Math 370/408, Spring 2008 Prof. A.J. Hildebrand Actuarial Exam Practice Problem Set 4 Solutions About this problem set: These are problems from Course 1/P actuarial exams that I have collected over the years, grouped by topic, and ordered by difficulty. All of the exams have appeared on the SOA website http: //www.soa.org/ at some point in the past, though most of them are no longer there. The exams are copy-righted by the SOA. Note on the topics covered: This problem set covers problems on joint (multivariate) distributions (Chapter 4 of Hogg/Tanis).. Note on the ordering of the problems: The problems are loosely grouped by topic, and very roughly ordered by difficulty (though that ordering is very subjective). The last group of problems (those with triple digit numbers) are harder, either because they are conceptually more difficult, or simply because they requite lengthy computations and may take two or three times as long to solve than an average problem. Answers and solutions: I will post an answer key and hints/solutions on the course ‘webpage, ww#.math.uiuc. edu/~hildebr/370.‘Math 370/408 Spring 2008 Actuarial Exam Practice Problem Set 4 Solutions 1. [4] Let X and ¥ be discrete random variables with joint probability function = [AF for (2) = (0,1), (0,2),(1,2), 1,3), none Oe Si Determine the marginal probability function of X, 1/6 forr=0, (A) plz) = 45/6 forr=1, 0 otherwise. 1/4 for =0, 3/4 forz=1, 0 otherwise. 1/3. forr=0, (C) p(x) = 42/3 for x =1, 0 otherwise. 2/9 _ 3/9 ©) Pl) = Fig 0 otherwise. y/12 for z=0, (E) plz) = 4 (2+y)/12 fore =1, o otherwise. Answer: B Solution: This is a routine calculation of a marginal distribution. 2. [4-9] ‘A car dealership sells 0, 1, or 2 luxury cars on any day. When selling a car, the dealer also to persuade the customer to buy an extended warranty for the car. Let X denote the number of luxury cars sold in a given day, and let Y denote the number of extended ‘warranties sold, and suppose that 1/6 for (xu) = (0,0), 1/12 for (x,y) = (1,0), 1/6 for (zy) = (1.1), 1/12. for (1xy) = (2.0). 1/3 for (2,9) =(2,1), 1/6 for (x,y) = (2,2). P(X=2,¥ =y)= ‘What is the variance of X? (A) 047 (B) 0.58 (©) 0.83 (D) 142 (©) 2.58 Answer: B: 0.58 Solution: This is routine: Find the marginal distribution of X', then use it compute the variance of X.Math 370/408 Spring 2008 Actuarial Exam Practice Problem Set 4 Solutions 3. [43] ‘An actuary determines that the annual numbers of tornadoes in counties P and Q are jointly distributed as follows: “nna number of tomadoos ir county Q ofi]2 3 Annual number | 0 | 0.12 | 0.06 | 0.05 | 0.02 oftomadoes |1 [0.13015 | 012] 008 in county P |2{}0.05 | 015 | 0.10] 0.02 Calculate the conditional variance of the annual number of tornadoes in county Q, given that there are no tornadoes in county P. (A) 0.51 (B) 0.84 (c) 0.88 () 0.99 (&) 1.76 Answer: D: 0.99 Solution: We need to compute Var(Q|P =0), where P and Q denote, respectively, the number of tornadoes in counties P and Q. Now Var(Q|P = 0) = E(Q?|P = 0)-E(Q|P 0)2, and the two conditional expectations here are computed just like ordinary expectations, but with the conditional density of Q given P = 0 in place of the ordinary density of Q. This conditional density is obtained from the first row of the distribution table (corresponding to P = 0) by dividing by the row sum, 0.12 + 0.06 + 0.05 + 0.02 = 0.25, resulting in values 0.48, 0.24,0.20, 0.08 for Q = 0,1,2,3. Using these probabilities in computing the expectations £(Q2|P = 0) and E(Q|P = 0) gives the result, 0.99. 