FINS5513 Lecture 3A

Capital Allocation and

Optimal Risky Portfolios
Lecture Outline
❑ 3.1 Introducing the Risk-Free Asset
➢ The Risk-Free Asset

➢ Combining Risky and Risk-Free Assets

• Complete portfolios: return and risk
➢ Capital Allocation Line

➢ Deriving Optimal Portfolio Weights

❑ 3.2 Deriving Optimal Portfolio Weights

➢ Optimal Risky Portfolio 𝑃 ∗

➢ Optimal Complete Portfolio 𝐶 ∗

• Optimal allocation to risky assets 𝑦 ∗
➢ Separation Theorem
• Summary of the Markowitz Optimisation Model
➢ Leveraged Portfolios
• Kinked CAL
2
3.1 Introducing the
Risk-Free Asset

FINS5513
The Risk-Free Asset

FINS5513 4
Brief Review
❑ Last week we introduced Modern Portfolio Theory (MPT)
❑ Portfolio optimisation under MPT is often called the Markowitz optimisation model
❑ We can think of the Markowitz model as having 3 steps:
➢ Last week we covered Step #1 of the Markowitz model – derivation of the efficient frontier

➢ But where along the efficient frontier should we invest?

• When we add a risk-free asset, we can optimise our answer further
❑ The next step in the Markowitz optimisation model tells us how to find the optimal point on the
efficient frontier
➢ We will add a risk-free asset

➢ We will find the point on the efficient frontier which provides the highest Sharpe ratio

5
The Risk-Free Asset
❑ Previously we only considered risky assets
❑ Now let’s add a risk-free asset:
➢ Short-term Government bills (T-bills) are often considered as the risk-free asset, as they
have almost no default risk and limited interest rate risk
❑ The return on the risk-free asset (the “risk-free rate”) is denoted rf
❑ Even if we don’t take risk, we still want a positive return
➢ Hence 𝑟𝑓 is generally positive as shown

❑ The risk premium 𝑅 on a risky asset is: 𝐸(𝑟) – 𝑟𝑓

➢ Note that excess return/risk premium is denoted 𝑹
while total return is denoted 𝒓

Risk-free
rf
rate

6
Combining Risky and
Risk-Free Assets

FINS5513 7
Video 3AV1: “What is a Cash Investment” Video 3AV2: “Vanguard: Moving to Cash”

Combining Risky and Risk-Free Assets

❑ One of the most important capital allocation decisions is determining how much of our
portfolio to allocate to risk-free assets vs risky assets Video 3AV3: “Druckenmiller on Treasuries”
➢ Let’s now combine a risky asset portfolio with a risk-free asset

❑ First, let’s denote all asset class returns and risks correctly:
➢ We have identified an efficient risky portfolio on the efficient frontier, and denote it 𝑃
Risk-Free Assets 𝑭 Risky Assets 𝑷
▪ T-Bills/Govt Bonds ▪ Equities: wE1,wE2,wE3… wEn
▪ Money Market Funds ▪ Risky bonds: wB1,wB2,wB3… wBn
▪ Bank Deposits ▪ Alternatives: wA1,wA2,wA3… wAn
Expected Return 𝑟𝑓 𝐸(𝑟𝑝 )
Risk 𝜎𝑟𝑓 = 0 𝜎𝑝
Weighting (1 − 𝑦) 𝑦
Combination Complete Portfolio C
Comb. Expected Return 𝐸(𝑟𝐶 ) = 1 − 𝑦 𝑟𝑓 + 𝑦𝐸(𝑟𝑝 ) = 𝑟𝑓 + 𝑦[𝐸(𝑟𝑝 ) – 𝑟𝑓 ]
Comb. Risk 𝜎𝐶 = 𝑦𝜎𝑝 8
Complete Portfolio Expected Return
❑ What we are trying to derive is the appropriate weighting between the risky assets portfolio 𝑃
and the risk-free asset(s) 𝐹 in our Complete Portfolio 𝐶
❑ As usual, to determine weightings, we must first derive the expected return and risk of 𝐶
❑ The return on our complete portfolio 𝐶 is:
𝑟𝐶 = 1 − 𝑦 𝑟𝑓 + 𝑦𝑟𝑝
❑ So, the expected return on the Complete Portfolio 𝐶 is given by:
𝐸(𝑟𝐶 ) = 1 − 𝑦 𝑟𝑓 + 𝑦𝐸(𝑟𝑝 )
❑ As we saw previously, portfolio expected return is simply a weighted average of the
component asset expected returns
➢ It is useful to express this equation as a risk premium. Rearranging we have:
𝐸(𝑟𝐶 ) = 𝑟𝑓 + 𝑦[𝐸(𝑟𝑝 ) – 𝑟𝑓 ]
[𝐸(𝑟𝑝 ) – 𝑟𝑓 ] is often referred to as the risk premium
9
Complete Portfolio Risk
❑ What about risk? We know that portfolio risk for a 2-asset portfolio (𝑃 and 𝐹) is given by:
𝜎𝐶2 = 𝑦 2 𝜎𝑃2 + (1 − 𝑦)2 𝜎𝑟2𝑓 + 2𝑦 1 − 𝑦 𝐶𝑜𝑣 𝑟𝑝 , 𝑟𝑓

