F520 Asset Valuation and Strategy Overview Risk and Return F520 – Portfolio Concepts 1 Overview of Market Participants and Financial Innovation • What Types of Risk does a Corporation or a Financial Intermediary Encounter? F520 – Portfolio Concepts 2 Overview (Cont.) • How can Financial Products or Intermediaries reduce these risks F520 – Portfolio Concepts 3 Risk and Return - Outline • How is the return on an asset affected by the risk of the asset? • How do we measure risk and return on an asset? – Unique Risk (diversifiable, unsystematic, residual, or specific) – Market Risk (undiversifiable, systematic, or covariance) • Constructing Portfolios -- How do we measure risk and return on a portfolio of assets? • Choosing Stocks -- Development of the Efficient Frontier and use of Indifference Curves F520 – Portfolio Concepts 4 Outline - Cont. • More on Systematic Risk Beta The Capital Asset Pricing Model (CAPM) Security Market Line (SML) • • • • Obtaining Estimates of Beta Uses of Beta Tests of the Capital Asset Pricing Line and Beta. Arbitrage Pricing Theory (APT), an alternative to CAPM F520 – Portfolio Concepts 5 Measuring Risk - Single Period r 1 1 P P P D 0 0 P1 = the market value at the end of the interval P0 = the market value at the beginning of the interval D = the cash distributions during the interval F520 – Portfolio Concepts 6 Measuring Return - Multiple Periods Arithmetic N ^ Ra R i 1 i N • Assumes no reinvestment of cash flows at the end of each period F520 – Portfolio Concepts 7 Measuring Return - Multiple Periods Geometric R [(1 R t )(1 R p 2)(1 R p 3)...(1 R pN )] 1/ N p1 1 • Also referred to as Time-Weighted Rate of Return • Assumes reinvestment of cash flows at the end of each period. F520 – Portfolio Concepts 8 Measuring Return - Multiple Periods Internal Rate of Return C3 C N VN C1 C2 V0 ... 1 2 3 N (1 RD ) (1 RD ) (1 RD ) (1 RD ) • Also referred to as Dollar-Weighted Rate of Return • Allows additions and withdrawals • When no further additions or withdrawals occur and all dividends are reinvested, the Geometric and the IRR will yield the same F520 – Portfolio Concepts 9 Comparing Return Calculations Without Dividend (Income) Cash Flows Growth of $1 investment assuming Dividend Return reinvestment 1.00 2.00 0 100% 1.00 -50% Example of Return Calculations: Period Price 10 0 20 1 10 2 Arithmetic Return: Geometric Return IRR without reinvestment IRR with reinvestment 25.00% 0.00% 0.00% 0.00% IRR Cash Flows Cash flows Shares Cash flows for IRR with owned with for IRR no reinvestment reinvestment reinvestment -10 1 -10 0 1.00 0 10 1.00 10 F520 – Portfolio Concepts 10 Comparing Return Calculations With Dividend (Income) Cash Flows Example of Return Calculations: Period Price 0 10 1 18 2 9 Growth of $1 investment assuming Dividend Return reinvestment 1.00 2 100% 2.00 -50% 1.00 Arithmetic Return: Geometric Return IRR without reinvestment IRR with reinvestment 25.00% 0.00% 5.39% 0.00% IRR Cash Flows Cash flows Shares Cash flows for IRR no owned with for IRR with reinvestment reinvestment reinvestment -10 1 -10 2 1.11 0 9 1.11 10 F520 – Portfolio Concepts 11 Measuring Total Risk Variance of actual returns 1 n Variance(r ) N 1 t 1 r r 2 t • Measures of the dispersion of returns • Standard Deviation (STD) Variance Standard deviation measures dispersion in percents F520 – Portfolio Concepts 12 Historical Returns, Standard Deviations, and Frequency Distributions: 1926-2009 F520 – Portfolio Concepts 13 Example Frequency Distribution • Frequency distribution is a histogram of yearly returns Goal: Select the lowest risk portfolio • • • • • • 0% stock, 100% bond 20% stock, 80% bond 40% stock, 60% bond 60% stock, 40% bond 80% stock, 20% bond 100% stock, 0% bond F520 – Portfolio Concepts 15 Constructing Portfolios • Investors seek to maximize the expected