Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Chapter 4: Random Variables and Probability Distributions Hildebrand, Ott and Gray Basic Statistical Ideas for Managers Second Edition Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 1 2005 Brooks/Cole, a division of Thomson Learning, Inc. Learning Objectives for Ch. 4 • Defining a random variable and probability distribution • Determining the expected value of a random variable • Determining the standard deviation of a random variable • Defining the joint probability distribution of two random variables • Determining marginal probability distributions • Determining conditional probability distributions • Determining statistical independence • Determining the covariance and correlation • Determining the expected value and standard deviation of sums of random variables Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 2 2005 Brooks/Cole, a division of Thomson Learning, Inc. Section 4.1 Random Variable: Basic Ideas Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc. 3 4.1 Random Variable: Basic Ideas • Concept: A random variable assigns numerical values to the possible outcomes of a random experiment. Example: Today's closing price of a security relative to yesterday. Let Y=1, if price goes up; = 0, if price is unchanged; = -1, if price goes down. Y Up +1 y1 Same 0 y2 Down -1 y3 General notation • Notation: Y,X,Z (Upper case letters denote a r.v.) • Values are determined by chance. • Notation: y,x,z (Lower case letters denote a general value of a r.v.) Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 4 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.1 Random Variable: Basic Ideas • Types of random variables: discrete and continuous. • A discrete r.v. has a countable number of values. -1 • 0 Real no. line (Closing Price of a Security) +1 A continuous r.v. has an uncountable number of values. 8:45 Real no. line (Time at which you arrive for class) 9:00 Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 5 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.1 Random Variable: Basic Ideas Example (Game of Chuck-A-Luck): Three dice are rolled. You wager on any number from 1 to 6, say a ‘4’. If any one of the three dice comes up with a ‘4’, you win the amount of your wager. You also get back your original stake. It could happen that more than one dice comes up with a ‘4’. For example, if you had a $1 bet on a ‘4’, and each of the three dice came up with a ‘4’, you would win $3. If two of the three dice came up with a ‘4’, you would win $2. Let Y be the net amount you win in one play of the game. Possible values of Y: -1, 1, 2, 3. Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc. 6 Section 4.2 Probability Distribution: Discrete Random Variables Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 7 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.2 Probability Distribution: Discrete Random Variables • A probability distribution for a discrete r.v. Y is the collection of possible values of Y (denoted by y) and their probabilities, P[Y = y]. • Notation: P[Y = y] = PY(y) • Possible representations of a probability distribution: • Table • Graph • Mathematical Formula Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 8 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.2 Probability Distribution: Discrete Random Variables Example (Game of Chuck-A-Luck): • Y is the net amount won in one play of the game. • Possible values of Y: -1, 1, 2, 3 Find the probability distribution of Y and also find P(Y ≤ 1). Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc. 9 4.2 Probability Distribution: Discrete Random Variables Probability Distribution of Y Py(Y) 0.6 -1 (5/6) (5/6) (5/6) 1 3 (1/6) (5/6) (5/6) = 75/216 2 3 (1/6) (1/6) (5/6) = 15/216 3 (1/6) (1/6) (1/6) 0.5 = 125/216 = 1/216 0.4 P(Y=y) y 0.3 0.2 Sum = 1 0.1 0.0 -1 0 1 y 2 3 P(Y ≤ 1) = P(Y=-1) + P(Y=+1) = 200/216 Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 10 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.2 Probability Distribution: Discrete Random Variables • Requirements for a probability distribution: • PY(y) ≥ 0, for all y • ∑ P ( y ) =1 Y all y Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 11 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.2 Probability Distribution: Discrete Random Variables • The cumulative distribution function (cdf) of a random variable Y is denoted FY(y). • Gives the probability that Y is less than or equal to y, for all values of y. • Need to find P(Y ≤ y). Example (Game of Chuck-A-Luck): Find the cumulative distribution function of Y and the P(Y ≤ 1) by using the cdf. Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc. 12 4.2 Probability Distribution: Discrete Random Variables Recall Probability Distribution of Y 0.6 Py(Y) -1 125/216 1 75/216 2 15/216 3 1/216 Sum 1 0.5 0.4 P(Y=y) y 0.3 0.2 0.1 0.0 -1 0 1 y 2 3 FY(y) For y < - 1, FY(y) = 0 For -1 ≤ y < 1, FY(y) = 125/216 For 1 ≤ y < 2, FY(y) = 200/216 1 215/216 200/216 125/216 For 2 ≤ y < 3, FY(y) = 215/216 For y ≥ 3, FY(y) = 1 -1 0 1 2 Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 3 y 13 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.2 Probability Distribution: Discrete Random Variables • Find P(Y ≤ 1) • P(Y ≤ 1) = FY(1) = 200/216 • Same as before: P(Y ≤ 1) = PY(-1) + PY(1) = 200/216 Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 14 2005 Brooks/Cole, a division of Thomson Learning, Inc. Section 4.3 Expected Value and Standard Deviation Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc. 15 4.3 Expected Value and Standard Deviation • The expected value (or mean) of a discrete random variable Y is µY = E[Y] = ∑ y PY(y) Interpretation: Long-run average Example (Game of Chuck-A-Luck): Find the expected value of Y and provide an interpretation. Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 16 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.3 Expected Value and Standard Deviation y PY(y) y PY(y) -1 125/216 -(125/216) 1 75/216 75/216 2 15/216 30/216 3 1/216 3/216 E(Y) = -(17/216) = - $.079 Interpretation: In the long run, you expect to lose about 8 cents out of every dollar wagered. Suppose a game is defined to be fair if its expected value is 0. Is the game of Chuck-A-Luck fair? No! Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 17 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.3 Expected Value and Standard Deviation • The variance of a discrete random variable Y is 2 2 σ Y = V(Y) = ∑(y - µ Y ) PY ( y) • An alternative computational formula for the variance is 2 2 2 σ Y = ∑ y P Y ( y) - µ y all y • The standard deviation σY is the positive square root of the variance. σ Y = σ Y2 Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc. 18 4.3 Expected Value and Standard Deviation • A distribution that is compact has a small standard deviation, while a distribution that is dispersed has a large standard deviation. Example: Matching fair coins Two players each toss a coin. The equally likely outcomes are: (H,H), (H,T) (T,H), (T,T). One player calls “odds” or “evens.” If the player calls “odds” and the coins differ, the player wins. If the player calls “evens” and the coins match, the player wins. Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 19 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.3 Expected Value and Standard Deviation Suppose each player wagers $1. Let X be the net amount you win. Find σX. PX(x) .5 -1 0 1 x E(X) = 0 V(X) = (-1-0)2 (.5) + (1-0)2 (.5) = 1 ($)2 σX = $1 Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 20 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.3 Expected Value and Standard Deviation Suppose each player wagers $10. Let Y be the net amount you win. Find σY. PY(y) .5 -10 0 10 y E(Y) = 0 V(Y) = (-10-0)2 (.5) + (10-0)2 (.5) = 100 ($)2 σY = $10 Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc. 21 4.3 Expected Value and Standard Deviation • The standard deviation measures the dispersion of the probability distribution relative to the expected value. Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 22 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.3 Expected Value and Standard Deviation Example (Game of Chuck-A-Luck): Find the standard deviation of Y. (y -µ)2 17 − 1 − − 216 17 1 − − 216 2 2 17 2 − − 216 17 3 − − 216 2 2 PY(y) (y - µ)2 PY(y) 125 216 199 125 − 216 216 75 216 233 75 216 216 2 2 2 15 216 449 15 216 216 1 216 665 1 216 216 2 1.239 ($) 2 σ = V(Y) = $1.11 Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 23 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.3 Expected Value and Standard Deviation Example (Game of Chuck-A-Luck): Find the probability that Y takes on a value within one standard deviation of the mean. P[ µ − σ ≤ Y ≤ µ + σ ] = P[ -.08 − 1.11 ≤ Y ≤ -.08 + 1.11 ] = P[ -1.19 ≤ Y ≤ 1.03 ] = P(Y=-1) + P(Y=1) = 200/216 Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc. 24 4.3 Expected Value and Standard Deviation Example: A financial analyst is trying to choose between two investments. Typical returns on the first investment, denoted by X, are: 0% and 20%, with probabilities 0.5 for each. Typical returns on the second investment, denoted by Y, are: -10%, 10%, 20%, and 30%, with probabilities 0.15, 0.35, 0.35 and 0.15, respectively. These probabilities have been subjectively determined. The probability distributions of X and Y follow. x 0% 20% P(X=x) 0.50 0.50 y -10% 10% 20% 30% P(Y=y) 0.15 0.35 0.35 0.15 Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 25 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.3 Expected Value and Standard Deviation a. What is the expected return for investment 1? What is the expected return for investment 2? Obviously, E(X) = 10% { The probability distribution of X is symmetric about 10. } E(Y) = -10(0.15) + 10(0.35) + 20(0.35) + 30(0.15) = 13.5% Using the expected value criterion, which investment would you choose? Answer: Choose Y. Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 26 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.3 Expected Value and Standard Deviation b. Calculate the standard deviation of the percentage returns for investments 1and 2. 2 ( xi − µ X ) V(X) = ∑ i =1 = (0 – 2 PX ( xi ) 10)2(.5) + (20 –10)2(.5) = 100 ⇒ σX = 10% 4 ( yi − µY ) PY ( yi ) V(Y) = ∑ i =1 = (-10 – 13.5)2( .15 ) + (10 – 13.5)2( .35 ) + (20 – 13.5)2(.35 ) + (30 – 13.5)2( .15) 2 = 142.75 ⇒ σY = 11.95% Using the criterion of minimizing risk, which investment would you choose? Answer: Choose X. Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc. 27 4.3 Expected Value and Standard Deviation c. Using both the mean and standard deviation, which investment would you choose? Are you a risk seeker? Risk averse? Risk neutral? Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 28 2005 Brooks/Cole, a division of Thomson Learning, Inc. Section 4.4 Joint Probability Distributions and Independence Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 29 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.4 Joint Probability Distributions and Independence • The joint probability distribution for two discrete random variables X and Y, denoted by PXY (x,y), is the possible values of (x, y) together with their joint probabilities. • Notation: PXY (x,y) = P(X = x, Y = y) Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc. 30 4.4 Joint Probability Distributions and Independence Example: Portfolio analysis $1 is available to invest and there are two securities under consideration. How should one allocate the $1 between the two securities? The return on the first security, denoted by X, has two possible values: 0% and 20%. The return on the second security, denoted by Y, has four possible values: -10%, 10%, 20% and 30%. • The joint probability distribution follows. Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 31 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.4 Joint Probability Distributions and Independence • The joint probability distribution of X and Y is given in the table below. y x -10 10 20 30 0 .10 .20 .15 .05 20 .05 .15 .20 .10 • Let P[X = x, Y = y] denote the joint probability that the random variable X takes on the value x and, at the same time, the random variable Y takes on the value y. • Thus, P[X = 0, Y = -10] = .10 Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 32 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.4 Joint Probability Distributions and Independence • General requirements for a discrete joint probability distribution: • • PXY(x,y) ≥ 0 ∑∑ P XY y Probability can’t be negative. ( x, y ) = 1 All the probabilities must add to 1. x Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc. 33 4.4 Joint Probability Distributions and Independence • The marginal probability distribution of X, denoted PX(x), is the probability distribution of X by itself. • Need to find PX ( x) = ∑ PXY ( x, y) for each value of X. all y • Similarly, PY ( y) = ∑ PXY ( x, y) for each value of Y. all x Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 34 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.4 Joint Probability Distributions and Independence Example: Portfolio analysis a. Find the marginal distribution of the return (X) for security 1. The joint probability distribution is given below: y x -10 10 20 30 0 .10 .20 .15 .05 20 .05 .15 .20 .10 P[X = 0] = ∑ P[X = 0, Y = y] = .10 +.20 +.15 +.05 = .50 y P[X = 20] = ∑ P[X = 20, Y = y] = .05 +.15 + .20 +.10 =.50 y Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 35 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.4 Joint Probability Distributions and Independence The marginal probability distribution of X is: x P[X=x] 0 .5 20 .5 Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc. 36 4.4 Joint Probability Distributions and Independence b. Find the marginal distribution of the return (Y) for security 2. y x 0 -10 10 20 30 .10 .20 .15 .05 20 .05 .15 .20 .10 P[Y=y] .15 .35 .35 .15 Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 37 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.4 Joint Probability Distributions and Independence c. What are the expected returns for securities 1 and 2 ? Previously we saw that E(X) = 10% and E(Y) = 13.5%. d. What are the standard deviations of the returns for securities 1 and 2 ? Previously we saw that σX = 10% and σY = 11.95%. • Only need the marginal distributions to find the expected values and standard deviations of X and Y, respectively. Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 38 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.4 Joint Probability Distributions and Independence • The conditional probability distribution of X given that Y = y is denoted PX|Y (x | y). • Need to find PX| Y ( x | y) = PXY ( x, y) for each value of X. PY ( y) • Similarly, the conditional probability distribution of Y, given that X=x, is PY|X ( y | x) = PXY ( x, y) PX ( x) Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc. 39 4.4 Joint Probability Distributions and Independence Example: Portfolio analysis e. Find the conditional probability distribution of the return for security 2, if the return for security 1 is 20%. The joint and marginal probability distributions are given below: y x 0 20 P[Y=y] -10 10 20 30 P[X=x] .10 .20 .15 .05 .50 .05 .15 .20 .10 .50 .15 .35 .35 .15 Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 40 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.4 Joint Probability Distributions and Independence • General expression: P[Y = y | X = 20] = P[Y = y and X = 20] P[X = 20] • In table format: y -10 10 20 30 P[Y=y|X=20] .05/.50 =.10 .15/.50 =.30 .20/.50 =.40 .10/.50 =.20 Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 41 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.4 Joint Probability Distributions and Independence • How many conditional distributions of Y are there? One for each value of X. • How many conditional distributions of X are there? One for each value of Y. Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc. 42 4.4 Joint Probability Distributions and Independence • Sequence of generating the distributions Joint Probability Distribution Marginal Distributions Conditional Distributions • Can we go the other way? Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 43 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.4 Joint Probability Distributions and Independence • Random variables X and Y are independent if and only if PXY (x, y) = PX (x) PY(y) for all (x, y). • Equivalently, X and Y are independent if and only if PY|X (y | x) = PY (y) for all (x, y). Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 44 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.4 Joint Probability Distributions and Independence Example: Portfolio Analysis f. Are the returns from securities 1 and 2 independent? Is P[Y = -10 | X = 20] = P[Y = -10]? .10 .15 Therefore, the returns from securities 1 and 2 are not independent. Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc. 45 Section 4.5 Covariance and Correlation of Random Variables Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright 46 2005 Brooks/Cole, a division of Thomson Learning, Inc. © 4.5 Covariance and Correlation of Random Variables • Concept: Covariance measures the degree of dependence between 2 random variables • The covariance of X and Y is denoted by Cov(X, Y). • Need to find Cov(X, Y) = ∑ ∑ ( x - µ X )( y - µ Y ) PXY ( x, y) x y • An alternative computational formula is Cov(X, Y) = ∑ ∑ x y PXY ( x, y) - µ X µ Y x y Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 47 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.5 Covariance and Correlation of Random Variables Example: Portfolio analysis g. Find the covariance between the returns for securities 1 and 2. y x -10 10 20 30 0 .10 .20 .15 .05 20 .05 .15 .20 .10 Cov(X,Y) = ∑xy PXY(x, y) - µx µy = 160 – (10)(13.5) = 25 {Units ?} where ∑xy PXY (x, y) = 0 + 0 + 0 + 0 + (20)(-10)(.05) + (20)(10)(.15) + (20)(20)(.20) + (20)(30)(.10) Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc. 48 4.5 Covariance and Correlation of Random Variables • Since the covariance is positive, the returns on the two securities tend to move in same direction. • There is a tendency for both returns to be below their means at the same time or above their means at the same time. Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 49 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.5 Covariance and Correlation of Random Variables • Standardizing the covariance • The coefficient of correlation is denoted by ρ. • Need to find ρ = Cov( X , Y ) σ Xσ Y • Measures the strength of the linear relationship between X and Y. • Property: −1 ≤ ρ ≤ +1 Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 50 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.5 Covariance and Correlation of Random Variables Example: Portfolio analysis h. Find the correlation between the returns for securities 1 and 2. ρ= Cov( X , Y ) σ XσY 25 (10)(11.95) = = 0.21 Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc. 51 4.5 Covariance and Correlation of Random Variables Note: If X and Y are independent, then Cov(X, Y) = 0 and ρ = 0. Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 52 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.5 Covariance and Correlation of Random Variables Example: Consider the following joint distribution. Find the covariance between X and Y. Are X and Y independent? Y X • E(X) = 0, by symmetry of PX(x) 1 4 -2 0 1/4 -1 1/4 0 1 1/4 0 2 0 1/4 • Cov(X, Y) = ∑ ∑ xy PXY(x, y) - µX µY = 0 •ρ=0 • Are X and Y independent? No, because PXY(-2,1) ≠ PX(-2)PY(1) 0 ≠ (1/4)(1/2) • Are X and Y related? Yes, because Y = X2 Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 53 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.5 Covariance and Correlation of Random Variables Extensions: For any two constants b and c, E(bX + cY) = bE(X) + cE(Y) V(bX + cY) = b2 V(X) + c2 V(Y) + 2 bc Cov(X,Y) Equivalently, V(bX + cY) = b2 V(X) + c2 V(Y) + 2 bc σxσy ρ. Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc. 54 4.5 Covariance and Correlation of Random Variables Example: Portfolio analysis i. Suppose $0.40 is invested in security 1 and $0.60 in security 2. What is the expected value of the percentage return on the portfolio? How does this compare to the expected value for each security? E[.4 X + .6 Y] = .4E(X) + .6E(Y) = .4(10%) + .6(13.5%) = 12.1% vs. E(X) = 10% and E(Y) = 13.5% 55 Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.5 Covariance and Correlation of Random Variables Example: Portfolio analysis j. What is the standard deviation of the percentage return on the portfolio? How does this compare to the risks of the individual securities? V[.4 X + .6 Y] = (.4)2V(X) + (.6)2V(Y) + 2(.4)(.6)(25) = 79.39 (%)2 ⇒ σPORTFOLIO = 79.39 = 8.91% vs. σX = 10% and σY = 11.95% • Moral: “Don’t put all your eggs in one basket.” 56 Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.5 Covariance and Correlation of Random Variables Example: Portfolio analysis k. Sketch the investment frontier and specify your location as an investor. %Asset1 E(X) %Asset2 E(Y) E(P) Cov(X,Y) V(P) SD(P) 1 10% 0 13.50% 10.00% 25 100 10.00% 0.9 10% 0.1 13.50% 10.35% 25 86.92803 9.32% 0.8 10% 0.2 13.50% 10.70% 25 77.7121 8.82% 0.7 10% 0.3 13.50% 11.05% 25 72.35223 8.51% 0.6 10% 0.4 13.50% 11.40% 25 70.8484 8.42% 0.5 10% 0.5 13.50% 11.75% 25 73.20036 8.56% 0.4 10% 0.6 13.50% 12.10% 25 79.4089 8.91% 0.3 10% 0.7 13.50% 12.45% 25 89.47323 9.46% 0.2 10% 0.8 13.50% 12.80% 25 103.3936 10.17% 0.1 10% 0.9 13.50% 13.15% 25 121.17 11.01% 0 10% 1 13.50% 13.50% 25 142.8025 11.95% Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc. 57 4.5 Covariance and Correlation of Random Variables Example: Portfolio analysis The investment frontier is shown below. Where are you? Graph of Expected Value vs. Standard Deviation 14% 12% Expected Value 10% 8% Ignore? 6% 4% 2% 0% 0% 2% 4% 6% 8% 10% 12% Standard Deviation Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 58 2005 Brooks/Cole, a division of Thomson Learning, Inc. 4.5 Covariance and Correlation of Random Variables Extensions (cont'd): • For X - Y, b = +1 and c = -1 E(X - Y) = E(X) – E(Y) V(X - Y) = V(X) + V(Y) -2Cov(x,y) • If X and Y are independent, V(bX + cY) = b2V(X) + c2V(Y) Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 59 2005 Brooks/Cole, a division of Thomson Learning, Inc. Keywords: Chapter 4 • • • • • • • • • • • • Random variable Probability distribution Cumulative distribution function Expected value Standard deviation Joint probability distribution Marginal probability distribution Conditional probability distribution Covariance Correlation Expected value for sums of random variables Variance for sums of random variables Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc. 60 Summary of Chapter 4 One Random Variable • General Case – Discrete Probability Distribution y PY(y) • • • • • • µ = E(Y) = ΣyPY(y) σ2 = V(Y) = Σ(y - µ)2PY(y) Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 61 2005 Brooks/Cole, a division of Thomson Learning, Inc. Summary of Chapter 4 Two or More Random Variables • General Case – Discrete Joint Probability Distribution Y • • • X • • • Joint probabilities are in the cells of the table. Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 62 2005 Brooks/Cole, a division of Thomson Learning, Inc. Summary of Chapter 4 • Joint dist. → Marginal dist. • Joint dist. & Marginal dist. Conditional dist. P(X=x|Y=y) = P(X=x,Y=y)/P(Y=y) Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc. 63 Summary of Chapter 4 • Cov(X,Y) : Do X and Y tend to vary together? If there is a tendency for both Y and X to be below (or above) their means at the same time, then Cov(X,Y) > 0. • Corr(X,Y) = Cov(X,Y)/σXσY Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 64 2005 Brooks/Cole, a division of Thomson Learning, Inc. Summary of Chapter 4 • X and Y are independent if P(X = x, Y = y) = P(X = x) P(Y = y) • If X and Y are independent, Cov(X,Y) = 0 Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 65 2005 Brooks/Cole, a division of Thomson Learning, Inc. Summary of Chapter 4 • E(aX + bY) = aE(X) + bE(Y) • V(aX + bY) = a2V(X) + b2 V(Y) + 2 ab Cov(X,Y) Hildebrand, Ott & Gray, Basic Statistical Ideas for Managers, 2nd edition, Chapter 4 Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc. 66