CAPM Lecture Notes
CAPM Lecture Notes
Lecture Notes 9
VII. The CML and SML VIII. Overpricing/Underpricing and the SML IX. X. Uses of CAPM in Corporate Finance Additional Readings
Equilibrium Process, Supply Equals Demand, Market Price of Risk, Cross-Section of Expected Returns, Risk Adjusted Expected Returns, Net Present Value and Cost of Equity Capital.
Buzz Words:
I.
2.
There are many dubious assumptions. The main implication of the CAPM concerns expected returns, which cant be observed directly.
B. Implications of the CAPM: A Preview If everyone believes this theory then (as we will see next): 1. There is a central role for the market portfolio: a. This simplifies portfolio selection. b. Provides a rationale for a market-indexing investment strategy.
2. There is explicit risk-return trade-off for individual stocks: The model specifies expected returns for use in capital budgeting, valuation, and regulation. Risk premium on an individual security is a function of its systematic risk, measured by the covariance with the market. a. can use the model to evaluate given estimates of expected returns relative to risk b. Can obtain estimates of expected returns through estimates of risk. (This is more precise statistically than obtaining direct estimates of expected returns based on averages of past returns)
price per share of company is stock, number of shares outstanding, the market value of is equity.
The total market value of all stocks: V = v1+v2+...+vn The weight of stock i in the market portfolio wi,M = vi / V. Example
Suppose the weight of IBM is: wIBM = 1.5% If we put $100,000 in the market portfolio, $1,500 should be invested in IBM.
V.
Er
CAL
P=T
rf
(Risky assets)
B. If T is the same for everybody (all investors agree on what are the tangent weights), then T is the Market portfolio (M). That is, each assets weight in the tangent portfolio, wi,T, is simply its weight in the market portfolio: wi,T = wi,M V = total market value of all stocks = total funds invested in the tangent portfolio = market value of i's equity = total funds invested in firm i = wi,T (total funds invested in the tangent portfolio) = wi,T V
6
vi
Therefore, wi,T =
wi ,T V V
vi V
= wi,M
Example Suppose based on the Mean-Variance analysis, IBMs weight in the tangent portfolio is wIBM,T = 1% (all investors agree on that), and all investors combined have $12 trillion to invest (so V= $12 trillion). Then, IBMs market value is vIBM = wIBM,T V = 1% 12 trillion = $120 billion and IBMs market weight is wIBM,M = vIBM/V = $120 billion/12 trillion = 1% = wIBM,T
So everyone holds some combination of the value weighted market portfolio M and the riskless asset.
C.
Capital Market Line (CML) The CAL, which is obtained by combining the market portfolio and the riskless asset is known as the Capital Market Line (CML): E rM r f C E rC = r f + M
7
Er M
rf
D. Indexing The portfolio strategy of matching your portfolio (of risky assets) to a popular index. 1. Indexing is a passive strategy. (No security analysis; no market timing.) 2. Some stock indices (e.g., the S&P 500 index) use market value weights. 3. If the total value of the stocks in the index is close to the value of all stocks, then it may approximate the market portfolio. 4. Conclusion: the investor can approximate the market portfolio by matching a market index.
8
Er
CML
M
ErIBM IBM
rf
IBM
9
B. How should we measure the risk of an asset (IBM)? As an investor, I care only about C (the risk in my combined portfolio). The risk in C derives entirely from the risk in the market portfolio M: C = y M What does IBM contribute to M?
We can answer that using a thought experiment, where we consider the covariance of the asset with the market, iM, or equivalently its beta, because iM = iM/2M. [iM is often written simply as i (note: M = 1), and it measures how much an assets return is driven by the market return.]
So now consider the following marginal portfolio formation scenario: An investor holds the market portfolio. A new stock is issued. The investor is considering adding it to his portfolio.
10
If the stock has a high positive : - It will have large price swings driven by the market - It will increase the risk of the investors portfolio (in fact, will make the entire market more risky ) - The investor will demand a high Er in compensation. If the stock has a negative : - It moves against the market. - It will decrease the risk of the market portfolio - The investor will accept a lower Er (in exchange for the risk reduction, and Er can be negative).