return from their investment given some level of risk, or • Investors seek to minimize the risk they are exposed to given some target expected return. F520 – Portfolio Concepts 16 Constructing Portfolios Portfolio Return • Expected Return of a Portfolio equals the weighted average return on the portfolio Rp = wa * Ra + wb * Rb wa wb Ra Rb = = = = weight of asset a weight of asset b Expected return of asset a Expected return of asset b • General Formula R p n wi R i 1 i – Weights must add to 1 w1 + w2 + ... + wn = 1 F520 – Portfolio Concepts 17 Constructing Portfolios Portfolio Variance • Two Asset Case Var(Rp) = Var(wa * Ra + wb * Rb ) w w 2w w 2 2 2 2 a a b b a b ab • General Case G w2g g 1 – – – – G w2g g 1 G G g g1 h1 wg wh gh 2 for h g since 12 = 21, each covariance term is included in this equation twice. i is the variance of asset i gh is the covariance between asset g and asset h G G g g1 h1 wg wh 2 p gh g h F520 – Portfolio Concepts where p gh gh g h 18 Portfolio Variance Using Correlation p gh • Correlation gh g h is the covariance standardized by the standard deviation of the two variables. – p = 1, perfect positive correlation – p = -1, perfect negative correlation – p = 0, no correlation • Two Asset Case VAR( R p) x 2 2 a a xb b 2 xa xb 2 2 p ab a b • General Case G 2 VAR( R p ) w g g 1 G G g g1 h1 wg wh 2 F520 – Portfolio Concepts p gh g h 19 Input Data A B Efficient Frontier Correlation = 1 Efficeint Frontier (Corr = 1) Return 12% 16% Std. Dev. 10% 20% Correlation 17.0% 1.00 Expected Return 16.0% 15.0% 14.0% 13.0% 12.0% 11.0% 10.0% 0.0% 3.0% 6.0% 9.0% 12.0% 15.0% 18.0% 21.0% Standard Deviation xb b 2 xa xb p xa a 2 2 p p 2 2 2 2 x a a xb b x a a x b b p ab a b Correlation = 1 2 F520 – Portfolio Concepts 20 Input Data A B Efficient Frontier Correlation = -1 Efficeint Frontier (Corr = -1) Return 12% 16% Std. Dev. 10% 20% Correlation 17.0% -1.00 16.0% Expected Return 15.0% 14.0% 13.0% 12.0% 11.0% 10.0% 0.0% 3.0% 6.0% 9.0% 12.0% 15.0% 18.0% 21.0% Standard Deviation x x 2x x p x x 2x x x a a xb b F520 – Portfolio Concepts x a a xb b 2 2 2 2 2 p a a b b 2 2 2 2 2 p a a b b 2 a a b b ab a a b Correlation = -1 b 2 p p 21 Input Data A B Efficient Frontier Correlation = 0 Efficeint Frontier (Corr = 0) Return 12% 16% Std. Dev. 10% 20% Correlation 17.0% 0.00 Expected Return 16.0% 15.0% 14.0% 13.0% 12.0% 11.0% 10.0% 0.0% 3.0% 6.0% 9.0% 12.0% 15.0% 18.0% 21.0% Standard Deviation Correlation = -1 p x a a xb b 2 x a xb 2 2 2 2 2 p x a a xb b 2 p 2 2 x 2 2 2 a a p ab a Correlation = 1 Correlation = 0 b 2 xb b 2 2 F520 – Portfolio Concepts 22 Portfolio Diversification Average annual standard deviation (%) 49.2 Diversifiable risk 23.9 19.2 Nondiversifiable risk 1 10 20 30 40 F520 – Portfolio Concepts 1000 Number of stocks in portfolio 23 Efficient Frontier Conclusions • The covariance of two assets is important in determining the variance of a portfolio • As long as assets are not perfectly correlated, combining them in a portfolio reduces risk • Systematic risk cannot be eliminated by diversification because it is the covariance risk. Also called nondiversifiable or market risk, since it is primarily from economy wide factors. • Unsystematic risk (also called diversifiable risk, unique risk, or firm specific risk) comes from circumstances unique to the firm. This is why in a well diversified portfolio, unique risk is unimportant. F520 – Portfolio Concepts 24 Covariance – the key to diversification Mathematical Example • Assume a Special Case: G wi2 i 1 2 i Cov(i,h) = 0 G G wi i 1 h 1 wh ih • As our portfolio gets large, the variances of the portfolio gets vary small if all the covariances are 0. n p xi i 2 2 2 i 1 • If all assets have weight Yn then x = 1 / n • 2 2 2 1 j j p 2 i 1 n i 1 n If the largest variance is V 2 n n p 2 i 1 • • V 2 n n nV 2 n V n As n gets