❑ However, we know 𝑟𝑓 is a constant and has no variance (i.e. it is certain)

Therefore: 𝜎𝑟2𝑓 = 0 and 𝐶𝑜𝑣 𝑟𝑝 , 𝑟𝑓 = 0
➢ So the portfolio risk equation above simplifies to:
𝜎2𝐶 = 𝑦2𝜎2𝑃
𝜎𝐶 = 𝑦2𝜎2𝑃
𝜎𝐶 = 𝑦𝜎𝑃
𝜎𝐶
And rearranging: 𝑦=
𝜎𝑃
Now replace y into our expected return derived on the last slide:
𝜎𝐶
𝐸(𝑟𝐶 ) = 𝑟𝑓 + [𝐸(𝑟𝑝 ) – 𝑟𝑓 ]
𝜎𝑃
10
Capital Allocation Line

FINS5513 11
Graphical Representation
❑ Lets look at this equation carefully:
𝜎𝐶
𝐸(𝑟𝐶 ) = 𝑟𝑓 + [𝐸(𝑟𝑝 ) – 𝑟𝑓 ]
𝜎𝑃

❑ What do you notice about this equation?

➢ It is a linear line in the 𝐸 𝑟 − 𝜎 space with intercept 𝑟𝑓 and slope equal to the Sharpe ratio:

𝐸 𝑟𝑝 − 𝑟𝑓
Slope =
𝜎𝑝

❑ So graphically, if we drew a straight line connecting 𝑟𝑓 (𝑦-intercept 𝜎 = 0) to the risky asset 𝑃:

➢ The slope of the line will equal 𝑃’s Sharpe ratio
➢ 𝑦 (the weight in risky assets) will tell us where along the line we sit
➢ When this line connects with the optimal portfolio on the efficient frontier (which we will
define in 3.2), it is known as the Capital Allocation Line (CAL)

12
Capital Allocation Line
❑ The higher is 𝑦, the more we allocate to risky assets, and therefore the higher the return and
the risk of the combined portfolio
❑ If 𝑦 > 1, it means borrowing at 𝑟𝑓 (instead of investing in the risk-free asset at 𝑟𝑓 ) and investing
the proceeds into risky assets i.e. taking a levered position in risky assets
❑ We can label the line that links with risky portfolio 𝑃 𝐶𝐴𝐿𝑃 – the Capital Allocation Line for 𝑃

E(r)
𝑟𝑓 is the
𝑦-intercept as C(y=0.75) 𝑦=1
CALP
C(y=0.25)

𝜎=0
P
rf C(y=1.25)
C(y=0.5)

