You are on page 1of 17


Capital Market Theory

Chapter 20 follows Chapter 19 because capital market theory builds on portfolio theory by examining how asset prices are determined in a world of Markowitz diversifiers Chapter 20 also contains some important concepts relevant to a better understanding of such topics as systematic and nonsystematic risk and beta !he first part of Chapter 20 outlines the necessary assumptions to derive capital market theory and introduces the concept of e"uilibrium in the capital markets #mportant related concepts are introduced and discussed$ primarily the market portfolio %oth its importance and its composition are considered &sing the concepts developed to this point$ the e"uilibrium risk''return tradeoff is analyzed in detail !he capital market line is developed and illustrated !his line applies to efficient portfolios$ with the slope of the line showing the market price of risk for efficient portfolios !he e"uation is explained$ and certain points about the line are emphasized !he security market line is developed next !he e"uation is developed$ and beta as a measure of volatility is considered in some detail !he C()M*s expected return''beta relationship is analyzed$ with each of the components of the re"uired rate of return analyzed and explained !he process of identifying undervalued and overvalued securities using the +M, follows this discussion )roblems in estimating the +M, are described$ and this leads into a detailed discussion of the accuracy of beta estimates and tests of the C()M !he characteristic line is also explained !he chapter concludes with a complete discussion of (rbitrage )ricing !heory in terms of what beginners need to know (lthough this concept probably has not advanced in terms of being widely used in the investments world as much as some have predicted$ it is an important development that can be used 28

for discussion purposes if the instructor so chooses -actor models are explained as part of this discussion ( reasonably detailed discussion on understanding the ()! is included Consistent with the emphasis in this text$ the use of ()! in investment decisions is considered +tudents should be able to see how the model could be applied in actual practice NOTE. !he discussion of ()! probably contains all that beginners need to know and can reasonably handle CHAPTER OBJECTIVES !o develop the concept of asset pricing theory as a natural extension of portfolio theory !o develop the concepts of the CM, and +M,$ explain what they mean$ and consider how they can be used !o discuss related issues such as what beta measures and the problems with estimating beta$ systematic and nonsystematic risk$ problems in testing asset pricing models$ and so forth !o provide the necessary information about ()!$ including what it means and how it could be used to make investment decisions


MAJOR CHAPTER HEA IN!S "Co#te#t$% The A$$&'ptio#$ O( The CAPM /assumptions used to derive capital market theory0

1"uilibrium in the Capital Markets /2hat are the implications of the assumptions30

The Market Port(olio /definition4 graph showing the market portfolio on the efficient frontier0

!he #mportance of the Market )ortfolio /2hy do all investors hold identical risky portfolios30 Composition of the Market )ortfolio /what it consists of4 limitations0

The E)&ili*ri&' Ret&r#++Ri$k Tra,eo(( /C()M is an e"uilibrium relationship encompassing both the CM, and the +M,0

!he Capital Market ,ine /e"uation4 diagram4 example4 important points to note about the CM,0 !he +ecurity Market ,ine /deriving the e"uation4 graph4 what beta is4 the C()M5s expected return''beta relationship4 over'and'undervalued securities with explanation and diagram0

E$ti'ati#- the SM. /estimating the variables in the e"uation0

1stimating %eta

/the accuracy of beta estimates4 characteristic line for Coca'Cola4 problems when estimating beta4 stability of portfolio beta0 Te$t$ O( The CAPM /predictions of the model4 findings0 Ar*itra-e Pri/i#- Theory /basic description of the model4 assumptions4 factor models4 characteristics of factors4 example of factor model4 e"uations for actual return and expected return4 factors that have been identified through research0

&sing ()! in #nvestment 6ecisions /possible strategies when using ()! to make decisions0 +ome Conclusions (bout (sset )ricing /controversy remains0

