when X follows Gompertz (α, λ, is Hence (3.6 (3.7 f T (x(t(x = αλ (x+t(x e L(t(x, α, λ = dlog(l(t(x, α, λ dα dlog(l(t(x, α, λ dλ αλ x (λ t(x 1 logλ α >, λ > 1, t(x and x >. αλ n x (λ ti(x 1 αλ x+ti(x e logλ is the corresponding likelihood function. = n α n = αλx 1 logλ + λ x (λ ti(x 1 ; logλ ( (λ ti(x 1 1( logλ x λti(x t i (x (x + t i (x. λ Equating (3.6 and (3.7 to zero, the MLEs of α and λ are obtained in simulations section using numerical approximation. The following section discusses the estimation of parameters of different continuous future lifetime distributions when some observations are censored in a given set of observations.
8 8 4. Estimation of parameters in censored case In insurance sector, it is possible that some insured persons do not continue the contract due to some reasons. They surrender the policy or don t pay the renewal premium after few years to draw benefit from the contract. Such individuals are considered to be censored and their time of death is not known. The time of occurrence of such events is not known in advance and is a random variable, named as censoring variable. This type of censoring is known as right censoring. For an individual aged x, let the random variable C(x be the time at which an individual either leaves the contract or stops paying premium and is randomly censored. T (x and C(x are assumed to be independent. In the subsequent discussion, we study the estimation under term assurance contract and whole life assurance contract in case of censoring Estimation for term assurance contract. We consider a term assurance contract, tenable for t years. Under this contract, an insurer pays the sum assured only if insured dies within the term of the contract. This means that at the time of inception of the policy, a policy holder aged x gets the benefit if he dies between x and x + t. Hence, the insured ones surviving at time t and those who surrender the policy due to one or the other reason are lost to further follow up by the company. Surviving persons are right censored at a predetermined time t and those who surrender the policy are randomly censored. We attempt to estimate the unknown parameters of the distributionsof future lifetime. For this type of situation, we attempt to estimate the unknown parameters of the distributions of future lifetime. Let the random variable Z be defined as T (x if T (x < C(x and T (x < t ; Z = min(t (x, C(x, t = C(x if C(x < T (x and C(x < t ; t if t < T (x and t < C(x. For i th life aged x, we take t i (x to be the future lifetime and c i (x, the censoring time. This means that the policy is either surrendered or gets lapsed. We assume that, there are n individuals aged x in the cohort and a total of m leave the cohort due to death or some other reason. Out of m, r is assumed to be the random no. of individuals that leave the cohort due to death. Hence, there will be (n m observations in the cohort at time t and the likelihood function [1] can be written as r m L(z = f T (x(t i (xḡc(x(t i (x g C(x (c i (x F T (x (c i (x (4.1 [ FT (x (t ḠC(x(t ] n m where F T (x (. : survival function of T (x, Ḡ C(x (. : survival function of C(x, g C(x (. : probability density function of C(x. In the following discussion, we explore the maximum likelihood estimation of the
9 9 parameters of exponential, Weibull, Pareto, burr and Gompertz future lifetime distributions when some observations are censored. For all the cases, the distribution of C(x is assumed to follow U(, η where the value of η depends on the percentage of censoring Exponential Future Lifetime Distribution If λ is the parameter of exponential distribution followed by X, the likelihood function for T (x using (4.1, is written as ( r L(z, λ =λ r e λ r ti(x Maximizing the loglikelihood, we have ˆλ = ( 1 t i(x η r r t i(x + m c. i(x + (n mt Weibull Future Lifetime Distribution ( e λ m Let X follow Weibull (α, λ and C(x follow U(, η, then ci(x η m r ( ( e λt 1 t n m (x. η Hence L(z, α, λ = e n(λxα dlog(l(z, α, λ dα r m ( αλ α (x + t i (x α 1 e (λ(x+ti(xα 1 t i(x η e (λ(x+ci(xα η = r α + rlogλ + n(λxα log(λx ( (e (λ(x+tα 1 t n m (x. η + (log(x + t i (x λ α (x + t i (x α log(λ(x + t i (x m (λ(x + c i (x α log(λ(x + c i (x (4.2 (n m(λ(x + t α log(λ(x + t ; (4.3 dlog(l(z, α, λ dλ = αr λ + αλα 1 (nx α (n m(x + t α αλ α 1 ( (x + t i (x α + m (x + c i (x α ML estimates of α and λ are found by solving the nonlinear equations obtained as a result of equating (4.2 and (4.3 to zero. This is done using numerical approximation and corresponding RMSEs are calculated in Section 5.
10 Pareto Type I Future Lifetime Distribution If X follows Pareto (θ, θ >, the MLE of θ is r ˆθ = r log(x + t i(x nlogx + m log(x + c i(x + (n mlog(x + t Burr Future Lifetime Distribution For X Burr(γ, θ γ, θ >, the ML estimators of θ and γ can be obtained using nonlinear equations given by dlog(l(z, γ, θ = r m dθ θ + nlog(1 + xγ log(1 + (x + t i (x γ log(1 + (x + c i (x γ dlog(l(z, γ, θ dγ (n mlog(1 + (x + t γ ; = r γ + nθxγ logx (1 + x γ + θ m ( 1 θ(x + ti (x γ log(x + t i (x (1 + (x + t i (x γ (x + c i (x γ log(x + c i (x (1 + (x + c i (x γ (n mθ(x + t γ log(x + t (1 + (x + t γ. The estimation of γ and θ shall be explored in Section 5 by using numerical methods Gompertz Future Lifetime Distribution If X follows Gompertz (α, λ, then using the loglikelihood function, we have (4.4 (4.5 dlog(l(z, α, λ dα dlog(l(z, α, λ dλ = nλx logλ + r α + (x + t i (xlogλ ( λx λ ti(x + logλ m λ ci(x (n m λx+t logλ. x 1 xlogλ 1 =nαλ (logλ 2 (n mαλ x+t 1 logλ(x + t 1 (logλ 2 + m x + t i (x λ αλ x+ti(x 1 logλ(x + t i(x 1 (logλ 2 αλ x+ci(x 1 logλ(x + c i(x 1 (logλ 2 ; By applying NewtonRaphson Method and using (4.4 and (4.5, the estimates are found in Section 5.
