MATH4512 Fundamental of Mathematical Finance (2022 Spring)

Suggested Solution of Assignment 3

Problem 1 (Required, 30 marks)
We consider a market with 1 riskfree asset and 𝑁 risky assets. The following table shows the
Sharpe ratio and expected return of 5 mutual funds.
Expected return Standard Sharpe ratio
deviation of
return
Fund A 0.30 ?? 0.45
Fund B 0.19 ?? 0.4
Fund C 0.12 0.18 ??
Fund D 0.07 0.085 ??
Fund E 0.03 ?? 0
You are given that some of the funds are efficient. We also assume that CAPM formula is valid
in this problem. Answer the following questions based on the above information:
(a) Is it possible to construct a portfolio with expected return 𝜇𝑃 ≥ 0.24 and 𝜎𝑃2 ≤ 0.15.
Explain your answer.
(b) Suppose that there is another fund (called Fund X) available in the market, the following
table summarizes the correlation coefficients between the return of fund X and the
return of each of the five funds.
Fund A Fund B Fund C Fund D Fund E
Correlation 𝜌𝑋𝐴 = 0.1 𝜌𝑋𝐵 = 0.3 𝜌𝑋𝐶 𝜌𝑋𝐷 = 0.2 𝜌𝑋𝐸
coefficient = 0.25 = −0.1
You are also given that 𝜎𝑋 = 0.3 for fund 𝑋. Calculate the expected return of fund X.
Solution
(a) Recall that
 the Sharpe ratio of efficient portfolio equals the slope of CML;
 the Sharpe ratio of other portfolios less than the slope of CML;
So the efficient portfolio (which lies on CML) should have the largest possible Sharpe
ratio. To check the feasibility of the investment goal, we first identify which of the
five funds is efficient by computing the Sharpe ratios of these 5 funds.
Since Fund E has Sharpe ratio 0, we deduce that
𝑟̅𝐸 − 𝑟𝑓
= 0 ⇒ 𝑟𝑓 = 𝑟̅𝐸 = 0.03.
𝜎𝐸
Then the Sharpe ratios of these 5 funds are found to be
Expected Standard deviation Sharpe ratio
return of return
Fund A 0.3 ?? 0.45
Fund B 0.19 ?? 0.4
Fund C 0.12 0.18 0.12 − 0.03
= 0.5
0.18
Fund D 0.07 0.085 0.07 − 0.03
≈ 0.47
0.085
Fund E 0.03 ?? 0
 Since fund C has the largest Sharpe ratio, so fund C is the efficient portfolio.

Given that 𝜇𝑃 ≥ 0.24 and 𝜎𝑃2 ≤ 0.15, the Sharpe ratio of the target portfolio is
𝜇𝑃 − 𝑟𝑓 0.24 − 0.03
𝑆𝑃 = ≥ = 0.542218 > 0.5.
𝜎𝑃2 √0.15
Recall that the Sharpe ratio of efficient portfolio (which is 0.5 in our case) represents
the largest possible Sharpe ratio among all feasible portfolio, thus the target
portfolio cannot be constructed.

(b) Using the extension of CAPM formula (use efficient portfolio (fund C) to replace the
market portfolio), the expected return of the fund X is
𝑟̅𝑋 = 𝑟𝑓 + 𝛽𝑋𝐶 (𝑟̅𝑐 − 𝑟𝑓 ).
Note that
𝑐𝑜𝑣(𝑟𝑋 , 𝑟𝐶 ) 𝜌𝑋𝐶 𝜎𝑋 𝜎𝐶 0.25(0.3)
𝛽𝑋𝐶 = = = = 0.416667.
𝑣𝑎𝑟(𝑟𝐶 ) 𝜎𝐶2 0.18
Hence, we deduce that
𝑟̅𝑋 = 0.03 + (0.416667)(0.12 − 0.03) = 0.0675 (𝑜𝑟 6.75%).