POINTS TO NOTE ABO0T CHAPTER 20 Ta*le$ a#, 1i-&re$ !here are no tables or boxed inserts in Chapter 20 !he four ma7or figures in this chapter are all standard figures of the efficient frontier with borrowing and lending$ the CM,$ the +M,$ and so forth (s such$ they are interchangeable with virtually any other comparable figures that an instructor may already have developed !hey are not uni"ue although they are keyed to the discussion in the text -igure 20'1 shows the Markowitz efficient frontier and the borrowing and lending possibilities resulting from introducing a risk'free asset !his figure is the same as -igure 8'9 except the point of tangency has been changed from ! to M to emphasize the market portfolio and its importance

-igure 20'2 shows the capital market line and the components of its slope #t is important to emphasize that standard deviation is on the horizontal axis and to emphasize what the slope of this line measures -igure 20': shows the +M, !he emphasis now is on beta as the measure of risk on the horizontal axis -igure 20'; illustrates how an overvalued and an undervalued security can be identified by using the +M, #t could be pointed out here that in some sense this is how to think of modern security analysis''the search for securities not on the e"uilibrium tradeoff that should exist -igure 20'9 shows the characteristic line for Coca'Cola$ using monthly data

ANSWERS TO EN +O1+CHAPTER 20ESTIONS 20+34 .e#,i#- po$$i*ilitie$ change part of the Markowitz efficient frontier from an arc to a straight line !he straight line extends from <-$ the risk'free rate of return$ to M$ the market portfolio !his new opportunity set$ which dominates the old Markowitz efficient frontier$ provides investors with various combinations of the risky asset portfolio M and the riskless asset Borro5i#- po$$i*ilitie$ complete the transformation of the Markowitz efficient frontier into a straight line extending from <- through M and beyond #nvestors can use borrowed funds to lever their portfolio position beyond point M$ increasing the expected return and risk beyond that available at point M 20+24 &nder the C()M$ all investors hold the market portfolio because it is the optimal risky portfolio %ecause it produces the highest attainable return for any given risk level$ all rational investors will seek to be on the straight line tangent to the efficient set at the steepest point$ which is the market portfolio !he basic difference between graphs of the +M, and the CM, is the label on the horizontal axis -or the CM,$ it is standard deviation while for the +M,$ beta (lso$ the CM, is applicable to portfolios while the +M, applies to individual securities and to portfolios #n theory$ the 'arket port(olio =portfolio M> is the portfolio of all risky assets$ both financial and real$ in their proper proportions +uch a portfolio would be completely diversified4 however$ it is a risky portfolio #n e"uilibrium$ all risky assets must be in portfolio M because all investors are assumed to hold the same risky portfolio #f they do$ in e"uilibrium this portfolio must be the market portfolio consisting of all risky assets 20+84 !he slope of the CML is



1<M ' <???????? +6M where 1<M is the expected return on the market M portfolio$ <- is the rate of return on the risk'free asset$ and +6M is the standard deviation of the returns on the market portfolio The slope of the CML is the market price of risk for efficient portfolios; that is, it indicates the equilibrium price of risk in the market #t shows the additional return that the market demands for each percentage increase in a portfolio5s risk 20+94 !he CM, extends from <-$ the risk'free asset$ through M$ the market portfolio of all risky securities =weighted by their respective market values> !his portfolio is efficient$ and the CM, consists of combinations of this portfolio and the risk'free asset (ll asset combinations on the CM, are efficient portfolios consisting of M and the risk'free asset !he contribution of each security to the standard deviation of the market portfolio depends on the size of its covariance with the market portfolio !herefore$ investors consider the relevant measure of risk for a security to be its covariance with the market portfolio &sing some methodology =such as the dividend valuation model> to estimate the expected returns for securities$ investors can compare these expected returns to the re"uired returns obtained from the +M, +ecurities whose expected returns plot above the +M, are undervalued because they offer more expected return than investors re"uire4 if they plot below the +M,$ they are overvalued because they do not offer enough expected return for their level of risk 2hen a security is recognized by investors undervalued$ they will purchase it because more return than re"uired$ given its risk will drive up the price of the security as purchased !he return will be driven down as it offers !his demand more of it is until it