σ
13
Example: Risky and Risk-Free Assets
❑ Example 3A1: Low Risk Fund has identified an efficient portfolio 𝑃 of risky assets on the
efficient frontier with an expected return 𝐸(𝑟𝑃 ) = 9.21% and risk of 𝜎𝑃 = 16.92%
a) What is the Sharpe ratio of 𝑃?
b) If the fund prefers a lower risk level, how would it efficiently combine risky portfolio 𝑃 and
the risk-free asset (𝑟𝑓 = 2.0%) to create a complete portfolio 𝐶 with risk level of 𝜎𝐶 = 12%?
c) What is the expected return 𝐸(𝑟𝐶 ) and Sharpe ratio of this complete portfolio 𝐶?
➢ Sharpe ratio of 𝑃 = (.0921-.02)/.1692 = .426
𝜎𝐶
➢ Share in P at preferred risk level of 12%: 𝑦 = = .12 / .1692 = 70.9%
𝜎𝑃
𝜎𝐶
➢ Expected return of complete portfolio 𝐶: 𝐸(𝑟𝐶 ) = 𝑟𝑓 + [𝐸(𝑟𝑝 ) – 𝑟𝑓 ]
𝜎𝑃
= .02 + .709 (.0921 -.02) = 7.11%
➢ Sharpe ratio of 𝐶 = (.0711 - .02) / .12 = .426
➢ By putting 70.9% of its capital in the risky asset portfolio 𝑃 and 29.1% in the risk-free asset,
the fund can create an efficient portfolio (same Sharpe Ratio as 𝑃) with 𝜎𝐶 = 12%
14
Example: Risky and Risk-Free Assets
The Capital Allocation Line
14.0%
At 𝑪: 𝐸(𝑟𝐶 ) = 7.11% When linked to an
and 𝜎𝐶 = 12.0% 12.0%
optimal risky portfolio, the

Portfolio Expected Return E(r)

10.0% P CAL depicts optimal risk-
Slope of CAL is 𝑃’s 8.0% return combinations
Sharpe ratio available to investors
6.0% C
4.0%

2.0%
𝑟𝑓 = 2.0% At 𝑷: 𝐸(𝑟𝑃) = 9.21%
0.0%
and 𝜎𝑝 = 16.92%
0 0.05 0.1 0.15 0.2 0.25 0.3

Portfolio Standard Deviation σ

σ𝐶
𝐸(𝑟𝐶 ) = 𝑟𝑓 + (𝐸(𝑟𝑝) − 𝑟𝑓 ) = .02 + .426 𝜎𝐶
σ𝑃

➢ This equation describes the Capital Allocation line. The 𝑦-intercept is 𝑟𝑓 = 2% and the
slope is the Sharpe ratio

15
3.2 Deriving Optimal
Portfolio Weights

FINS5513
 ∗
Optimal Risky Portfolio 𝑷

FINS5513 17
Why Add the Risk-free Asset?

𝐶1 offers the same

return as 𝑃1 but 𝐶3 dominates 𝑃3
lower risk: 𝐶1
dominates 𝑃1

❑ Efficient risky portfolios, 𝑃1 and 𝑃3 , are dominated by complete portfolios of the risk-free asset
and the (inefficient) risky portfolio, 𝑃𝐼 (portfolios 𝐶1 and 𝐶3 )
❑ Hence, combinations of risky portfolios with risk-free assets can dominate the efficient frontier
➢ The risk-free asset has significantly increased our investment opportunities

18
Which Risky Portfolio is Optimal?
❑ So, which efficient portfolio is optimal?
➢ Which portfolio combination should we choose?
CALPE
PE
E(r)
CPE CALPI
PI
CPI
rf

σ
❑ Clearly it is not optimal to combine the risk-free asset with an interior (inefficient) risky
portfolio, 𝑃𝐼
❑ But 𝐶𝐴𝐿𝑃𝐸 and portfolios 𝐶𝑃𝐸 and 𝑃𝐸 are also not optimal. Why?
❑ What CAL would be optimal?
19
Which Risky Portfolio is Optimal?
Note these principles:
❑ Remember, the slope of the CAL is the Sharpe ratio of the efficient risky portfolio it links to
➢ As we want to maximise the Sharpe ratio, we want to maximize the slope of the CAL for any
possible efficient portfolio, 𝑃
➢ The steeper the slope of the CAL to an efficient risky portfolio 𝑃, the better the risky portfolio
❑ The steepest sloping CAL maximises return at each level of risk (or minimises risk at each
level of return). In other words it maximises the Sharpe ratio
❑ Therefore we want to invest along the CAL where:
1. The complete portfolios along the CAL dominate all other portfolios - the steepest possible
CAL
2. The investment opportunity is attainable

20
The Optimal Risky Portfolio 𝑷∗
❑ The optimal CAL is the one which is tangent to the efficient frontier 𝑪𝑨𝑳𝑷∗
❑ The point of tangency 𝑷∗ is called the Optimal Risky Portfolio
❑ The line 𝐶𝐴𝐿𝑃∗ is often simply denoted CAL - the Capital Allocation Line