Problem 2 (Required, 30 marks)

The following table shows the betas and actual expected returns of 4 mutual funds in the
market:
Beta Actual expected
return
Fund 1 0.5 7.2%
Fund 2 0.8 10%
Fund 3 1.4 12%
Fund 4 1.9 16%
The return rate of riskfree asset is 𝑟𝑓 = 5% and the market portfolio exists. You are given that
 Exactly two of the funds perform better than the market expectation (i.e. Actual
expected return is strictly higher than market expected return)
 Exactly one of the funds performs worse than the market expectation.
(a) Which of the funds lie on security market line (SML)? Explain your answer.
(😊Hint 1: Try to draw a suitable figure and visualize the problem.)
(😊Hint 2: Which performance indexes will be useful in this context?)
(b) A fund manager launches a new fund (called Fund X) which 25% of the capital will be
invested in Fund 1, 35% of the capital will be invested in Fund 3 and 40% of the capital
will be invested in Fund 4. Compute the Jensen alpha and Treynor ratio of the Fund X.
Solution of (a)
Recall that
 If the fund lies on the SML, then the Treynor ratio 𝑇𝑖 should be equal to slope of SML.
 If the fund performs better than the market expectation, then it must lie above SML
and the corresponding Treynor ratio 𝑇𝑖 should be greater than the slope of SML.
  If the fund performs worse than the market expectation, then it lies below SML and
the Treynor ratio 𝑇𝑖 should be less than the slope of SML.
Using the given information, we deduce that the Treynor ratio of the required fund (lie on
SML) should rank third among those 4 funds. The Treynor ratios of these 4 funds are found
to be
Beta Actual expected Treynor ratio
return
Fund 1 0.5 7.2% 0.072 − 0.05
= 0.044 (4𝑡ℎ )
0.5
Fund 2 0.8 10% 0.1 − 0.05
= 0.0625 (1𝑠𝑡 )
0.8
Fund 3 1.4 12% 0.12 − 0.05
= 0.05 (3𝑟𝑑 )
1.4
Fund 4 1.9 16% 0.16 − 0.05
= 0.0579 (2𝑛𝑑 )
1.9
So we conclude that fund 3 lies on SML.
Solution of (b)
Firstly, the actual expected return of the fund X is
𝑟̅𝑋𝐴 = 0.25(0.072) + 0.35(0.12) + 0.4(0.16) = 0.124.
On the other hand, the beta of the fund X is
𝛽𝑋𝑀 = 0.25(0.5) + 0.35(1.4) + 0.4(1.9) = 1.375.
Since Fund 3 lies on SML, so the equation of SML is found to be
𝑟̅ − 0.05 0.12 − 0.05
= ⇒ 𝑟̅ = 0.05 + 0.05𝛽.
𝛽−0 1.4 − 0
Thus, the market expected return of fund X can be computed as
𝑟̅𝑋 = 0.05 + 0.05(1.375) = 0.11875.
Therefore, the Jensen alpha of fund X can be computed as
𝛼 = 𝑟̅𝑋𝐴 − 𝑟̅𝑋 = 0.00525 (𝑜𝑟 0.525%).
On the other hand, the Treynor ratio can be computed as
𝑟̅𝑋𝐴 − 𝑟𝑓
𝑇𝑋 = = 0.053818.
𝛽𝑋𝑀
(*Remark of (b): Alternatively, one can calculate the market expected return of fund X via
CAPM formula. Since fund 3 lies on SML, then the expected return of fund 3 should satisfy
the CAPM formula
0.12
⏟ = 0.05 ⏟ + 1.4(𝜇𝑀 − 0.05) ⇒ 𝜇𝑀 = 0.1.
𝑟̅3 𝑟𝑓 +𝛽3 (𝜇𝑀 −𝑟𝑓 )
Thus, the market required return of fund X can be calculated as
𝑟̅𝑋 = ⏟
0.05 + 1.375(0.1 − 0.05) = 0.11875.
𝑟𝑓 +𝛽𝑋 (𝜇𝑀 −𝑟𝑓 )
Problem 3 (Required, 40 marks)
We consider a market with 1 riskfree asset and 𝑁 risky assets. The following table shows the
expected return 𝜇𝑃 and standard deviation 𝜎𝑃 of 4 portfolios consists of these assets:
Expected return Standard
𝑟̅𝑃 deviation 𝜎𝑃
Portfolio 1 𝑤 ⃗⃗ 1 0.21 0.42
Portfolio 2 𝑤 ⃗⃗ 2 0.17 0.405
Portfolio 3 𝑤 ⃗⃗ 3 0.13 0.31
Portfolio 4 𝑤 ⃗⃗ 4 0.08 0.25
The return rate of riskfree asset is 𝑟𝑓 = 0.04. You are given that some of the portfolios are
efficient.
(a) Suppose that an investor is seeking for minimum variance portfolio with return not less
than 12%, find this portfolio and express your answer in terms of 𝑤 ⃗⃗ 𝑓 . 𝑤
⃗⃗ 1 , … , 𝑤
⃗⃗ 4. Here,
𝑤
⃗⃗ 𝑓 denotes the riskfree asset.
(b) Suppose that an investor is seeking for portfolio with maximum expected return and the
variance of portfolio return not more than 0.15, find this portfolio and express your
answer in terms of 𝑤 ⃗⃗ 𝑓 . 𝑤
⃗⃗ 1 , … , 𝑤
⃗⃗ 4. Here, 𝑤
⃗⃗ 𝑓 denotes the riskfree asset.
Provide full justification to your answer.
Solution
(a) To obtain the optimal portfolio required, we consider the following diagram:
𝑟̅
Optimal
Portfolio
𝜇𝑃 = 0.12