large, this goes to zero. Therefore, our portfolio choices are dominated by concern over the covariance terms. In other words, well diversified investors need only price the risk F520of – Portfolio 25 associated with the covariance assets. Concepts Covariance the key to diversification - Intuitive Example # of Assets in the Portfolio # of # of Variance Covariance Terms terms 1 2 3 4 5 10 20 50 100 1 2 3 4 5 10 20 50 100 F520 – Portfolio Concepts 0 1 3 6 10 45 190 1225 4950 26 Conclusions on Covariance • Question What will the addition of this asset to my portfolio do to my level of risk? • Answer: Look at the covariance of the asset with my portfolio, rather than the variance. F520 – Portfolio Concepts 27 Choosing Stocks • Investors maximize their welfare by choosing the: – Set of securities (investments) that maximize return for a given level of risk. – Set of securities (investments) that minimize risk for a given level of return. F520 – Portfolio Concepts 28 Input Data Efficient Frontier Correlation = 0 Efficeint Frontier A B Return 6.5% 12% Std. Dev. 7.1% 16% Correlation 0.00 Expected Return 12.0% 10.0% 8.0% 6.0% 4.0% 2.0% 0.0% 0.0% 3.0% 6.0% 9.0% 12.0% 15.0% 18.0% Standard Dev iation Correlation = 0 QU: How do Investors Choose a F520 – Portfolio Concepts Portfolio on the Efficient Frontier? 29 Use Indifference Curves – measures of investor risk aversion Efficeint Frontier Expected Return 12.0% 10.0% 8.0% 6.0% 4.0% 2.0% 0.0% 0.0% 3.0% 6.0% 9.0% 12.0% 15.0% 18.0% Standard Dev iation Correlation = 0 QU: How Does this Change when a F520 – Portfolio Concepts Risk-free asset is offered? 30 Investors can move to a higher indifference curve – greater utility. Efficeint Frontier 12.0% Expected Return 10.0% 8.0% 6.0% 4.0% 2.0% 0.0% 0.0% 3.0% 6.0% 9.0% 12.0% 15.0% 18.0% Standard Deviation Correlation = 0 QU: Can you identify the important parts in the graph. F520 – Portfolio Concepts 31 Important points on the graph. AAL – Asset Allocation or CML – Capital Market Line Efficeint Frontier Borrowing 12.0% Expected Return 10.0% Risk-free rate 8.0% Lending Market Portfolio 6.0% 4.0% 2.0% 0.0% 0.0% 3.0% 6.0% 9.0% 12.0% 15.0% 18.0% Standard Deviation Correlation = 0 QU: What is meant by two-fund separation? F520 – Portfolio Concepts 32 Measuring Risk and Return for the CML • The risk free asset has no variance and its return is known with certainty (proxy – T-bill) • Portfolio Return on CML R x R x R p F RF m m • Portfolio Risk on CML 2w w p VAR( R ) w 0 w 2w w 0 0 VAR( R ) w STD ( R ) w Standard Deviation is a linear function of the STD of the market portfolio STD ( R ) w VAR( R p ) wa 2 p p 2 2 b 2 2 2 2 a a b b 2 2 b b p p wb a 2 b 2 2 b b a a b b a ab ab a b b b F520 – Portfolio Concepts 33 Conclusions from Efficient Frontier and CML • As long as there are only risky assets, it makes sense for investors to hold a portfolio on the efficient frontier. The existence of a risk-free asset changes this. The new efficient frontier (called the capital market line) will connect the risk free asset to some risky portfolio. • The market portfolio (Rm) should be chosen because any other security will lead to a lower return for a given level of risk (Tangent portfolio). • All investors will hold some combination of the risk-free asset and the market portfolio, since this will maximize their risk-return trade-off. (called two-fund separation) • The CML portfolio chosen by an investor depends upon their risk aversion F520 – Portfolio Concepts 34 • The Capital Market Line (CML) is Rp = Rf + slope (Standard Deviation) RM RF R R M p F p • The CML is a linear relationship between the efficient portfolio’s standard deviation and its expected return. QU: Can we transform the CML to another measure of risk which only accounts for systematic risk? F520 – Portfolio Concepts 35 SML, Beta, and