11 Estimation for whole life assurance contracts in case of censoring. In insurance, there are some policies which pay benefit till or at the time of the death of the insured, known as whole life assurance contracts. This happens provided the insured does not leave the cohort due to any other reason except death. To evaluate benefits accurately under this situation, the insurance company takes account of all the conditions under which insured may leave the contract. For example, the insured may surrender the policy at a random time giving rise to a censored observation. We take account of this censoring for estimation of parameters of the future lifetime distributions under whole life assurance contract. We write { T (x if T (x < C(x Z = min(t (x, C(x = C(x if C(x < T (x where T (x and C(x are defined in Section 4.1. The likelihood function, under random right censoring [1] is written as (4.6 L(z = r f T (x(t i (xḡc(x(t i (x n g C(x (c i (x F T (x (c i (x. In the following subsections, ML estimation of unknown parameters of different future lifetime distributions shall be discussed by assuming that C(x U(, η where the value of η depends upon the assumed percentage of censoring Exponential Future Lifetime Distribution Let X follow Exponential (λ, then using (4.6 L(z, λ = λ r e λ r r ( ti(x 1 t i(x e λ n ci(x η η n r. Differentiating the loglikelihood with respect to λ and solving the normal equation, we get ˆλ = r r t i(x + n c i(x Weibull Future Lifetime Distribution If X follows Weibull (α, λ, then considering the log likelihood and differentiating with respect to α and λ, we get (4.7 dlog(l(z, α, λ dα = r α + rlogλ + n(λxα log(λx + log(x + t i (x (x + t i (x α λ α log(λ(x + t i (x (λ(x + c i (x α log(λ(x + c i (x;
12 12 (4.8 dlog(l(z, α, λ dλ ( = αr λ αλα 1 nx α + (x + t i (x α (x + c i (x α. When (4.7 and (4.8 are equated to zero, the resulting nonlinear equations can be solved using numerical methods. The results are demonstrated in Section Pareto Type I Future Lifetime Distribution If X follows Pareto (θ, θ >, then differentiation of log likelihood with respect to θ and equating the resulting expressions to zero, gives r ˆθ = r log(x + t i(x nlogx + n log(x + c i(x Burr Future Lifetime Distribution Let X follow Burr (γ, θ, γ, θ >, then (4.9 (4.1 dlog(l(z, γ, θ dθ dlog(l(z, γ, θ dγ = r θ + nlog(1 + xγ = r γ + rθxγ log(x (1 + x γ θ log(1 + (x + t i (x γ log(1 + (x + c i (x γ ; + log(x + t i (x (θ + 1 (x + t i (x γ log(x + t i (x (1 + (x + t i (x γ (x + c i (x γ log(x + c i (x (1 + (x + c i (x γ + θ(n rxγ logx (1 + x γ = r γ + nθxγ logx [ (1 + x γ + (x + t i (x γ ] log(x + t i (x 1 (θ + 1 (1 + (x + t i (x γ θ (x + c i (x γ log(x + c i (x (1 + (x + c i (x γ. In Section 5, the estimates of γ and θ are obtained by equating (4.9 and (4.1 to zero and solving the resulting equations using numerical methods.
13 Gompertz Future Lifetime Distribution If X follows Gompertz (α, λ, dlog(l(z, α, λ dα dlog(l(z, α, λ dλ = nλx logλ + r α + (x + t i (xlogλ [ ] λx λ ti(x + λ ci(x. logλ x 1 xlogλ 1 =nαλ (logλ 2 + ( x + t i (x λ αλx 1 (logλ 2 λ ti(x (logλ(x + t i (x 1 + λ ci(x (logλ(x + c i (x 1 The maximum likelihood estimates are presented in Section Simulations In this section, we generate random samples of sizes 5, 1 and 5 from different continuous future lifetime distributions and estimate their parameters, using BFGS method in R. The root mean square errors (RMSEs are evaluated to check the extent of the deviation of the estimate from the true value. The number of replications is 1 for starting age x = 2, 4 and Uncensored case for whole life assurance contract. Tables 25 give the estimates and the corresponding RMSEs when the data sets are generated from Weibull (α = 1.5 and λ = 2.5, Burr (θ = 3 and γ = 2, exponentiated exponential (α = 1.5 and λ =.5 and Gompertz (α =.5 and λ = 1.8 future lifetime distributions. Table 2: Estimates and RMSEs for Weibull future lifetime dist. x n = 5 n = 1 n = (.626 (.2134 (.435 (.1947 (.112 ( (.173 (.134 (.97 (.891 (.55 ( (.85 (.82 (.44 (.543 (.39 (.131 Table 3: Estimates and RMSEs for Burr future lifetime dist. x n = 5 n = 1 n = (.2532 (.243 (.2232 (.214 (.1532 ( (.1917 (.1859 (.163 (.1215 (.951 ( (.959 (.1123 (.751 (.825 (.275 (.372 ;
14 14 Table 4: Estimates and RMSEs for Exponentiated Exp. future lifetime dist. x n = 5 n = 1 n = (.2468 (.74 (.124 (.512 (.991 ( (.2215 (.76 (.985 (.511 (.925 ( (.276 (.685 (.832 (.434 (.64 (.198 Table 5: Estimates and RMSEs for Gompertz future lifetime dist. x n = 5 n = 1 n = (.117 (.3 (.8 (.2 (.34 ( (.15 (.5 (.13 (.3 (.44 ( (.256 (.2 (.17 (.8 (.78 (.2 On the basis of values in Tables 25, we conclude that for all future lifetime distributions under study RMSEs decrease as sample size increases, that is, the estimates get closer to the assumed values of the parameters for large samples; RMSEs decrease as initial age increases, except when X follows Gompertz distribution Censored case. As mentioned in Section 4.1, C(x, the censoring variable is assumed to follow U(, η. The parameter η for C(x is determined by assuming 3% censoring in the data, that is, by solving (5.1 P [T (x > C(x] =.3. We find the values of η for different combinations of parameters of T (x. The values are generated from the future lifetime and Uniform distributions. These generated sets of values are used to estimate the values of parameters when x = 2, 4, 6; n = 5, 1, 5 and t = 1, 2, 3, 4 and Term Life Assurance Contract Tables 6 8 display the values of η, estimates of parameters of future lifetime distributions and the corresponding RMSEs when the random samples are generated from Weibull (α = 1.5 and λ = 2.5, Burr (θ = 3 and γ = 2, and Gompertz (α =.5 and λ = 1.8 future lifetime distributions. The values in the parentheses give the RMSEs.