reaches the level indicated by the +M, as appropriate for its degree of risk 20+304 !he difficulties involved in estimating a security5s beta include deciding on the number of observations and the length of the periods to use in calculating the beta !he regression estimate of beta is only an estimate of the true beta$ and sub7ect to error (lso$ the beta is not perfectly stationary over time !he ma7or problem in testing capital market theory is that the theory is formulated ex'ante$ concerning what is expected to happen !he only data we typically have are ex'post !he C()M can be tested empirically by regressing the average return on security i over some number of periods on security #*s beta !his is usually done for a large number of securities !he e"uation involved is. @ <i A a1 B a2bi @ where <i is the average return on security i over some number of periods and bi is the estimated beta for security i !he expected results of regressing average returns on beta are that a1 should approximate the average risk'free rate during the period studied and a2 should approximate the average market risk premium during the period studied =<M ' <-> 20+364 20+374 !he law of one price states that two otherwise identical assets cannot sell at different prices <oll has argued that the C()M is untestable because the market portfolio$ which consists of all risky assets$ is unobservable !he C()M is a useful model for estimating re"uired returns !hese re"uired returns can be used in con7unction with independently derived expected returns to determine overvalued and undervalued securities !his model is also useful in estimating the cost of e"uity capital for a security (nd$ as we shall see in




Chapter 22$ the C()M provides a basis for measuring portfolio performance 20+394 !he CM, is drawn tangent to the Markowitz efficient frontier 2hen this is done$ it can be seen that the CM, dominates the Markowitz efficient frontier !he CM, is a straight line tangent to the efficient frontier at point M$ the market portfolio$ and with an intercept of <#nvestors decide where they are to be on the new efficient frontier =the straight line dominating the Markowitz efficient frontier> by their risk preferences #f they are conservative$ they will be on the lower end of the line toward <-4 if aggressive$ they will be on the upper end$ which represents larger expected returns and larger risks !he CM, is the trade off between expected returns and risk for efficient portfolios !he slope of the CM, indicates the e"uilibrium price of risk in the market ( diagram of the +M, is simply an upward'sloping tradeoff between expected return on the vertical axis and risk as measured by beta on the horizontal axis #n effect$ this is a diagram of the whole concept of investing$ which is$ in fact$ best described simply as an upward'sloping tradeoff between expected return and risk =a> #f the risk'free rate shifts upward$ and nothing else changes$ the diagram would show a new upward' sloping line above the old line$ running parallel with it !he difference between the two vertical intercepts would reflect the increase in the risk' free rate #n this case$ the +M, would rotate upward to the left to reflect a greater tradeoff (s investors become pessimistic$ the line becomes steeper =rotates upward to the left>4 as they become optimistic$ the line rotates downward to the right$ approaching a horizontal line at the extreme






&nlike the C()M$ ()! does not assume.

1 2 : ;

a single'period investment horizon the absence of taxes borrowing and lending at the rate <investors select portfolios on the basis of expected return and variance

()!$ like the C()M$ does assume. 1 2 : ; 20+234 investors have homogeneous beliefs investors are risk'averse utility maximizers markets are perfect returns are generated by a factor model

( (a/tor 'o,el is based on the view that there are underlying risk factors that affect realized and expected security returns !hese risk factors represent broad economic forces and not company'specific characteristics and$ by definition$ they represent the element of surprise in the risk factor''the difference between the actual value for the factor and its expected value !he factors must possess three characteristics. 1 1ach risk factor must have a pervasive influence on stock returns -irm'specific events are not ()! risk factors !hese risk factors must influence expected return$ which means they must have non'zero prices !his issue must be determined empirically$ by statistically analyzing stock returns to see which factors pervasively affect returns (t the beginning of each period$ the risk factors must be unpredictable to the market as a whole