CAPM • The CML shows that all investors must hold a combination of the risk-free asset and the market portfolio to maximize their utility. Furthermore, it shows that their is a linear relationship between risk and return. Knowing that two points make a line, let’s form the SML by plotting these points. Return Rm Rf F520 – Portfolio0.0 Concepts 1.0 Beta 36 Security Market Line Return Rm Rf 0.0 1.0 Beta • Ri = Rf + (Rm - Rf) • Where (Rm - Rf) is the slope of the line • Beta measures the risk of a stock in regards to the market portfolio (similar to the average stock). F520 – Portfolio Concepts 37 Understanding Beta and Calculating Portfolio Betas • Beta measures the relative volatility of stock i with the market portfolio. • The beta of a portfolio is the market value weighted average of the betas in the portfolio. n B x B p i 1 i i F520 – Portfolio Concepts 38 Example: Portfolio Beta Calculations Stock (1) Market Portfolio Value Weights (2) (3) Haskell Mfg. $ 6,000 Cleaver, Inc. 4,000 Rutherford Co. 2,000 50% 33% 17% Portfolio $12,000 100% F520 – Portfolio Concepts Beta (4) (3) x (4) 0.90 1.10 1.30 0.450 0.367 0.217 1.034 39 Beta, Expected Return and the Choice of Projects (Stock) • The concept that all assets must lie on the SML can also be Shown through an arbitrage argument. Consider Assets A, B, C, and D below. What will happen to the prices and expected returns of these assets in a competitive market using diversification techniques to eliminate all unsystematic risk? Return B Rm A Rf C D 0.0 1.0 Beta QU: How do I set up a trade to take advantage of this “mis-pricing”? F520 – Portfolio Concepts 40 Hedge Fund Example • How should I invest in these securities to take advantage of my expectations in returns relative to the required return. (Think about a hedge fund.) Return B CML = 5+B(6) Rm A Rf C D 0.0 1.0 Beta Beta E(Return) Req. Ret A 0.6 8.6 5+.6*6 = 8.6 B 0.8 12.0 5+.8*6 = 9.8 C 1.4 D 0.6 F520 – Portfolio Concepts 10 5+1.4*6 = 13.4 4 5+.6*6 = 8.6 41 Hedge Fund Example • Some may think of having a net investment of zero, but look at the returns with market movements. None of our securities moved closer to efficiency in the example below. They each just followed the market as their risk would suggest. A B C D Beta E(Return) 0.6 8.6 0.8 12 1.4 10 0.6 4 Req. Portfolio Market Ret Invest Beta +10% 8.6 0 6.00% 9.8 2000 0.40 8.00% 13.4 -1000 (0.35) 14.00% 8.6 -1000 (0.15) 6.00% 4000 (0.10) -1.00% Absolute Profit Market (Loss) -10% $ -6.00% $ 160 -8.00% $(140) -14.00% $ (60) -6.00% $ (40) 1.00% Profit (Loss) $ $ (160) $ 140 $ 60 $ 40 • How can we reduce our market risk while still taking a position on our expectations? F520 – Portfolio Concepts 42 Hedge Fund Example • How can we reduce our market risk while still taking a position on our expectations? A B C D Beta E(Return) 0.6 8.6% 0.8 12.0% 1.4 10.0% 0.6 4.0% Req. S=-1 Net Portfolio Market Ret Invest L=+1 Position Beta +10% 8.6 0 0.00% 9.8 2000 1 2,000 0.40 8.00% 13.4 500 -1 (500) (0.17) -14.00% 8.6 1500 -1 (1,500) (0.23) -6.00% 4000 0 0.00 0.00% Profit Market (Loss) -10% $ 0.00% $ 160 -8.00% $ (70) 14.00% $ (90) 6.00% $ 0 0.00% • Wb*Bb + Wc*Bc + Wd*Bd = 0 [no market risk] • Wb + Wc + Wd = 0 [no investment for arbitrage] • Having a portfolio beta of zero immunizes the portfolio from the market changes, and allows us to profit only from the unsystematic movements in prices, which is where one would find “mis-pricing”. • Remember this still has risk (betas could be incorrect, our estimates of over- and under-pricing could be incorrect). • Controlling for market movements, you expect prices of securities with expected returns that are higher relative to the required return to increase and lower expected returns to decrease. F520 – Portfolio Concepts 43 Hedge Fund Example • The prior example showed no