4. [44] |A diagnostic test for the presence of a disease has two possible outcomes: 1 for disease present and 0 for disease not present. Let X denote the disease state of a patient, and let Y denote the outcome of the diagnostic test. The joint probability function of X and Y is given by: (0.800. for (x,y) = (0,0), way =y) a) 0080 for vu) = (1,0), PERRY =W=) 9005. for (.u) = (0,1), 0.125 for (2,y) = (1,2). Calculate Vax(¥|X = 1) (A) 013 (B) 0.15, (©) 0.20 (0) 0.51 om Answer: C:0.20 Solution: We have Var(Y|X = 1) = E(¥#|X = 1) — E(Y[X = 1)?. The two conditional ‘expectations here are computed just ike ordinary expectations, but with hi(yjL) = h(ulX 1) = f(,y)/fx(1), the conditional density of ¥ given X = 1, in place of the ordinaryMath 370/408 Spring 2008 Actuarial Exam Practice Problem Set 4 Solutions 0,050 + 0.125 = 0.175, s0 h(y|1) equals 5/7 for y = 1. Hence density of Y. Now fx(1) = f(1,0) + £1) 0.050/0.175 = 2/7 for y = 0 and 0.125/0.175 E(YIX = i BW1x=1)=2 Var(¥|X = 1) = $ 5. (45) Let X and ¥ be random losses with joint density function fle.y) =eF” for z>Oand y >0. ‘An insurance policy is written to reimburse X +. Calculate the probability that the reimbursement is less than 1. ye (B) (1- (D) 1-26! (BE) 1-26? Answer: D Solution: The probability to compute is P(X +¥ < 1), which is given by the double integral of f(x,y) = e~*-¥ over the part of the first quadrant in which x+y <1. A sketch shows that this region is a triangular region described by the inequalities 0 < < 1,0 < yel—z. Therefore, pursvsae ff = [ere 6. [4-6] ‘The waiting time for the first claim from a good driver and the waiting time for the first claim from a bad driver are independent and follow exponential distributions with means 6 years and 3 years, respectively. What is the probability that the first claim from a good river will be filed within 3 years and the first claim from a bad driver will be filed within 2 years? (A) (le (B) yen* (1c eV pe! (D) Lee peoNS (E) 1- ge“ 1/2 4 e-1/8) 1794 ert/6 Answer: C Solution: ‘The probability to compute is P(X < 3,Y < 2), where X and Y denote the waiting times to the first claim for a good, resp. bad, driver. By the independence ‘of X and Y, this probability is equal to the product of the probabilities P(X < 3) and P(Y <2). Since X and ¥ ate exponentially distributed, with means 6 and 3, respectively, these probabilities are (1 — e~*/®) and (1—e~2/5). Multiplying these two expressions gives the correct answer, 1 — e~2/# —e-"/2 + @°1/6,Math 370/408 Spring 2008 Actuarial Exam Practice Problem Set 4 Solutions 7. 47] ‘A company has two electric generators. The time until failure for each generator follows fan exponential distribution with mean 10. ‘The company will begin using the second {generator immediately after the first one fails. What is the variance of the total time that the generators produce electricity? (A) 10 (B) 20 (©) 50 (D) 100 (E) 200 Answer: E: 200 Solution: We need to compute Var(X +¥), where X and ¥ denote the times that the two generators run. Assuming X and Y are independent, we have Var(X +Y) = Var(X)-+ Var(¥), Now X and Y are exponential with = = 10, so Var(X) = Var(Y) = 6? = 100, and hence Var(X +¥) = 200. Remark: The independence of X and Y here is not explicitly stated, but the only reason- able assumption given the context, and without this assumption the problem would not be doable. This is one of the rare cases (and, infact, the only case I am aware of) in which an independence assumption is not explicitly stated in the problem itself, but has to be made in order for the problem to be solvable. In general, if there is no explicit independence assumption in a problem, chances are that independence does not hold, so you should not make any assumptions about independence. 