: 20+264

Most empirical work suggests that three to five factors influence security returns and are priced in the market -or example$ <oll and <oss identify five systematic factors. 1 2 : ; changes in expected inflation unanticipated changes in inflation unanticipated changes in industrial production unanticipated changes in the default'risk premium

9 20+274 20+284

unanticipated changes in the term structure of interest rates

( factor model makes no statement about e"uilibrium ( portfolio manager could design strategies that would expose them to one or more types of these risk factors$ or CsterilizeD a portfolio such that its exposure to the unexpected change in the growth rate of profits matched that of the market as a whole !aking an active approach$ a portfolio manager who believes that he or she can forecast a factor realization can build a portfolio that emphasizes or deemphasizes that factor #n doing this$ the manager would select stocks that have exposures to the remaining risk factors that are exactly proportional to the market #f the manager is accurate with the forecast''and remember that such a manager must forecast the unexpected component of the risk factor''he or she can outperform the market for that period ()! is not critically dependent on an underlying market portfolio as is the C()M$ which predicts than only market risk influences expected returns #nstead$ ()! recognizes that several types of risk may affect security returns (n arbitrage profit$ in the context of the ()!$ refers to a situation where a zero investment portfolio can be constructed that will yield a risk'free profit #f arbitrage profits arise$ a relatively few investors can act to restore e"uilibrium 2e can evaluate how each security affects the standard deviation of the market portfolio by evaluating the way it would change if the proportion invested in a particular security changes =#n effect$ we take the partial derivative of the standard deviation of the market portfolio with respect to the proportion of portfolio funds invested in that particular security > !he result is that a security*s contribution to the risk of the market portfolio is given by. i$M ''' M




where i$M A the covariance between stock i and the market portfolio C1A 20+2<4 (ny three of the following are criticisms of beta as used in C()M 1 !heory does not measure up to practice #n theory$ a security with a zero beta should give a return exactly e"ual to the risk'free rate %ut actual results do not come out that way$ implying that the market values something besides a beta measure of risk %eta is a fickle short'term performer +ome short'term studies have shown risk and return to be negatively related -or example$ %lack$ Eensen and +choles found that from (pril 1998 through 6ecember 19F9$ securities with higher risk produced lower returns than less risky securities !his result suggests that =a> in some short periods$ investors may be penalized for taking on more risk4 =b> in the long run$ investors are not rewarded enough for high risk and are overcompensated for buying securities with less risk4 and =c> in all periods$ some unsystematic risk being valued by the market 1stimated betas are unstable Ma7or change in a company affecting the character of the stock$ some unforeseen event not reflected in past returns may decisively affect the security*s future returns %eta is easily rolled over <ichard <oll has demonstrated that by changing the market index against which betas are measured$ one can obtain "uite different measures of the risk level of individual stocks and portfolios (s a result$ one would make different predictions about the expected returns$ and by changing indexes$ one could change the risk'ad7usted performance ranking of a manager

C1A 20+604

&nder C()M$ the only risk that investors should be compensated for bearing is the risk that cannot be diversified away =systematic risk> %ecause systematic

risk =measured by beta> is e"ual to one for both portfolios$ an investor would expect the same return for )ortfolio # and ## +ince both portfolios are fully diversified$ it doesn5t matter if the specific risk for each individual security is high or low !he specific risk has been diversified away for both portfolios C1A 204634 C1A 204624 C1A 204664 C1A 204674 C1A 204684 C1A 20+694 b a d b d !he following comments are presented in the same order as the points made by +tatdud. 1 !his statement is incorrect +ince monthly observations =returns> are employed$ the constant =alpha> value of 99 indicates that if the return on the +G) 900 were zero in a given month$ the monthly return on the stock would tend to be 99H 6uring a year when the return on the +G) 900 is zero$ the annual return on the stock can be approximated as. = 99H> x =12> A 8 0IH !his statement is incorrect !he alpha value of 99 is the y'interest and represents a return on CC1 that is unrelated to the return on the market Jariability of the market is measured by variance or standard error of the estimate !he statement regarding the slope coefficient and the volatility of the return on CC1*s common stock