profit, because we assume that the returns on the stock were exactly equal to their expected return based on the market return and their beta. What is the hedge fund correctly predicted over and undervalued stocks? Exp Ret, A B C D Beta E(Return) 0.6 8.6% 0.8 12.0% 1.4 10.0% 0.6 4.0% • • • • Req. S=-1 Net Portfolio Mkt Actual Ret Invest L=+1 Position Beta +10% Return 8.6 0 0.00% 0.00% 9.8 2000 1 2,000 0.40 8.00% 10.00% 13.4 500 -1 (500) (0.17) -14.00% -12.00% 8.6 1500 -1 (1,500) (0.23) -6.00% -4.00% 4000 0 0.00 0.00% Profit (Loss) $ $ 200 $ (60) $ (60) $ 80 Stock B is undervalued (Exp Ret > Req Ret), so we purchased a long position. Based on a market return of 10%, we expected it to increase 8% (market * beta), but our hedge fund model prediction was correct, adding 2%, so we made a net 10%. Stock C is overvalued (Exp Ret < Req Ret), so we took a short position. Based on a market return of 10%, we expected it to increase 14% (market * beta), but our hedge fund model prediction was correct, reducing it by 2% for a net increase of 12%. Since we were short, we lost 12%. Stock D is overvalued (Exp Ret < Req Ret), so we took a short position. Based on a market return of 10%, we expected it to increase 6% (market * beta), but our hedge fund model prediction was correct, reducing it by 2% for a net increase of 4%. Since we were short, we lost 4%. Our portfolio has 0 beta and made money. F520 – Portfolio Concepts 44 Hedge Fund Example • What is the market had decreased in value? A B C D Beta E(Return) 0.6 8.6% 0.8 12.0% 1.4 10.0% 0.6 4.0% • • • • • Req. S=-1 Net Portfolio Exp Ret, Actual Ret Invest L=+1 Position Beta Mkt -10% Return 8.6 0 0.00% 0.00% 9.8 2000 1 2,000 0.40 -8.00% -6.00% 13.4 500 -1 (500) (0.17) 14.00% 16.00% 8.6 1500 -1 (1,500) (0.23) 6.00% 8.00% 4000 0 0.00 0.00% Profit (Loss) $ $ (120) $ 80 $ 120 $ 80 Stock B is undervalued (Exp Ret > Req Ret), so we purchased a long position. Based on a market return of -10%, we expected it to decrease 8% (market * beta), but our hedge fund model prediction was correct, adding 2%, so we made a lost 6%. Stock C is overvalued (Exp Ret < Req Ret), so we took a short position. Based on a market return of -10%, we expected it to decrease 14% (market * beta), but our hedge fund model prediction was correct, reducing it by 2% for a net decrease of 16%. Since we were short, we made 16%. Stock D is overvalued (Exp Ret < Req Ret), so we took a short position. Based on a market return of -10%, we expected it to decrease 6% (market * beta), but our hedge fund model prediction was correct, reducing it by 2% for a net decrease of 8%. Since we were short, we made 8%. Our portfolio has 0 beta and made money. As long as our hedge fund model to predict over and under-valued stocks is correct, we make money in either an up or a down market. F520 – Portfolio Concepts 45 Uses of Beta • Discount rates in capital budgeting • Discount rates for pricing assets (stocks) • Utilities often base rates on the rate of return investors demand. • Cost of capital calculations • QU: What does the SML tell about the risk that managers should be concerned with when choosing a real asset investment (specifically a capital budgeting decision)? F520 – Portfolio Concepts 46 Estimating Beta – Characteristic Line • Ri = Rf + (Rm - Rf) • rearranging terms Ri = Rf + *Rm - *Rf Ri = (1- ) Rf + * Rm • Characteristic Line (also called market model) Ri = ά + * Rm + eit • Where im Bi 2 m = covariance (Ri, Rm) / Var (Rm) • Based on the market model, we can also break down an assets total risk into systematic and unsystematic components. Total Risk = 2i = 2i 2m + 2ei F520 – Portfolio Concepts 47 Differences in Beta Calculations • • • • Merrill Lynch – 5 years of monthly returns Value Line – 5 years of weekly returns Historic Beta – Calculated with only the raw return