8, [48] A Joint density function is given by 10 e SomineE* (A) -1/6 (B)0 (C) 1/9 (D) 1/6 ©) 2/3 Answer: B: 0 First find the constant k by computing the integral of f(z, y) over the given range 0 1, Sketching this region, we see that it is described by the inequalities 0 1y>l. “Answer: E: 8/9 Solution: The probability sought is the conditional probability P(1 < Y < 3|X ‘To compute this, we first compute h(y|2), the conditional density of Y given X = 2: $2.9) (yay = L222, (WP) = 76) J0.0) = (Pw sx) [7 su) (20% a4 l 0y>024y<1, faye er Determine the probability that the portion of a claim representing damage to the house is less than 0.2. (A) 0.360 (B) 0.480 (C) 0.488 (D) 0512 (B) 0.520 Answer: C: 0.488 Solution: We need to find P(X < 0.2): This can be computed as a double integral over the joint density f(x,y) over the part of its range where < 0.2. Now the range of f(x,y) is the triangle given by 0<2<1,0 0.5). We are given that, X is uniform fn {0,1} and that, given X =z, ¥ is uniform on (z,2-+ 1). Thus, fx(z) =1 for 0<2<1 and N(yl2) =1 for 0.5 has area 1 — (1/2)0.5? = 7/8. Hence P(Y > 0.5) = (7/8)/1=7/8. , OSeSlesysrtl 2Math 370/408 Spring 2008 Actuarial Exam Practice Problem Set 4 Solutions 21. [4-55] ‘An insurance policy is written to cover a loss X, where X has density function }: for 0<2<2, 101 (PP has ‘The time (in hours) to process a claim of size x, where 0 3). From the problem, we know that, given X=, Y Ss uniformly isibuted on the interval fe, 2s), Thu, the conditional density of Y given X aris Ayla) = l/s for x 9 i described by the inequalities 3/2 < x < 2,3 2: Here the benefit is ¥ -2, and hence <5 if ¥ < 7, so the rectangle [0,1] x [2,7] has to be counted. The contribution of this area to Ris 5. 7 (iii) X > Land ¥ <2: Here the benefit is X —1, and hence < 6 if X < 6, so the rectangle [1,6] x [0,2] has to be counted. The contribution of this area to Reis 10. v (iv) X > Land ¥ > 2: Here the benefit is X+¥—3, and hence < 5 ifX-+Y < 8. Sketching this region, we see that itis a right triangle with side lengths 5, so it contributes an area of 52/2 = 12.5 to R. ‘Adding up the areas of regions corresponding to these cases gives 2 + 5+ 10+ 12.5 = 29.5 as the area of R. Dividing by the total area of the square (0, 10] x [0,10], we get the answer 0.295, 26. [4-106] ‘An insurance company sells two types of auto insurance policies: Basic and Deluxe, The time until the next Basic Policy claim is an exponential random variable with mean two days. The time until the next Deluxe Policy claim is an independent exponential random variable with mean three days. What is the probability that the next claim will be a Deluxe Policy claim? (a) 0.172 (B) 0.223, (©) 0.400 (D) 0487 (B) 0.500 Answer: C: 0.400 Solution: We need to compute P(Y < X), where X resp. Y denote the time until the next Basic resp. Deluxe Policy claim. We are given that X and Y are independent and have exponential distribution with means 2 and 3, respectively. Thus, the joint density of, X and Y is f(xy) = (alfvia) = pewbeW, O< 2 < 00,0 1). This probability is given by the double integral [Jy f(x,y)dedy, where f(r, y) is the given density and R is the part of the rectangle 0<2< 1,0 1. A direct computation of this double integral is feasible, but rather lengthy. It is easier to consider