15 15 Table 6: Estimates and RMSEs for censored Weibull future lifetime dist. n = 5 n = 1 n = 5 t x = 2, η = (.55 (.25 (.339 (.1596 (.12 ( (.462 (.2243 (.281 (.1375 (.97 ( (.451 (.272 (.262 (.1359 (.96 ( (.415 (.1689 (.259 (.1283 (.6 ( (.399 (.1374 (.16 (.932 (.58 (.19 t x = 4, η = (.223 (.288 (.172 (.271 (.73 ( (.221 (.283 (.155 (.269 (.5 ( (.198 (.251 (.143 (.25 (.46 ( (.189 (.198 (.14 (.225 (.31 ( (.166 (.152 (.14 (.178 (.27 (.56 t x = 6, η = (.225 (.354 (.165 (.185 (.68 ( (.188 (.321 (.141 (.163 (.63 ( (.179 (.313 (.14 (.154 (.62 ( (.145 (.292 (.138 (.153 (.58 ( (.97 (.112 (.75 (.132 (.55 (.94
16 16 Table 7: Estimates and RMSEs for censored Burr future lifetime dist. n = 5 n = 1 n = 5 t x = 2, η = (.4877 (.4392 (.4562 (.4278 (.4223 ( (.4583 (.4363 (.433 (.3959 (.3686 ( (.449 (.431 (.4184 (.363 (.3455 ( (.4223 (.4246 (.3799 (.357 (.2966 ( (.4164 (.447 (.3375 (.3427 (.2114 (.2697 t x = 4, η = (.1755 (.2133 (.1284 (.1396 (.1194 ( (.1612 (.1959 (.126 (.1232 (.845 ( (.1542 (.1897 (.1243 (.189 (.749 ( (.1486 (.1751 (.1212 (.189 (.69 ( (.1468 (.1678 (.789 (.186 (.322 (.326 t x = 6, η = (.1561 (.2291 (.12 (.1775 (.394 ( (.1548 (.2113 (.952 (.1261 (.26 ( (.124 (.1948 (.925 (.11 (.122 ( (.126 (.1938 (.778 (.951 (.117 ( (.73 (.1317 (.599 (.848 (.86 (.171
17 17 Table 8: Estimates and RMSEs for censored Gompertz future lifetime dist. n = 5 n = 1 n = 5 t x = 2, η = (.1395 (.3 (.1155 (.3 (.812 ( (.725 (.3 (.575 (.3 (.268 ( (.453 (.3 (.345 (.2 (.122 ( (.296 (.3 (.23 (.2 (.65 ( (.165 (.2 (.14 (.2 (.5 (.1 t x = 4, η = (.786 (.3 (.716 (.3 (.42 ( (.366 (.3 (.333 (.3 (.128 ( (.257 (.3 (.169 (.2 (.69 ( (.143 (.2 (.11 (.2 (.52 ( (.17 (.2 (.96 (.2 (.48 (.1 t x = 6, η = (.471 (.3 (.424 (.3 (.161 ( (.33 (.3 (.228 (.3 (.16 ( (.327 (.3 (.197 (.3 (.77 ( (.324 (.3 (.247 (.3 (.84 ( (.314 (.3 (.228 (.3 (.78 (.2 From Tables 6 8, it can be observed that RMSEs decrease with an increase in sample size n, initial age x and term of the contract t Whole Life Assurance contract Tables 9 11 show the estimates and the corresponding RMSEs (in parantheses when the data sets are generated from Weibull (α = 1.5 and λ = 2.5, Burr (θ = 3 and γ = 2 and Gompertz (α =.5 and λ = 1.8 future lifetime distributions.
18 18 Table 9: Estimates and RMSEs for censored Weibull future lifetime dist. n = 5 n = 1 n = η =.12 for x = 2 (.182 (.97 (.114 (.75 (.75 ( η =.85 for x = 4 (.224 (.141 (.154 (.88 (.79 ( η =.7 for x = 6 (.264 (.1188 (.356 (.511 (.124 (.156 Table 1: Estimates and RMSEs for censored Burr future lifetime dist. n = 5 n = 1 n = η = 12.5 for x = 2 (.444 (.4117 (.4418 (.4477 (.4643 ( η = 24.5 for x = 4 (.2199 (.2688 (.262 (.1936 (.175 ( η = 36.5 for x = 6 (.2152 (.269 (.1436 (.1845 (.931 (.946 Table 11: Estimates and RMSEs for censored Gompertz future lifetime dist. n = 5 n = 1 n = η = 13 for x = 2 (.135 (.3 (.93 (.2 (.41 ( η = 75 for x = 4 (.189 (.7 (.128 (.3 (.56 ( η = 34 for x = 6 (.339 (.83 (.232 (.23 (.96 (.4 On the basis of Tables 9 11, it is concluded that as sample size increases, RMSEs decrease; as initial age increases, RMSEs increase except for Burr future lifetime distribution.
19 19 Conclusions Estimation of the parameters of future lifetime distributions is helpful in modeling the remaining lifetime of the insured. This is helpful to insurer for setting the premium in order to make the insured policy mutually advantageous. We study the distributional properties of the complete future lifetime and explore the maximum likelihood estimation of the parameters when the insured does not leave the cohort till the end of the contract or leaves at a random time. Acknowledgement The first author is grateful to University Grants Commission, Government of India, for providing financial support for this work.the authors are extremely grateful to the refrees for providing valuable guidance and comments that helped in improving the manuscript. References [1] Borowiak, D. S. and Shapiro, A. F. Financial and Actuarial Statistics: An Introduction. A Series of Textbooks and Monographs, Marcel Dekker Publication, New York, 23. [2] Bowers, N. L., Gerber, H. U., Hickman, J. C, Jones, D. A. and Nesbitt, C. J. Actuarial Mathematics, The Society of Actuaries, USA, [3] Dickson, D. C. M., Hardy, M. R. and Waters, H. R. Actuarial Mathematics for Life Contingent Risks, International Series on Actuarial Science, Cambridge University Press, UK, 29. [4] Gerber, H.U. Life Insurance Mathematics, Springer, Germany, 199. [5] International Actuarial Notation. Journal of the Institute of Actuaries, 75, , [6] Jain, K. and Kapoor, H. P. S. Ageing Properties of Curtate Future Life Time. Sri Lankan Journal of Applied Statistics, 14 (1, 5584, 213. [7] Johnson, N. L., Balakrishnan, N. and Kotz, S. Continuous Univariate Distributions 1, Wiley Series in Probability and Statistics, John Wiley and Sons, USA, [8] Johnson, N. L., Balakrishnan, N. and Kotz, S. Continuous Univariate Distributions 2, Wiley Series in Probability and Statistics, John Wiley and Sons, USA, [9] Kleiber, C. and Kotz, S. Statistical Size Distributions in Economics and Actuarial Sciences, John Wiley and Sons, New Jersey, Hoboken, 23. [1] Klein, J. P. and Moeschberger, M. L. Survival Analysis: Techniques for Censored and Truncated Data, Second edition, Springer Verlag, USA, 23. [11] Lai, C. D. and Xie, M. Stochastic Ageing and Dependence for Reliability, Springer, USA, 26. [12] Zhu, Y. J. Actuarial Model: Life Insurance and Annuity, Series in Actuarial Science, International Press, USA, 27.