𝐶𝐴𝐿𝑃∗ = CAL has CALP* = CAL

the highest possible
Sharpe Ratio
E(r)
P*

𝑃∗ is the best possible

portfolio weighting
rf combination

σ
21
Summary Principles to Identify 𝑷∗
❑ Only risky portfolios on the efficient frontier should be considered
➢ The CAL links the risk-free asset to a portfolio on the efficient frontier

❑ The steeper the CAL the better eg 𝐶𝐴𝐿2 clearly dominates 𝐶𝐴𝐿1
❑ The point of tangency with the efficient frontier from 𝑟𝑓 provides the steepest possible CAL
and identifies the Optimal Risky Portfolio 𝑃∗
❑ Any portfolio beyond the efficient frontier is not attainable
CALP* = CAL

CAL2
E(r)
P*
Any CAL above
𝐶𝐴𝐿𝑃∗ is not attainable CAL1

rf

σ 22
Two Asset Optimal Risky Portfolio (𝑷∗ )
❑ Note the optimal risky portfolio 𝑃∗ is different from the GMVP
➢ The GMVP is the portfolio combination which minimises risk (regardless of return)

➢ The optimal risky portfolio 𝑃∗ is the portfolio combination which maximises the Sharpe ratio
❑ For a portfolio with just 2 risky assets (or asset classes), the optimal risky portfolio weights
are given by:
𝐸(𝑟1)𝜎22 − 𝐸(𝑟2 )𝐶𝑜𝑣(𝑟1 ,𝑟2 )
𝑤1 = 𝐸(𝑟1 )𝜎22 + 𝐸(𝑟2 )𝜎12 − [𝐸 𝑟1 +𝐸 𝑟2 ]𝐶𝑜𝑣(𝑟1,𝑟2 )
𝑤2 = 1 − 𝑤1

❑ For more than 2 assets this optimisation gets tricky – we often use Excel Solver

23
 ∗
Optimal Complete Portfolio 𝑪

FINS5513 24
The Optimal Allocation Along the CAL
❑ The remaining question is, where along the Capital Allocation Line should we invest?
➢ More risk averse investors would invest less in 𝑃 ∗
• That is, invest more in risk-free assets and invest less (𝑦 ∗ <100%) in risky assets 𝑃 ∗

➢ Less risk averse investors would invest more in 𝑃∗

• That is, invest less in risk-free assets or potentially even borrow at the risk-free rate 𝑟𝑓 (in
which case 𝑦 ∗ >100%) and invest more in risky assets 𝑃∗
❑ Therefore, where an investor invests along the CAL is a personal choice based on each
individual investor’s level of risk aversion i.e. their risk aversion coefficient 𝐴
❑ The exact location will be where the highest possible indifference curve is tangent to the CAL
(at allocation point 𝑦 ∗ )
❑ So, how do we determine the optimal allocation to risky assets 𝑦 ∗ ?

25
Optimal Allocation to Risky Assets 𝒚∗
❑ An investor chooses 𝑦 ∗ based on their individual risk preferences:
➢ The risk aversion coefficient 𝐴 is different for different investors

➢ Therefore, different investors have different utility functions and different indifference curves:
Unattainable
𝑪∗ is the Optimal Complete indifference curve
Portfolio – tangent between the ❑ These are different
CAL and the highest attainable indifference curves for
indifference curve
one individual. 𝑪∗ is
E(r) specific to this one
P* investor based on their
indifference curves
C* (and therefore their
risk aversion)
rf
𝑪∗ is determined at the
optimal risky share 𝑦 ∗

σ 26
Deriving the Optimal Risky Allocation 𝒚∗
❑ We can derive the optimal allocation to risky assets 𝑦 ∗ mathematically:
❑ Recall we derived the return and risk on a Complete Portfolio 𝐶
Expected Return: 𝐸(𝑟𝐶 ) = 𝑟𝑓 + 𝑦[𝐸(𝑟𝑝∗ ) – 𝑟𝑓 ]
Risk: 𝜎𝐶 = 𝑦𝜎𝑃∗ or re-stated as variance: 𝜎𝐶2 = 𝑦2𝜎𝑝∗
2

Replace these into the utility function and optimise:

1
Max 𝑈 = 𝐸(𝑟𝐶 ) − 𝐴 𝜎𝐶2
2
1 2
= 𝑟𝑓 + 𝑦[𝐸(𝑟𝑝∗ ) – 𝑟𝑓 ] − 𝐴 𝑦2𝜎𝑝∗
2

Take the 1st derivative to maximise U and solve for 𝑦 ∗ :

𝐸(𝑟𝑝∗ ) – 𝑟𝑓 − 𝐴 𝑦 ∗ 𝜎𝑝∗
2 =0

𝐸(𝑟𝑝∗ ) – 𝑟𝑓 𝑦 ∗ is positively correlated

𝑦∗ = 2
to the risk premium
𝐴 𝜎𝑝∗
𝑦 ∗ is negatively correlated to
risk aversion and portfolio risk 27
Optimal Complete Portfolio 𝑪∗
❑ 𝑦 ∗ identifies the optimal proportion an individual investor should allocate to risky assets to
achieve their Optimal Complete Portfolio 𝑪∗
➢ While 𝐶𝐴𝐿𝑃∗ is universal for all investors, each individual’s indifference curve determines
where they sit along this CAL
❑ Example 3A2: Low Risk Fund has identified 𝑃∗ as having an expected return 𝐸(𝑟𝑝∗ ) = 9.21%
and risk of 𝜎𝑝∗ = 16.92%. Given 𝑟𝑓 = 2.0% and Low Risk Fund’s risk aversion coefficient 𝐴 =
3.55, what is the Fund’s optimal allocation to risky assets 𝑦 ∗ ? What is the expected return
𝐸(𝑟𝐶∗ ), risk 𝜎𝐶∗ and utility score of the optimal complete portfolio 𝐶 ∗ ?
∗ 𝐸(𝑟𝑝∗ ) – 𝑟𝑓
➢ Optimal risky allocation: 𝑦 = 2 = (.0921 – .02) / (3.55 x .16922) = 70.9%
𝐴 𝜎𝑝∗

➢ Expected return of the optimal complete portfolio 𝐶 ∗ :

𝐸(𝑟𝐶∗ ) = 𝑟𝑓 + 𝑦[𝐸(𝑟𝑝∗ ) – 𝑟𝑓 ] = .02 + .709 (.0921 - .02) = 7.11%
➢ Risk of 𝐶 ∗ : 𝜎𝐶∗ = 𝑦𝜎𝑃∗ = .709 (.1692) = 12.0%
1 2
➢ Utility score of 𝐶 ∗ : U = 𝐸(𝑟𝐶∗ ) − 𝐴 𝜎𝐶∗ = .0711 - 1/2 (3.55 x .122) = .0456
2
28
Separation Theorem

FINS5513 29
The CAL is Common to All Investors
❑ 𝐶𝐴𝐿𝑃∗ (or simply the CAL) caters to all risk tolerances and provides the highest possible
return for each level of risk
❑ Therefore, regardless of an individual investors’ level of risk aversion – defined by their risk
aversion coefficient 𝐴 - EVERY rational investor will invest along 𝐶𝐴𝐿𝑃∗
➢ Because for EVERY level of risk aversion, 𝐶𝐴𝐿𝑃∗ gives the highest return

❑ 𝐶𝐴𝐿𝑃∗ is the Capital Allocation Line (CAL) for ALL investors

❑ ALL rational investors will form an Optimal Complete Portfolio 𝐶 ∗ along the CAL
➢ Where they sit along this CAL will depend on their optimal risky asset allocation 𝑦 ∗ which
depends on the investors’ level of risk aversion