D. The Pricing Implication of the CAPM for Any Asset 1. Security Market Line (SML) In a CAPM world, SML describes the relationship between and Er (holds for individual assets and portfolios)
Er SML
0
11
In order to fix the position of the SML, we need to know two points: For the risk-free security, = 0. (The risk-free return is constant; it isnt driven by the market.) For the market itself, =1. (If we run a regression of rM vs rM, the slope is 1.)
Er ErM rf M
SML
Eri = rf + i ( ErM rf ), where Eri is the expected return on an (arbitrary) stock or security i. i is security is beta.
12
The CAPM therefore states that in equilibrium, only the systematic (market) risk is priced, and not the total risk; investors do not require to be compensated for unique risk.
(Although it is somewhat similar to what we saw in the market model, recall that in the market model the market beta determines the expected return of a security simply by construction, whereas here it is due to an economic argument that all investors hold the same tangent portfolio. Moreover in the market model the intercept varies across assets, and it is not clear what are all the determinants of expected returns.)
Like the CML, the SML is a statement about expected returns. In any given year, a low- stock could have a high return and a high- stock could have a low return. In principle, a stock with <0 will have Er < rf. (If <<0, could have Er < 0.) Most stocks have s between 0 and 3.
3. Estimation Example 1:
13
Note: this estimation uses weekly return data to get beta from the slope of the regression. From a data source called SBBI (see BKM), ErM-rf 8.5%; also lets use rf = 5%; Then the SML implies: ErMRK = 5% + 1.19(8.5%) = 15.1%
Example 2
We saw before an example where we estimated GE = 1.44, MSFT = 0.88. Empirical evidence suggests that over time the betas of stocks move toward the average beta of 1. For this reason, a raw estimate of beta is often adjusted using the following formula: adj Z raw + (1-w) 1, where typically w=0.67.
Example 3
14
Both specify a relation between risk and Er: Expected Return = Time Premium + Risk Premium where Risk Premium = quantity of risk price of risk
Measure of risk.
15
B. All assets (and portfolios) lie on the SML yet only efficient portfolios which are combinations of the market portfolio and the riskless asset lie on the CML How can this be?
1. First note that since by definition, for any portfolio p [rp,rM] = [rp,rM] [rp] [rM] it follows that
p,M = [ r p , rM ] [ r p , rM ] [ r p ] [ rM ] [ r p , rM ] [ r p ] = = . [ rM ] 2 [ rM ] 2 [ rM ]
2.
[ rM ]
{ [ r p , r M ] [ r p ]} .
3.
Comparing this equation to the CML, which hold for efficient portfolios or assets (subscript ef below)
CML : E[ r ef ] = r f + E[ r M ] - r f
[ rM ]
[ r ef ]
we see that: an asset p lies on the SML and the CML if [rp,rM]=1. an asset p only lies on the SML and is not a combination of the riskless asset and the market portfolio if [rp,rM]<1
16
2.
Er
17
Regulated Project Specification: Investment= $200 million Required annual profit: 200 0.106 = $21.2 million Note: 10.6% is called, in this context, the cost of equity capital and it affects the prices the monopoly (e.g. power utility) will charge.
18
X.
Additional Readings
The article is by William Sharpe (a Nobel Laureate), one of the creators of what we refer to as the CAPM. He describes in the article the multi-index models (which he refers to as factor models), the CAPM, and the APT, and the relation between all these. The APT the Arbitrage Pricing Theory derives a general relationship between expected return and risk using only no arbitrage arguments (i.e., relying only on the investors preferring more to less). Although we do not cover the APT in the course (mainly because it applies to well diversified portfolios, but not necessarily to individual stocks), it is worthwhile to be familiar with this alternative approach to asset pricing (see also BKM, chapter 11). The APT, the CAPM, and the CAPMs extensions are all called Asset Pricing models because they tell us how assets are priced. We focused the discussion on the expected return, E[r], because it is exactly this quantity which we need to be able to price assets. Why? As we discussed in the context of over/under pricing, we know that an asset that pays a random cash flow CF next period will have a return: r = CF/P 1, where P is the current price. Taking expectations, we get E[r] = E[CF]/P 1. Rearranging, we find the price: P = E[CF]/(1+E[r]). That is, we discount the expected cash flow using an appropriate discount rate, which is provided to us by an asset pricing model in terms of the associated risks (and various models differ in their normative guidance of how exactly to calculate E[r]).
19