Maximum Likelihood Estimation
Math 541: Statistical Theory II Lecturer: Songfeng Zheng Maximum Likelihood Estimation 1 Maximum Likelihood Estimation Maximum likelihood is a relatively simple method of constructing an estimator for
More informationJANUARY 2016 EXAMINATIONS. Life Insurance I
PAPER CODE NO. MATH 273 EXAMINER: Dr. C. BoadoPenas TEL.NO. 44026 DEPARTMENT: Mathematical Sciences JANUARY 2016 EXAMINATIONS Life Insurance I Time allowed: Two and a half hours INSTRUCTIONS TO CANDIDATES:
More informationAssurancevie 2. Manuel Morales Department of Mathematics and Statistics University of Montreal. February 2006
Assurancevie 2 Manuel Morales Department of Mathematics and Statistics University of Montreal February 2006 Chapter 9 Multiple Life Functions 1. Joint Distributions 2. Jointlife status 3. Lastsurvivor
More informationSOA EXAM MLC & CAS EXAM 3L STUDY SUPPLEMENT
SOA EXAM MLC & CAS EXAM 3L STUDY SUPPLEMENT by Paul H. Johnson, Jr., PhD. Last Modified: October 2012 A document prepared by the author as study materials for the Midwestern Actuarial Forum s Exam Preparation
More informationINSURANCE RISK THEORY (Problems)
INSURANCE RISK THEORY (Problems) 1 Counting random variables 1. (Lack of memory property) Let X be a geometric distributed random variable with parameter p (, 1), (X Ge (p)). Show that for all n, m =,
More informationA Uniform Asymptotic Estimate for Discounted Aggregate Claims with Subexponential Tails
12th International Congress on Insurance: Mathematics and Economics July 1618, 2008 A Uniform Asymptotic Estimate for Discounted Aggregate Claims with Subexponential Tails XUEMIAO HAO (Based on a joint
More informationManual for SOA Exam MLC.
Chapter 4. Life Insurance. Extract from: Arcones Manual for the SOA Exam MLC. Fall 2009 Edition. available at http://www.actexmadriver.com/ 1/14 Level benefit insurance in the continuous case In this chapter,
More informationACTS 4301 FORMULA SUMMARY Lesson 1: Probability Review. Name f(x) F (x) E[X] Var(X) Name f(x) E[X] Var(X) p x (1 p) m x mp mp(1 p)
ACTS 431 FORMULA SUMMARY Lesson 1: Probability Review 1. VarX)= E[X 2 ] E[X] 2 2. V arax + by ) = a 2 V arx) + 2abCovX, Y ) + b 2 V ary ) 3. V ar X) = V arx) n 4. E X [X] = E Y [E X [X Y ]] Double expectation
More informationMATH4427 Notebook 2 Spring 2016. 2 MATH4427 Notebook 2 3. 2.1 Definitions and Examples... 3. 2.2 Performance Measures for Estimators...
MATH4427 Notebook 2 Spring 2016 prepared by Professor Jenny Baglivo c Copyright 20092016 by Jenny A. Baglivo. All Rights Reserved. Contents 2 MATH4427 Notebook 2 3 2.1 Definitions and Examples...................................
More informationPrinciple of Data Reduction
Chapter 6 Principle of Data Reduction 6.1 Introduction An experimenter uses the information in a sample X 1,..., X n to make inferences about an unknown parameter θ. If the sample size n is large, then
More informationNonparametric adaptive age replacement with a onecycle criterion
Nonparametric adaptive age replacement with a onecycle criterion P. CoolenSchrijner, F.P.A. Coolen Department of Mathematical Sciences University of Durham, Durham, DH1 3LE, UK email: Pauline.Schrijner@durham.ac.uk
More informationSurvival Distributions, Hazard Functions, Cumulative Hazards
Week 1 Survival Distributions, Hazard Functions, Cumulative Hazards 1.1 Definitions: The goals of this unit are to introduce notation, discuss ways of probabilistically describing the distribution of a
More information1 Cashflows, discounting, interest rate models
Assignment 1 BS4a Actuarial Science Oxford MT 2014 1 1 Cashflows, discounting, interest rate models Please hand in your answers to questions 3, 4, 5 and 8 for marking. The rest are for further practice.