30
Separation Theorem
❑ The Separation Theorem states portfolio optimisation may be separated into two independent
steps:
1. Determine the CAL and optimal risky portfolio, 𝑃∗ (common to all investors)
2. Determine the share of wealth which will be invested in 𝑃∗ (the optimal allocation to risky
assets 𝑦 ∗ ) based on individual risk aversion (specific to the individual investor). This
step defines the investor’s Optimal Complete Portfolio 𝐶 ∗
❑ Separation theorem
CA* is the Optimal
Complete Portfolio ➢ Investors with different risk aversion
for investor A C B *
(𝐴) have different 𝐶 ∗
E(r)
P* ➢ Investor A has steeper indifference
curves and is more risk averse than B
CA* (higher risk aversion coefficient 𝐴)
CB* is the Optimal
rf Complete Portfolio ➢ But they invest in the same 𝑷∗
for investor B
➢ They only differ in terms of their
allocation (𝒚∗ ) to 𝑷∗ 31
σ
Summary of the Markowitz Optimisation Model
❑ Lets summarise the Markowitz Portfolio Optimisation model in 3 important steps:
1. Identify the optimal risk-return combinations from all risky assets to form the minimum
variance frontier. Discard all portfolios below the GMVP, thereby identifying the efficient
frontier
2. Derive the CAL by linking the risk-free asset with the portfolio on the efficient frontier with the
highest Sharpe ratio – which is the point of tangency with the efficient frontier. This portfolio
is known as the Optimal Risky Portfolio 𝑷∗
3. All rational investors invest in 𝑃∗ , regardless of their risk aversion. Individually, investors
decide where they sit along the CAL by determining their optimal risky allocation (𝒚∗ ) to 𝑃∗
➢ More risk averse investors put more in the risk-free asset
➢ Less risk averse investors put more in P*
➢ This last step identifies the investors’ Optimal Complete Portfolio 𝑪∗
Video 3AV4: ““In Pursuit of the Perfect Portfolio: Harry M. Markowitz”
Further Reading 3AR1: “Markowitz Mellon Survey on MPT” 32
 Excel 3AE1: “3A - Deriving the Optimal Complete Portfolio - 2 Assets”

Summary of the Markowitz Optimisation Model

Curve/Line Graphical Representation Portfolio/Plot Point
Minimum Variance Frontier GMVP
(Lowest risk at each return) Turning point Global Minimum
Variance Portfolio
Common/Universal

(lowest risk)

Efficient Frontier
(All efficient portfolios at each return)
GMVP

Capital Allocation Line 𝑷∗

𝑷∗
(Highest Sharpe ratio on efficient From rf Tangent to EF
Optimal Risky Portfolio
frontier) 𝒓𝒇
Individual

Highest Indifference Curve 𝑷∗

𝑪∗
(Attainable risk-return combinations 𝑪∗ Tangent to CAL
Opt. Complete Portfolio
which give highest utility) (at 𝒚∗ - optimal risky share) 33
Leveraged Portfolios

FINS5513 34
Borrowing Constraints
❑ The optimal allocation to risky assets 𝑦 ∗ can be > 1 for (less risk averse) investors
❑ This means the investor takes a levered position in 𝑃∗ - that is, borrowing to invest
➢ Points on the CAL to the left of 𝑃 ∗ represent lending (investing) at the risk-free rate 𝑟𝑓 and
points to the right represent borrowing at 𝑟𝑓
❑ In reality, although we can invest at 𝑟𝑓 (by buying government bonds) we generally can’t
borrow at 𝑟𝑓
➢ Only AAA Governments borrow at 𝑟𝑓 , others borrow at a higher rate reflecting their higher
risk
❑ Let’s assume that we are able to borrow at some higher interest rate 𝑟𝑏 and that 𝑟𝑏 > 𝑟𝑓
➢ In this case, for those (less risk averse) investors who would like to borrow to invest, the
section of CAL above 𝑃∗ is not obtainable, as those levered positions are based on 𝑟𝑓

Further Reading 3AR2: “Default Risk Premiums”

35
Borrowing Constraints
❑ A new optimal risky portfolio (𝑃𝑏∗ ) and associated CAL is derived for the borrowing rate 𝑟𝑏
❑ Only the section of CAL above 𝑃𝑏∗ is relevant, as this is the CAL for borrowing (taking a
levered position in the optimal risky portfolio 𝑃𝑏∗ )
❑ Graphically the CAL has a kink where 𝑦 = 1 (at 𝑃∗ ) – the slope (Sharpe ratio) changes at 𝑃∗
❑ Unless explicitly asked, we assume no borrowing constraints
Lending at rf Borrowing at rb

E(r)
P* Pb* Sharpe ratio changes at 𝑃∗ .
The borrowing CAL has a
rb lower Sharpe ratio than the
investing CAL. This results
in a kink in the CAL
rf

36
σ
Next Lecture
❑ BKM Chapter 9

❑ 3.3 General Equilibrium: Derivation of the CAPM

❑ 3.4 Interpreting the CAPM

37