data Adjusted Beta – Begins with a firms historic beta and makes an adjustment for the expected future movement towards one. (Beta has been found to gradually approach 1 over time) • Fundamental Beta – Adjusts historic betas for variables such as financial leverage, sale volatility, etc. F520 – Portfolio Concepts 48 Data For Beta Calculation – Lilly Stock Calculations in yellow, WRETD = Value weighted return, DATE DIVAMT PRC CFACPR Adj Prc Adj Div Return 31-Jan-95 65.875 4 16.46875 0 28-Feb-95 0.645 67 4 16.75 0.16125 0.026869 31-Mar-95 73.125 4 18.28125 0 0.091418 28-Apr-95 74.75 4 18.6875 0 0.022222 31-May-95 0.645 74.625 4 18.65625 0.16125 0.006957 30-Jun-95 78.5 4 19.625 0 0.051926 31-Jul-95 78.25 4 19.5625 0 -0.003185 31-Aug-95 0.645 81.875 4 20.46875 0.16125 0.054569 29-Sep-95 89.875 4 22.46875 0 0.09771 31-Oct-95 96.625 4 24.15625 0 0.075104 30-Nov-95 0.685 99.5 4 24.875 0.17125 0.036843 29-Dec-95 0 56.25 2 28.125 0 0.130653 31-Jan-96 57.25 2 28.625 0 0.017778 29-Feb-96 0.3425 60.625 2 30.3125 0.17125 0.064934 29-Mar-96 65 2 32.5 0 0.072165 30-Apr-96 59.125 2 29.5625 0 -0.090385 31-May-96 0.3425 64.25 2 32.125 0.17125 0.092474 28-Jun-96 65 2 32.5 0 0.011673 31-Jul-96 56 2 28 0 -0.138462 30-Aug-96 0.3425 57.25 2 28.625 0.17125 0.028438 30-Sep-96 64.5 2 32.25 0 0.126638 31-Oct-96 70.5 2 35.25 0 0.093023 29-Nov-96 0.3425 76.5 2 38.25 0.17125 0.089965 31-Dec-96 73 2 36.5 0 -0.045752 31-Jan-97 87.125 2 43.5625 0 0.193493 28-Feb-97 0.36 87.375 2 43.6875 0.18 0.007001 31-Mar-97 82.25 2 41.125 0 -0.058655 30-Apr-97 87.875 2 43.9375 0 0.068389 30-May-97 0.36 93 2 46.5 0.18 0.062418 30-Jun-97 109.3125 2 54.65625 0 0.175403 31-Jul-97 113 2 56.5 0 0.033734 29-Aug-97 0.36 104.625 2 F520 52.3125 0.18 -0.070929 – Portfolio Concepts 30-Sep-97 121 2 60.5 0 0.156511 31-Oct-97 0 67.0625 1 67.0625 0 0.108471 VWRETD 0.0396228 0.0269823 0.0248828 0.0341465 0.0308407 0.0406674 0.0093427 0.0363905 -0.011145 0.042971 0.0153999 0.0280874 0.0160582 0.0112021 0.0251253 0.0267221 -0.007659 -0.05339 0.0322224 0.0529918 0.0139377 0.0657296 -0.011362 0.0530395 -0.000889 -0.044396 0.0424802 0.071263 0.0441994 0.0763158 -0.036456 0.0580094 -0.034116 49 Data For Beta Calculation – Lilly Stock SUMMARY OUTPUT Regression Statistics Multiple R 0.089765 R Square 0.008058 Adjusted R Square -0.006318 Standard Error 0.09864 Observations 71 0.008057732 Percent of total variation explained = MS(reg)/SS(total) Total Percent Systematic Risk 7.6811E-05 0.00806 R-squared Unsystematic Risk 0.00945575 0.99194 1- R-squared Total Risk 0.00953256 0.00953256 1 Should add to 1 due to the degrees of freedom in regression these are not exactly equal to each other ANOVA df Regression Residual Total Intercept X Variable 1 1 69 70 SS 0.00545357 0.671358409 0.676811978 Coefficients Standard Error 0.027862 0.01243276 0.196282 0.262175336 MS F Significance F 0.00545357 0.56049986 0.456603 0.009729832 t Stat P-value Lower 95% Upper 95% Lower 95.0% 2.241034391 0.02824458 0.00306 0.05266491 0.00305957 0.748665388 0.45660343 -0.326744 0.71930692 -0.3267437 Upper 95.0% 0.052664912 0.719306923 Alpha Beta estimate RESIDUAL OUTPUT Observation Predicted Y 1 0.035639 Residuals SS(residuals) -0.008770407 7.692E-05 F520 – Portfolio Concepts 50 Assumptions of the CAPM (SML) • Assumptions about investor behavior – Investors use only two measures to determine their strategy, expected return and risk, – Investors will choose portfolios as a risk reduction technique, – Investors make investment decisions over some single-period investment horizon, – Homogenous expectations with respect to asset returns, variances, and correlations • Assumptions about capital markets – Perfect competition, – No transaction costs • -No bid-ask spreads, -No commissions, -No information costs, -No taxes, -No regulation, and -all assets are marketable – Investors can borrow and lend at the risk-free rate. F520 – Portfolio Concepts 51 Test of the CAPM (SML) • Clearly the assumptions are unrealistic, but the true test of a model comes from answering two questions – Does the model change when the assumptions are changed? – How well does the model predict? • Empirical Findings – There is a significant positive relationship between realized returns and systematic risk. However, the slope is usually less than predicted by the CAPM. – The relationship between risk and return appears to be linear. No evidence of curvature has been found. – Tests assessing the importance of company specific risk after controlling for market risk are inconclusive. Econometrically controlling for market risk given its high correlation with total risk is difficult. – The CAPM should be valid for all assets; however, bonds do not track along the SML. – Betas of individual stocks are not stable over time; however, betas for portfolios are stable over time. F520 – Portfolio Concepts 52 Anomalies with using the CAPM • Small firm effect • Price-to-Book Ratios (Growth versus value stocks) • January effect F520 – Portfolio Concepts 53 Common Question: When using CAPM [Ri=Rf+i(Rm – Rf)], what is the Risk Premium (Rm – Rf) What is the Rf you are using? Should you use Large or Small Stocks? Should you use arithmetic or geometric returns? F301_CH12-54 Can CAPM be used for bond? • (August 9, 2013 data) Lehman Index (ticker = AGG) http://www.ishares.com/product_info/fund/overview/AGG.htm Effective Duration Average Yield to Maturity 5.05 years 2.16% http://finance.yahoo.com/q/rk?s=AGG+Risk Beta (against Standard Index) R-squared (against Standard Index) • 1.01 Yahoo 98.88 Yahoo betas are 5-years Lehman 1-3 year Treasury Bond Fund (ticker = SHY) http://www.ishares.com/product_info/fund/overview/SHY.htm Effective Duration 1.86 Average Yield to Maturity 0.32% http://finance.yahoo.com/q/rk?s=SHY+Risk Beta (against Standard Index) R-squared (against Standard Index) 0.14 Yahoo 24.31 What is the standard index in this case? – So what is beta in this case? 1.86 / 5.05 = 0.36, compare to Beta? F520 – Portfolio Concepts 55 Can CAPM be used for bond? • Lehman 7-10 Year Treasury Bond Fund (ticker = IEF) http://www.ishares.com/product_info/fund/overview/IEF.htm Effective Duration 7.48 Average Yield to Maturity 2.29% http://finance.yahoo.com/q/rk?s=IEF+Risk Beta (against Standard Index) R-squared (against Standard Index) 1.70 Yahoo 69.52 – So what is beta in this case? 7.48 / 5.05 = 1.48, compare to Beta? • Lehman 20+ Year Treasury Bond Fund (ticker = TLT) http://www.ishares.com/product_info/fund/overview/TLT.htm Effective Duration 16.43 Average Yield to Maturity 3.61% http://finance.yahoo.com/q/rk?s=TLT Beta (against Standard Index) R-squared (against Standard Index) 3.37 Yahoo 53.81 – So what is beta in this case? 16.43 / 5.05 = 3.25, compare to Beta? F520 – Portfolio Concepts 56 Cont. • • The concept of Beta, used by Yahoo Finance and MSN Money for bonds is not the same concept of beta referred to in stocks. When a bond index is used as the standard index, we obtain a relative measure of duration. When a stock index is used, we obtain the traditional measure of systematic risk. When using public betas, identify the index used to interpret the concept of beta reported. For many companies/funds, they state a “Standard Index”, to properly interpret the measures, you must clearly identify the index. (MSN Money provides identification, Yahoo does not.) – – For Ishare Austria Fund: http://investing.money.msn.com/investments/etf-management?symbol=ewo For Ishare Japan Fund: http://investing.money.msn.com/investments/etf-management?symbol=ewj Standard Index is MSCI EAFE NDTR_D EAFE stands for Europe, Australasia, and Far East. The index has stocks from 21 developed