More informationStatistical Analysis of Life Insurance Policy Termination and Survivorship
Statistical Analysis of Life Insurance Policy Termination and Survivorship Emiliano A. Valdez, PhD, FSA Michigan State University joint work with J. Vadiveloo and U. Dias Session ES82 (Statistics in Actuarial
More information1 The Pareto Distribution
Estimating the Parameters of a Pareto Distribution Introducing a Quantile Regression Method Joseph Lee Petersen Introduction. A broad approach to using correlation coefficients for parameter estimation
More informationErrata and updates for ASM Exam C/Exam 4 Manual (Sixteenth Edition) sorted by page
Errata for ASM Exam C/4 Study Manual (Sixteenth Edition) Sorted by Page 1 Errata and updates for ASM Exam C/Exam 4 Manual (Sixteenth Edition) sorted by page Practice exam 1:9, 1:22, 1:29, 9:5, and 10:8
More informationSection 5.1 Continuous Random Variables: Introduction
Section 5. Continuous Random Variables: Introduction Not all random variables are discrete. For example:. Waiting times for anything (train, arrival of customer, production of mrna molecule from gene,
More informationLecture 15 Introduction to Survival Analysis
Lecture 15 Introduction to Survival Analysis BIOST 515 February 26, 2004 BIOST 515, Lecture 15 Background In logistic regression, we were interested in studying how risk factors were associated with presence
More information4. Life Insurance. 4.1 Survival Distribution And Life Tables. Introduction. X, Ageatdeath. T (x), timeuntildeath
4. Life Insurance 4.1 Survival Distribution And Life Tables Introduction X, Ageatdeath T (x), timeuntildeath Life Table Engineers use life tables to study the reliability of complex mechanical and
More informationExam C, Fall 2006 PRELIMINARY ANSWER KEY
Exam C, Fall 2006 PRELIMINARY ANSWER KEY Question # Answer Question # Answer 1 E 19 B 2 D 20 D 3 B 21 A 4 C 22 A 5 A 23 E 6 D 24 E 7 B 25 D 8 C 26 A 9 E 27 C 10 D 28 C 11 E 29 C 12 B 30 B 13 C 31 C 14
More informationGamma Distribution Fitting
Chapter 552 Gamma Distribution Fitting Introduction This module fits the gamma probability distributions to a complete or censored set of individual or grouped data values. It outputs various statistics
More informationTABLE OF CONTENTS. 4. Daniel Markov 1 173
TABLE OF CONTENTS 1. Survival A. Time of Death for a Person Aged x 1 B. Force of Mortality 7 C. Life Tables and the Deterministic Survivorship Group 19 D. Life Table Characteristics: Expectation of Life
More informationOverview of Monte Carlo Simulation, Probability Review and Introduction to Matlab
Monte Carlo Simulation: IEOR E4703 Fall 2004 c 2004 by Martin Haugh Overview of Monte Carlo Simulation, Probability Review and Introduction to Matlab 1 Overview of Monte Carlo Simulation 1.1 Why use simulation?
More informationPremium Calculation. Lecture: Weeks 1214. Lecture: Weeks 1214 (Math 3630) Annuities Fall 2015  Valdez 1 / 32
Premium Calculation Lecture: Weeks 1214 Lecture: Weeks 1214 (Math 3630) Annuities Fall 2015  Valdez 1 / 32 Preliminaries Preliminaries An insurance policy (life insurance or life annuity) is funded
More informationTABLE OF CONTENTS. GENERAL AND HISTORICAL PREFACE iii SIXTH EDITION PREFACE v PART ONE: REVIEW AND BACKGROUND MATERIAL
TABLE OF CONTENTS GENERAL AND HISTORICAL PREFACE iii SIXTH EDITION PREFACE v PART ONE: REVIEW AND BACKGROUND MATERIAL CHAPTER ONE: REVIEW OF INTEREST THEORY 3 1.1 Interest Measures 3 1.2 Level Annuity
More informationPractice problems for Homework 11  Point Estimation
Practice problems for Homework 11  Point Estimation 1. (10 marks) Suppose we want to select a random sample of size 5 from the current CS 3341 students. Which of the following strategies is the best:
More informationSection 6.1 Joint Distribution Functions
Section 6.1 Joint Distribution Functions We often care about more than one random variable at a time. DEFINITION: For any two random variables X and Y the joint cumulative probability distribution function
More informationModeling Individual Claims for Motor Third Party Liability of Insurance Companies in Albania
Modeling Individual Claims for Motor Third Party Liability of Insurance Companies in Albania Oriana Zacaj Department of Mathematics, Polytechnic University, Faculty of Mathematics and Physics Engineering
More informationMonitoring Software Reliability using Statistical Process Control: An MMLE Approach
Monitoring Software Reliability using Statistical Process Control: An MMLE Approach Dr. R Satya Prasad 1, Bandla Sreenivasa Rao 2 and Dr. R.R. L Kantham 3 1 Department of Computer Science &Engineering,
More informationAnnuities. Lecture: Weeks 911. Lecture: Weeks 911 (STT 455) Annuities Fall 2014  Valdez 1 / 43
Annuities Lecture: Weeks 911 Lecture: Weeks 911 (STT 455) Annuities Fall 2014  Valdez 1 / 43 What are annuities? What are annuities? An annuity is a series of payments that could vary according to:
More informationNovember 2012 Course MLC Examination, Problem No. 1 For two lives, (80) and (90), with independent future lifetimes, you are given: k p 80+k
Solutions to the November 202 Course MLC Examination by Krzysztof Ostaszewski, http://www.krzysio.net, krzysio@krzysio.net Copyright 202 by Krzysztof Ostaszewski All rights reserved. No reproduction in
More informationParametric Survival Models
Parametric Survival Models Germán Rodríguez grodri@princeton.edu Spring, 2001; revised Spring 2005, Summer 2010 We consider briefly the analysis of survival data when one is willing to assume a parametric
More informationBayesian Estimation of Joint Survival Functions in Life Insurance
ISBA 2, Proceedings, pp. ISBA and Eurostat, 21 Bayesian Estimation of Joint Survival Functions in Life Insurance ARKADY SHEMYAKIN and HEEKYUNG YOUN University of St. Thomas, Saint Paul, MN, USA Abstract:
More informationMay 2012 Course MLC Examination, Problem No. 1 For a 2year select and ultimate mortality model, you are given:
Solutions to the May 2012 Course MLC Examination by Krzysztof Ostaszewski, http://www.krzysio.net, krzysio@krzysio.net Copyright 2012 by Krzysztof Ostaszewski All rights reserved. No reproduction in any
More information4. Life Insurance Payments
4. Life Insurance Payments A life insurance premium must take into account the following factors 1. The amount to be paid upon the death of the insured person and its present value. 2. The distribution
More informationChapter 2 Premiums and Reserves in Multiple Decrement Model
Chapter 2 Premiums and Reserves in Multiple Decrement Model 2.1 Introduction A guiding principle in the determination of premiums for a variety of life insurance products is: Expected present value of