𝑟𝑓 = 0.04 Feasible region

𝜎
Hence, the optimal portfolio (denoted by 𝑤 ⃗⃗ ∗ ) is the minimum variance portfolio with
expected return 𝜇𝑃 = 0.12.
By one fund theorem, the portfolio can be expressed as the combination of riskfree
asset and market portfolio 𝑤 ⃗⃗ 𝑀
⃗⃗ ∗ = 𝛼𝑤
𝑤 ⃗⃗ 𝑓 + (1 − 𝛼)𝑤 ⃗⃗ 𝑀 .
Note that some portfolios are efficient, it follows from one-fund theorem that any
efficient portfolio (portfolio on frontier curve, denoted by 𝑤 ⃗⃗ 𝐸 ) can be expressed as
1 𝛽
𝑤 ⃗⃗ 𝑓 + (1 − 𝛽)𝑤
⃗⃗ 𝐸 = 𝛽𝑤 ⃗⃗ 𝑀 ⇒ 𝑤 ⃗⃗ 𝑀 = 𝑤
⃗⃗ 𝐸 − 𝑤
⃗⃗ .
1−𝛽 1−𝛽 𝑓
Hence, the optimal portfolio can also expressed as
1 𝛽 𝛽(1 − 𝛼) 1−𝛼
⃗⃗ ∗ = 𝛼𝑤
𝑤 ⃗⃗ 𝑓 + (1 − 𝛼) ( 𝑤
⃗⃗ 𝐸 − 𝑤⃗⃗ 𝑓 ) = (𝛼 − )𝑤⃗⃗ 𝑓 + 𝑤
⃗⃗ .
1−𝛽 1−𝛽 ⏟ 1−𝛽 1−𝛽 𝐸
⏟
γ 1−𝛾
It remains to identify an efficient portfolio.
 We first recall the following fact:
 The Sharpe ratio of any efficient portfolio equals the slope of CML.
 The Sharpe ratio of other portfolio (which is not efficient or not optimal)
must be less than the slope of CML.
This implies that
𝑆ℎ𝑎𝑟𝑝𝑒 𝑟𝑎𝑡𝑖𝑜 𝑆ℎ𝑎𝑟𝑝𝑒 𝑟𝑎𝑡𝑖𝑜 𝑜𝑓
{ } ≤ {𝑆𝑙𝑜𝑝𝑒 𝑜𝑓 𝐶𝑀𝐿} = { }
𝑜𝑓 𝑎𝑛𝑦 𝑝𝑜𝑟𝑡𝑓𝑜𝑙𝑖𝑜 𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡 𝑝𝑜𝑟𝑡𝑓𝑜𝑙𝑖𝑜
So the efficient portfolio should have the highest Sharpe ratio among all feasible
portfolio. Since some of the portfolios are efficient in the problem, it follows that the
portfolio with highest Sharpe ratio is efficient.
Next, we shall compute the Sharpe ratio of these four portfolios:
Expected return Standard Sharpe ratio
𝑟̅𝑃 deviation 𝜎𝑃
Portfolio 1 0.21 0.42 0.404762
Portfolio 2 0.17 0.405 0.320988
Portfolio 3 0.13 0.31 0.290323
Portfolio 4 0.08 0.25 0.16
Since portfolio 1 has largest Sharpe ratio, so the portfolio 1 should be efficient.
By the above formula, the optimal portfolio can be expressed as
⃗⃗ ∗ = γ𝑤
𝑤 ⃗⃗ 𝑓 + (1 − 𝛾)𝑤 ⃗⃗ 1 .
∗
Since the expected return of 𝑤 ⃗⃗ is 0.12, we have
9
0.12 = γ(0.04) + (1 − 𝛾)(0.21) ⇒ 𝛾 = .
17
Therefore, the optimal portfolio is given by
9 8
⃗⃗ ∗ =
𝑤 𝑤
⃗⃗ 𝑓 + 𝑤 ⃗⃗ .
17 17 1