markets, excluding the U.S. and Canada. For Ishare S&P Small Cap 600 Index, (uses S&P500) http://investing.money.msn.com/investments/etf-management?symbol=ijr For Ishare NAREIT Industrial/Office Index Fund (uses MSCI World) http://investing.money.msn.com/investments/etf-management?symbol=fnio F520 – Portfolio Concepts 57 Multifactor CAPM • Multi-Factor CAPM E(Ri) = Rf + i,M[E(RM) - Rf] + i,f1[E(Rf1) - Rf]+ i,f2[E(Rf2) - Rf] +…+ i,fn[E(Rfn) - Rf] • By rearranging terms we get the multiple regression typically used. E(Ri) = Rf + i,M*E(RM) - i,M*Rf + i,f1*E(Rf1) - i,f1*Rf + i,f2*E(Rf2) - i,f2*Rf +…+ i,fn*E(Rfn) - i,fn*Rf E(Rit) = + i,Mt*E(RMt) + i,f1*E(Rf1t) + i,f2*E(Rf2t) +…+ i,fn*E(Rfnt) + eit where = Rf - i,M*Rf - i,f1*Rf - i,f2*Rf -…- i,fn*Rf Rf = Riskfree Rate Rf1 = Expected Return on factor 1 F520 – Portfolio Concepts 58 Arbitrage Pricing Theory (APT), an alternative to the CAPM • E(Ri) = Rf + i,f1[E(Rf1) - Rf] + i,f2[E(Rf2) - Rf] +…+ i,fn[E(Rfn) - Rf] • By rearranging terms we get the multiple regression typically used. E(Rit) = + i,f1*E(Rf1t) + i,f2*E(Rf2t) +…+ i,fn*E(Rfnt) + eit where = Rf - i,f1*Rf - i,f2*Rf -…- i,fn*Rf Rf = Risk-free Rate Rf1 = Expected Return on factor 1 F520 – Portfolio Concepts 59 Assumptions of APT • APT assumes returns are a function of several factors, not just one as in the CAPM • Suggested factors (Roll & Ross 1983) – – – – Index of Industrial Production, Changes in the default risk premium on bonds, Changes in the yield curve, Unanticipated inflation • Other factors frequently considered – Factors for size – Factors for book-to-market value F520 – Portfolio Concepts 60 Principles to Take Away from the APT and CAPM • Investing has two dimensions, risk and return. • It is inappropriate to look at the risk of an individual asset when deciding whether it should be included in a portfolio. What is important is how the inclusion of an asset into a portfolio will affect risk of the portfolio (covariance and/or beta must be considered). • Risk can be divided into two categories, systematic and unsystematic • Investors should only be concerned about systematic risks. F520 – Portfolio Concepts 61 Commonly used Portfolio Performance Criteria are based on the Efficient Frontier or CAPM concepts Global Tech Fund: Return: +37.2%, Beta: 1.29, Std. Dev 25%, Riskfree = 5.0%, RiskPremium = 6.0% • Sharpe Ratio = (Rp – Rf)/σp Efficeint Frontier 12.0% Expected Return 10.0% = (37.2 – 5)/25 = 1.29 8.0% 6.0% 4.0% • Treynor Ratio = (Rp – Rf)/Bp 2.0% 0.0% 0.0% 3.0% 6.0% 9.0% 12.0% 15.0% 18.0% Standard Deviation Correlation = 0 = (37.2 – 5)/1.29 = 25.0% • Jensen’s alpha (αp) Return R = Rp – CAPM = Rp – [Rf+Bp(RM – Rf)] m = 37.2 – [5+1.29(6)] = 24.5% Rf 0.0 F520 – Portfolio Concepts 1.0 Beta 62 M-Squared Measure (Modigliani and Modigliani) Global Tech Fund: Return: +37.2%, Beta: 1.29, Std. Dev 25%, Riskfree = 5.0%, RiskPremium = 6.0% This fund has a beta of 1.29, substantially greater than the market beta of 1.0. To compare it to the market, we must determine what portion can be invested in the risk-free rate and what portion invested in the Global Tech Fund to have the same risk as the market. Let x = the percent invested in the Global Tech Fund, subsequently (1-x) is the percent in the risk-free asset. Note that the beta of the risk-free asset is equal to zero. (1-x)(0) + x(1.29) = 1.0 Solve for x. x = 1/1.29 = .78 portion of the portfolio in the Global Tech Fund 1-x = .22 portion invested in the risk-free asset Now calculate your risk-adjusted return: The Global Tech Fund earned 37.2% and the risk-free asset over this 3-year period earned 5%. The proportions in each asset are calculated above. .22(5%) + .78(37.2%) = 30.1% This value can be compared to what the market earned during this period, since it has a beta of 1. F520 – Portfolio Concepts 63