relative to the market is correct$ assuming that the +G) 900 represents the average stock ; !his statement is incorrect !he high t'statistic for the slope coefficient =beta> suggests the value is statistically significant at the 01 level !his statement is incorrect !he <2 of 219 indicates that 21 9H of the dependent variable =return on CC1 common stock> is explained by the independent variable =return on the market> !his statement is incorrect !he slope term =beta> is statistically significant in this problem and alpha is not %eta values found by regressing stock returns against market returns tend to be more stable over time than alpha values !his statement is incorrect 2hile rerunning the regression using data over a longer period should improve the statistical reliability of the estimated coefficients$ it is not without a price !he new regression would be constructed using data that may be so old that it does not reflect the current situation

ANSWERS TO EN +O1+CHAPTER PROB.EMS 20+34 Characteristic ,ine K @ M M 20+24 A :0F1 1I9L:9;F 2I9 A 88981:1 @ A 11 F1 N A I ;9 A 11 F1'= 88981:1>=I ;9> A 9 09922; A 9 099 B 0 88FN CM, slope A =12'I>L20 A (ffiliated Omega #vy Jalue ,ine Pew Qorizons I I I I I B B B B B 2 2 2 2 2 2 A A A A A 10 11 12 1: 1; IH 2H 0H 0H 0H

=a> =b>

=1;> =1F> =20> =29> =:0>

=c> =d> 20+64 =a>

!he rank order is the same as in =b>$ which is from low risk to high risk #vy does because it has the same risk as measured by the standard deviation -rom the +M,. +tock 1 2 : ; 9 I I I I I B 9=;> A 11 B 1 :=;> A 1: B 9=;> A 10 B 1 1=;> A 12 B 1 0=;> A 12 FH 2H 0H ;H 0H


!his is the +M,$ with the formula <i A I B ;bi


-unds 1$ :$ and ; are undervalued because each has an expected return greater than its re"uired return as given by the +M, !he slope of the +M,$ or =12'I> A ; #n order to calculate the beta for each stock$ it is necessary to calculate each of the covariances with the market$ using the

=d> 20+74 =a>

correlation coefficient for the stock with the market$ the standard deviation of the stock$ and the standard deviation of the market +tock ( cov A = I>=29>=20> A ;00

beta A ;00L=20>2 A 1 0 +tock % cov A = F>=:0>=20> A :F0 9

beta A :F0L=20>2 A =b> -rom the +M,$ <i +tock ( +tock %

A I B =12'I>bi A I B =12'I>=1 0> A 12H A I B =12'I>= 9> A 11 FH

20+84 =a> =b> =c> 20+94 1< A = F>=8> B = ;>=12> A 9H +6 A ;=20> A IH

1< A =' 9>=8> B =1 9>=12> A 1; 9H 1< A =0>=8> B =1 0>=12> A 12H

+6 A 1 9=20> A :0H +6 A 20H

1=<i> A 8 0 B =1: 0'8 0>bi A 8 0 B F 0bi R)epsiCo #%M PCP% 1RGR 1(, 8 8 8 8 8 8 B B B B B B F= F= F=1 F=1 F=1 F=1 I> 9> 0> 2> 2> 9> A A A A A A 11 12 1: 1; 1; 1F IH ;H 0H 2H 2H 0H S S S T S T 12H 1:H 1;H 11H 20H 10H undervalued undervalued undervalued overvalued undervalued overvalued


#f 1xxon is in e"uilibrium$ the relationship is. 1; A <- B /1=<M ' <-0 K A F B /1=<M ' F0 1 1 !herefore$ a the slope of the +M, must be /1=<M ' F0 or approximately 8 :H in order for the relationship to hold on both sides

the expected return on the market is 8 :H B FH or 1: :H =approximately>