first the complementary probability, P(X +Y <1). The corresponding region of integration is the triangle 0 < x < 1, OSyS1—z, Thus, y pacsy n= f feao Lf (axraa-n- 2528) ae | ft (Saved) (Capea) gtatg) =a 708. [oo Glare aida 2 Hence P(X +¥ > 1) =1-7/24 28, [4-108] ‘The stock prices of two companies at the end of any given year are modeled with random variables X and Y that follow a distribution with joint density function or for01, 10 haa where 2 is the amount of a claim in thousands. Suppose 3 such claims will be made. What is the expected value of the largest of the three claims? (A) 2025 (B) 2700 (©) 3282 (0) 3375 (B) 4500 Answer: A: 2025 Solution: Let X1,X2,Xg denote the three claims, and let X* = max(Xy,X2,Xs) be the largest (maximum) of these claims. We need to compute E(X"). ‘To this end we first compute the c.f. F*(2) = P(X* <2) using the “maximum trick”: F'(z) (X* <2) = Plmax(X1, Xo, Xs) <2) = P(X, <2, Xe S 2, Xg <2) = P(Xs < 2)P(X2 < 2)P(Xa <2) = (ys, where F(x) is the ef, of a single Xj. Differentiating, we get the density of X* f*(@) = 3F (2)? P'(2) = 3F(z)*f(2). 7Math 370/408 Spring 2008 Actuarial Exam Practice Problem Set 4 Solutions 30. 31, 2-8 and therefore Now, f(t) =32~4 for 2 > 1, so F(x) = J 3t-4dt = +27) (221). Hence Since the units here are thousands, we get 2025 as answer. (4-110) A company insures homes in three cities, J, K, L. The losses occurring in these cities are independent. ‘The moment-generating functions for the loss distributions of the cities are Ma(t) = (1-273, M(t) = (1-207, Ma) = (0-2-4 Let X represent the combined losses from the three cities. Calculate E(X°). (A) 1,320 (B) 2,082 (©) 5,760 (D) 8,000 (B) 10,560 Answer: E: 10,560 Solution: Let J, K, L denote the losses from the three cities. Then X = J +K +L. Since J, K, L are independent, the moment-generating function for theit sum, X, is equal to the product of the individual moment-generating functions, ie., 1-24)", M(t) = Mic(t)My(®)Mr(t) = (1 = 24) 28-48 = Differentiating this fanction, we get Mie) = (-2)(-10)01 = 2)", M(t) = (-2)%(-10)(-11)(0 = 28)", M(t) = (~2)%(—20)(~11)(-12)(1 = 28-8. Hence, E(X) = Mi (0) = (—2)*(—10)(—11)(—12) = 10, 560. (41] ‘An insurance contract reimburses a family's automobile accident losses up to a maximum ‘of two accidents per year. The joint probability distribution for the number of accidents of a three person family (X,Y, Z) is p(,y.2) = K(x+2y+2), where 2 =0,1, y=0,1,2,2= 0,1, 2, and 2, y, 2 are the numbers of accidents incurred by X,Y, Z, respectively. Determine the expected number of unreimbursed accident losses given that X is not involved in any accidents, (A) 5/21 (B) 1/3 (©) 5/9 (D) 46/63 (E) 7/9 E7/9 There are a total of 18 possible values for (x,y,z). In principle, each of these would have to be considered. However, the condition “given that X is not involved in any accident” allows us to. simply ignore the z-variable in the joint p.m.f, leaving 18‘Math 370/408 Spring 2008 32, Actuarial Exam Practice Problem Set 4 Solutions (2.4.2) = p(0,y,2) = K(2y + 2), with 9 values for (y,). Adding up these 9 terms gives ke 21, 90 k= 1/27. Next, compute the expected number of unreimbursed losses. If the total number of acci- dents, ie., y+ 2, is < 2, all accident losses are reimbursed, so these cases do not contribute to this expectation, This leaves us with three cases for (y,2), and their corresponding probabilities: (0 (v2) = (1,2) (probability (2+ 1-+-2)/27 = 4/27); (i) (v2) = (2,1) (probability (2-2-4 1)/27 = 5/27); (Gi) (2) = (2,2) (probability (2-2 + 