More information7.1 The Hazard and Survival Functions
Chapter 7 Survival Models Our final chapter concerns models for the analysis of data which have three main characteristics: (1) the dependent variable or response is the waiting time until the occurrence
More informationComparison of resampling method applied to censored data
International Journal of Advanced Statistics and Probability, 2 (2) (2014) 4855 c Science Publishing Corporation www.sciencepubco.com/index.php/ijasp doi: 10.14419/ijasp.v2i2.2291 Research Paper Comparison
More informationOperational Risk Modeling Analytics
Operational Risk Modeling Analytics 01 NOV 2011 by Dinesh Chaudhary Pristine (www.edupristine.com) and Bionic Turtle have entered into a partnership to promote practical applications of the concepts related
More informationIntroduction to Basic Reliability Statistics. James Wheeler, CMRP
James Wheeler, CMRP Objectives Introduction to Basic Reliability Statistics Arithmetic Mean Standard Deviation Correlation Coefficient Estimating MTBF Type I Censoring Type II Censoring Eponential Distribution
More informationProbability and Statistics Prof. Dr. Somesh Kumar Department of Mathematics Indian Institute of Technology, Kharagpur
Probability and Statistics Prof. Dr. Somesh Kumar Department of Mathematics Indian Institute of Technology, Kharagpur Module No. #01 Lecture No. #15 Special DistributionsVI Today, I am going to introduce
More informationActuarial Mathematics and LifeTable Statistics. Eric V. Slud Mathematics Department University of Maryland, College Park
Actuarial Mathematics and LifeTable Statistics Eric V. Slud Mathematics Department University of Maryland, College Park c 2006 Chapter 4 Expected Present Values of Insurance Contracts We are now ready
More informationECE302 Spring 2006 HW5 Solutions February 21, 2006 1
ECE3 Spring 6 HW5 Solutions February 1, 6 1 Solutions to HW5 Note: Most of these solutions were generated by R. D. Yates and D. J. Goodman, the authors of our textbook. I have added comments in italics
More informationThe Actuary s Free Study Guide for. Second Edition G. Stolyarov II,
The Actuary s Free Study Guide for Exam 3L Second Edition G. Stolyarov II, ASA, ACAS, MAAA, CPCU, ARe, ARC, API, AIS, AIE, AIAF First Edition Published in AugustOctober 2008 Second Edition Published in
More informationNIST GCR 06906 Probabilistic Models for Directionless Wind Speeds in Hurricanes
NIST GCR 06906 Probabilistic Models for Directionless Wind Speeds in Hurricanes Mircea Grigoriu Cornell University NIST GCR 06906 Probabilistic Models for Directionless Wind Speeds in Hurricanes Prepared
More informationPremium calculation. summer semester 2013/2014. Technical University of Ostrava Faculty of Economics department of Finance
Technical University of Ostrava Faculty of Economics department of Finance summer semester 2013/2014 Content 1 Fundamentals Insurer s expenses 2 Equivalence principles Calculation principles 3 Equivalence
More informationInstitute of Actuaries of India Subject CT3 Probability and Mathematical Statistics
Institute of Actuaries of India Subject CT3 Probability and Mathematical Statistics For 2015 Examinations Aim The aim of the Probability and Mathematical Statistics subject is to provide a grounding in
More informationUNIT I: RANDOM VARIABLES PART A TWO MARKS
UNIT I: RANDOM VARIABLES PART A TWO MARKS 1. Given the probability density function of a continuous random variable X as follows f(x) = 6x (1x) 0
More informationExact Confidence Intervals
Math 541: Statistical Theory II Instructor: Songfeng Zheng Exact Confidence Intervals Confidence intervals provide an alternative to using an estimator ˆθ when we wish to estimate an unknown parameter
More informationA LOGNORMAL MODEL FOR INSURANCE CLAIMS DATA
REVSTAT Statistical Journal Volume 4, Number 2, June 2006, 131 142 A LOGNORMAL MODEL FOR INSURANCE CLAIMS DATA Authors: Daiane Aparecida Zuanetti Departamento de Estatística, Universidade Federal de São
More informationParametric Models Part I: Maximum Likelihood and Bayesian Density Estimation
Parametric Models Part I: Maximum Likelihood and Bayesian Density Estimation Selim Aksoy Department of Computer Engineering Bilkent University saksoy@cs.bilkent.edu.tr CS 551, Fall 2015 CS 551, Fall 2015
More informationACTUARIAL MATHEMATICS FOR LIFE CONTINGENT RISKS
ACTUARIAL MATHEMATICS FOR LIFE CONTINGENT RISKS DAVID C. M. DICKSON University of Melbourne MARY R. HARDY University of Waterloo, Ontario V HOWARD R. WATERS HeriotWatt University, Edinburgh CAMBRIDGE
More informationInsurance Benefits. Lecture: Weeks 68. Lecture: Weeks 68 (STT 455) Insurance Benefits Fall 2014  Valdez 1 / 36
Insurance Benefits Lecture: Weeks 68 Lecture: Weeks 68 (STT 455) Insurance Benefits Fall 2014  Valdez 1 / 36 An introduction An introduction Central theme: to quantify the value today of a (random)
More informationREGS 2013: Variable Annuity Guaranteed Minimum Benefits
Department of Mathematics University of Illinois, UrbanaChampaign REGS 2013: Variable Annuity Guaranteed Minimum Benefits By: Vanessa Rivera Quiñones Professor Ruhuan Feng September 30, 2013 The author
More informationA Fractional Survival Model
Journal of Data Science 7(29), 487495 A Fractional Survival Model Cheng K. Lee 1 and JenqDaw Lee 2 1 Bank of America and 2 National Cheng Kung University Abstract: A survival model is derived from the
More informationDepartment of Mathematics, Indian Institute of Technology, Kharagpur Assignment 23, Probability and Statistics, March 2015. Due:March 25, 2015.
Department of Mathematics, Indian Institute of Technology, Kharagpur Assignment 3, Probability and Statistics, March 05. Due:March 5, 05.. Show that the function 0 for x < x+ F (x) = 4 for x < for x
More information**BEGINNING OF EXAMINATION** The annual number of claims for an insured has probability function: , 0 < q < 1.
**BEGINNING OF EXAMINATION** 1. You are given: (i) The annual number of claims for an insured has probability function: 3 p x q q x x ( ) = ( 1 ) 3 x, x = 0,1,, 3 (ii) The prior density is π ( q) = q,
More information1 Prior Probability and Posterior Probability
Math 541: Statistical Theory II Bayesian Approach to Parameter Estimation Lecturer: Songfeng Zheng 1 Prior Probability and Posterior Probability Consider now a problem of statistical inference in which
More informationAggregate Loss Models
Aggregate Loss Models Chapter 9 Stat 477  Loss Models Chapter 9 (Stat 477) Aggregate Loss Models Brian Hartman  BYU 1 / 22 Objectives Objectives Individual risk model Collective risk model Computing
More informationManual for SOA Exam MLC.