2)/27 = 6/27), In cases (i) and (ii) the total number of accidents is 3, so 1 accident is unreimbursed. In case (ii) the total number of accidents is 4, so 2 accidents are unreimbursed. Multiplying the number of unreimbursed accidents with the probabilities gives the desired expectation: 1 (4/27) +1 (6/21) + 2+ (6/21) = 7/9. (4112) Suppose the remaining lifetimes of a husband and wife are independent and uniformly distributed on the interval [0,40]. An insurance company offers two products to married couples: one which pays when the husband dies, and one which pays when both the husband and wife have died, Caleulate the covariance of the two payment times. (A) 00 (B) 44.4 (©) 66.7 (D) 200.0 (B) 466.7 Answer: C: 66.7 Solution: Let X and Y denote the lifetimes of the husband and wife. Then the time until both husband and wife have died is given by Z = max(X,Y), the larger of the two numbers. Thus, we need to compute Cov(X,Z) = E(XZ) ~ E(X)E(2). Oftthe three expectations here, £(X) is the easiest to find: Since X is uniformly distributed on (0,40), we have E(X) = 20. ‘The remaining expectations require double integrals computations. Since X and Y are independent, each uniformly distributed on [0,40], the joint density is equal to f(x,y) = (1/40)? on the square (0,40] x [0,40]. Thus, 1 10 p40 E(2) = Elmax(X,Y)) = hy [ _ [ _ menos ‘To calculate the latter integral, we need to first get rid of the “max” inside the integral ‘To this end, note frst that . max, y) = {r ecu sin® le a> eD> yelon@ Now, break the region of integration into two, according to whether x < y or x > y, and replace max(x,y) by y in the first of these regions (the part where x < y), and by in the second of these regions (the part where zr > y). These two regions represent the parts of the 0 by 40 square that lie above, respectively below, the diagonal x = y, as can be seen from a sketch. We thus get © YR [of nwstenmie [rates [Pye 0 [2m [ran fz=0 fy=O 3 rMath 370/408 Spring 2008 Actuarial Exam Practice Problem Set 4 Solutions 33. . [4113] Dividing by 40 we get E(Z) = E(max(X,Y)) = (2/3)40. A similar calculation, with fan additional factor z inside each integral gives E(XZ) = (3/8)40*, Hence Cov(X,¥) = (3/8)40? — 20(2/3)40 = 66.66. 115) Let X and ¥ denote the values of two stocks at the end ofa five-year period. X is uniformly distributed on the interval (0,12). Given X = 2, ¥ is uniformly distributed on the interval (0,2). Determine Cov(X,¥) according to this model. (ajo (B)4 (6 (D) 12 (E) 24 C6 ‘To find Cov(X,¥), we need to compute the three expectations E(XY), E(X) and E(Y’). Now E(X) = 6, since X is uniform on the interval (0,12). To find the remaining two expectations we need to first compute the joint density f(x,y). To this end we use the formula f(2,y) = h(ylz) fx (2). Since fx(2) is uniform on the interval (0,12), we have fx(c) = 1/12 for 0 < x < 12, Since the conditional density of ¥ given X = a is uniform ‘on (0,2), we have A(y[x) = 1/x for 0 0,p > 0. To get P(C < xP) we need to integrate this joint density over the region (inside the first quadrant of the cp-plane) in which ¢ < ap, or, equivalently, p 2 ¢/2. (Keep 20Math 370/408 Spring 2008 Actuarial Exam Practice Problem Set 4 Solutions in mind that 2 here is fixed—the variables are and p.) Thus, for x > 0, Pesapy= [0 [ane tnte [ccvinicn fF esintna - _ 20 ~ TF 1/@a) ~ T+ ‘This gives the e.dif. of X, F(a), for « > 0. Differentiating this function with respect to 2, the density of X: f(x) = F(x) = 2(1 + 22)? for x > 0. MP ne in OP gee with cb, tek ac a1