Chapter 4. Life Insurance. c 29. Miguel A. Arcones. All rights reserved. Extract from: Arcones Manual for the SOA Exam MLC. Fall 29 Edition. available at http://www.actexmadriver.com/ c 29. Miguel A. Arcones.
More informationSurvival Analysis of Left Truncated Income Protection Insurance Data. [March 29, 2012]
Survival Analysis of Left Truncated Income Protection Insurance Data [March 29, 2012] 1 Qing Liu 2 David Pitt 3 Yan Wang 4 Xueyuan Wu Abstract One of the main characteristics of Income Protection Insurance
More informationSome Observations on Variance and Risk
Some Observations on Variance and Risk 1 Introduction By K.K.Dharni Pradip Kumar 1.1 In most actuarial contexts some or all of the cash flows in a contract are uncertain and depend on the death or survival
More informationINSTITUTE AND FACULTY OF ACTUARIES EXAMINATION
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 INSTITUTE AND FACULTY OF ACTUARIES EXAMINATION 8 October 2015 (pm) Subject CT5 Contingencies Core Technical
More informationEDUCATION COMMITTEE OF THE SOCIETY OF ACTUARIES MLC STUDY NOTE SUPPLEMENTARY NOTES FOR ACTUARIAL MATHEMATICS FOR LIFE CONTINGENT RISKS VERSION 2.
EDUCATION COMMITTEE OF THE SOCIETY OF ACTUARIES MLC STUDY NOTE SUPPLEMENTARY NOTES FOR ACTUARIAL MATHEMATICS FOR LIFE CONTINGENT RISKS VERSION 2.0 by Mary R. Hardy, PhD, FIA, FSA, CERA David C. M. Dickson,
More informationExponential Distribution
Exponential Distribution Definition: Exponential distribution with parameter λ: { λe λx x 0 f(x) = 0 x < 0 The cdf: F(x) = x Mean E(X) = 1/λ. f(x)dx = Moment generating function: φ(t) = E[e tx ] = { 1
More informationSOCIETY OF ACTUARIES/CASUALTY ACTUARIAL SOCIETY EXAM C CONSTRUCTION AND EVALUATION OF ACTUARIAL MODELS EXAM C SAMPLE QUESTIONS
SOCIETY OF ACTUARIES/CASUALTY ACTUARIAL SOCIETY EXAM C CONSTRUCTION AND EVALUATION OF ACTUARIAL MODELS EXAM C SAMPLE QUESTIONS Copyright 005 by the Society of Actuaries and the Casualty Actuarial Society
More informationNPTEL STRUCTURAL RELIABILITY
NPTEL Course On STRUCTURAL RELIABILITY Module # 02 Lecture 6 Course Format: Web Instructor: Dr. Arunasis Chakraborty Department of Civil Engineering Indian Institute of Technology Guwahati 6. Lecture 06:
More informationPremium Calculation  continued
Premium Calculation  continued Lecture: Weeks 12 Lecture: Weeks 12 (STT 456) Premium Calculation Spring 2015  Valdez 1 / 16 Recall some preliminaries Recall some preliminaries An insurance policy (life
More informationPremium Calculation. Lecture: Weeks 1214. Lecture: Weeks 1214 (STT 455) Premium Calculation Fall 2014  Valdez 1 / 31
Premium Calculation Lecture: Weeks 1214 Lecture: Weeks 1214 (STT 455) Premium Calculation Fall 2014  Valdez 1 / 31 Preliminaries Preliminaries An insurance policy (life insurance or life annuity) is
More informationEstimating survival functions has interested statisticians for numerous years.
ZHAO, GUOLIN, M.A. Nonparametric and Parametric Survival Analysis of Censored Data with Possible Violation of Method Assumptions. (2008) Directed by Dr. Kirsten Doehler. 55pp. Estimating survival functions
More informationBenchmark Rates for XL Reinsurance Revisited: Model Comparison for the Swiss MTPL Market
Benchmark Rates for XL Reinsurance Revisited: Model Comparison for the Swiss MTPL Market W. Hürlimann 1 Abstract. We consider the dynamic stable benchmark rate model introduced in Verlaak et al. (005),
More informationChapter 6: Point Estimation. Fall 2011.  Probability & Statistics
STAT355 Chapter 6: Point Estimation Fall 2011 Chapter Fall 2011 6: Point1 Estimat / 18 Chap 6  Point Estimation 1 6.1 Some general Concepts of Point Estimation Point Estimate Unbiasedness Principle of
More informationDistribution (Weibull) Fitting
Chapter 550 Distribution (Weibull) Fitting Introduction This procedure estimates the parameters of the exponential, extreme value, logistic, loglogistic, lognormal, normal, and Weibull probability distributions
More informationThe Exponential Distribution
21 The Exponential Distribution From DiscreteTime to ContinuousTime: In Chapter 6 of the text we will be considering Markov processes in continuous time. In a sense, we already have a very good understanding
More informationLecture 2 ESTIMATING THE SURVIVAL FUNCTION. Onesample nonparametric methods
Lecture 2 ESTIMATING THE SURVIVAL FUNCTION Onesample nonparametric methods There are commonly three methods for estimating a survivorship function S(t) = P (T > t) without resorting to parametric models:
More informationA SURVEY ON CONTINUOUS ELLIPTICAL VECTOR DISTRIBUTIONS
A SURVEY ON CONTINUOUS ELLIPTICAL VECTOR DISTRIBUTIONS Eusebio GÓMEZ, Miguel A. GÓMEZVILLEGAS and J. Miguel MARÍN Abstract In this paper it is taken up a revision and characterization of the class of
More information1 Sufficient statistics
1 Sufficient statistics A statistic is a function T = rx 1, X 2,, X n of the random sample X 1, X 2,, X n. Examples are X n = 1 n s 2 = = X i, 1 n 1 the sample mean X i X n 2, the sample variance T 1 =
More informationStatistics 100A Homework 8 Solutions
Part : Chapter 7 Statistics A Homework 8 Solutions Ryan Rosario. A player throws a fair die and simultaneously flips a fair coin. If the coin lands heads, then she wins twice, and if tails, the onehalf
More informationStatistics in Retail Finance. Chapter 6: Behavioural models
Statistics in Retail Finance 1 Overview > So far we have focussed mainly on application scorecards. In this chapter we shall look at behavioural models. We shall cover the following topics: Behavioural
More informationO MIA009 (F2F) : GENERAL INSURANCE, LIFE AND
No. of Printed Pages : 11 MIA009 (F2F) kr) ki) M.Sc. ACTUARIAL SCIENCE (MSCAS) N December, 2012 0 O MIA009 (F2F) : GENERAL INSURANCE, LIFE AND HEALTH CONTINGENCIES Time : 3 hours Maximum Marks : 100
More informationManual for SOA Exam MLC.
Chapter 5 Life annuities Extract from: Arcones Manual for the SOA Exam MLC Fall 2009 Edition available at http://wwwactexmadrivercom/ 1/70 Due n year deferred annuity Definition 1 A due n year deferred
More informationPlease write your name and student number at the spaces provided:
MATH 3630 Actuarial Mathematics I Final Examination  sec 001 Monday, 10 December 2012 Time Allowed: 2 hours (6:008:00 pm) Room: MSB 411 Total Marks: 120 points Please write your name and student number
More informationActuarial Science with
Actuarial Science with 1. life insurance & actuarial notations Arthur Charpentier joint work with Christophe Dutang & Vincent Goulet and Giorgio Alfredo Spedicato s lifecontingencies package Meielisalp
More informationPortfolio Using Queuing Theory
Modeling the Number of Insured Households in an Insurance Portfolio Using Queuing Theory JeanPhilippe Boucher and Guillaume CouturePiché December 8, 2015 Quantact / Département de mathématiques, UQAM.
More informatione.g. arrival of a customer to a service station or breakdown of a component in some system.
Poisson process Events occur at random instants of time at an average rate of λ events per second. e.g. arrival of a customer to a service station or breakdown of a component in some system. Let N(t) be
More information1. A survey of a group s viewing habits over the last year revealed the following
1. A survey of a group s viewing habits over the last year revealed the following information: (i) 8% watched gymnastics (ii) 9% watched baseball (iii) 19% watched soccer (iv) 14% watched gymnastics and
More information4. Continuous Random Variables, the Pareto and Normal Distributions
4. Continuous Random Variables, the Pareto and Normal Distributions A continuous random variable X can take any value in a given range (e.g. height, weight, age). The distribution of a continuous random
More informationModelling spousal mortality dependence: evidence of heterogeneities and implications
1/23 Modelling spousal mortality dependence: evidence of heterogeneities and implications Yang Lu Scor and AixMarseille School of Economics Lyon, September 2015 2/23 INTRODUCTION 3/23 Motivation It has
More informationThe Kelly criterion for spread bets
IMA Journal of Applied Mathematics 2007 72,43 51 doi:10.1093/imamat/hxl027 Advance Access publication on December 5, 2006 The Kelly criterion for spread bets S. J. CHAPMAN Oxford Centre for Industrial
More informationLecture 4 PARAMETRIC SURVIVAL MODELS
Lecture 4 PARAMETRIC SURVIVAL MODELS Some Parametric Survival Distributions (defined on t 0): The Exponential distribution (1 parameter) f(t) = λe λt (λ > 0) S(t) = t = e λt f(u)du λ(t) = f(t) S(t) = λ
More informationManual for SOA Exam MLC.
Chapter 5. Life annuities. Extract from: Arcones Manual for the SOA Exam MLC. Spring 2010 Edition. available at http://www.actexmadriver.com/ 1/114 Whole life annuity A whole life annuity is a series of
More informationContinuous Random Variables
Chapter 5 Continuous Random Variables 5.1 Continuous Random Variables 1 5.1.1 Student Learning Objectives By the end of this chapter, the student should be able to: Recognize and understand continuous
More informationPattern Analysis. Logistic Regression. 12. Mai 2009. Joachim Hornegger. Chair of Pattern Recognition Erlangen University
Pattern Analysis Logistic Regression 12. Mai 2009 Joachim Hornegger Chair of Pattern Recognition Erlangen University Pattern Analysis 2 / 43 1 Logistic Regression Posteriors and the Logistic Function Decision
More informationTime Series Analysis
Time Series Analysis hm@imm.dtu.dk Informatics and Mathematical Modelling Technical University of Denmark DK2800 Kgs. Lyngby 1 Outline of the lecture Identification of univariate time series models, cont.:
More informationMATH 3630 Actuarial Mathematics I Class Test 2 Wednesday, 17 November 2010 Time Allowed: 1 hour Total Marks: 100 points
MATH 3630 Actuarial Mathematics I Class Test 2 Wednesday, 17 November 2010 Time Allowed: 1 hour Total Marks: 100 points Please write your name and student number at the spaces provided: Name: Student ID:
More informationManual for SOA Exam MLC.
Chapter 5 Life annuities Extract from: Arcones Manual for the SOA Exam MLC Fall 2009 Edition available at http://wwwactexmadrivercom/ 1/94 Due n year temporary annuity Definition 1 A due n year term annuity
More informationJoint Exam 1/P Sample Exam 1
Joint Exam 1/P Sample Exam 1 Take this practice exam under strict exam conditions: Set a timer for 3 hours; Do not stop the timer for restroom breaks; Do not look at your notes. If you believe a question
More informationHeriotWatt University. BSc in Actuarial Mathematics and Statistics. Life Insurance Mathematics I. Extra Problems: Multiple Choice
HeriotWatt University BSc in Actuarial Mathematics and Statistics Life Insurance Mathematics I Extra Problems: Multiple Choice These problems have been taken from Faculty and Institute of Actuaries exams.
More informationManual for SOA Exam MLC.
Chapter 5. Life annuities Extract from: Arcones Fall 2009 Edition, available at http://www.actexmadriver.com/ 1/60 (#24, Exam M, Fall 2005) For a special increasing whole life annuitydue on (40), you
More informationMathematics of Life Contingencies MATH 3281
Mathematics of Life Contingencies MATH 3281 Life annuities contracts Edward Furman Department of Mathematics and Statistics York University February 13, 2012 Edward Furman Mathematics of Life Contingencies
More information