(b) We consider the following diagram:

𝑟̅
Optimal
Portfolio

𝑟𝑓 = 0.07 Feasible region

𝜎
𝜎𝑃 = √0.15

So the optimal portfolio is the efficient minimum variance portfolio with 𝜎𝑃 =

√0.15. Using the result of (a), the optimal portfolio can be expressed as
𝑤⃗⃗ ∗ = 𝛿𝑤
⃗⃗ 𝑓 + (1 − 𝛿)𝑤
⃗⃗ 1 .
2
By setting 𝜎𝑃 = 0.15, we deduce that
𝜎𝑃2 = 𝑉𝑎𝑟(𝑟 ∗ ) = (1 − 𝛿)2 𝑉𝑎𝑟(𝑟1 ) ⇒ 0.15 = (1 − 𝛿)2 (0.42)2
 ⇒ 𝛿 = 0.077861 𝑜𝑟 1.922139
- If 𝛿 = 0.077861, the expected return of the portfolio is 𝜇𝑃 = 0.077861(0.04) −
(1 − 0.077861)(0.21) = 0.196764.
- If 𝛿 = 1.922139, the expected return of the portfolio is 𝜇𝑃 = 1.922139(0.04) −
(1 − 1.922139)(0.21) = −0.11676.
Hence, we take 𝛿 = 0.077861 and the optimal portfolio is given by
𝑤⃗⃗ ∗ = 0.077861𝑤⃗⃗ 𝑓 + 0.922139𝑤⃗⃗ 1 .
(*Note: In fact, the portfolio with 𝛿 = 1.922139 corresponds to inefficient minimum
variance portfolio (on lower half of frontier curve) with 𝜎𝑃 = √0.15, see the above
diagram).

Bonus Problem 1 (Optional, 50 marks)

We consider a market with 1 riskfree asset and 𝑁 risky assets. It is given that
 The return of riskfree asset is 𝑟𝑓 .
𝑟̅1
𝑟̅2
 The expected return and covariance matrix of 𝑁 risky assets are 𝜇 = ( ) and Ω =
⋮
𝑟̅𝑁
𝜎11 𝜎12 ⋯ 𝜎1𝑁
𝜎21 𝜎22 ⋯ 𝜎2𝑁
( ⋮ ⋮ ⋱ ⋮ ).
𝜎𝑁1 𝜎𝑁2 ⋯ 𝜎𝑁𝑁

𝑏
We assume that 𝑎 > 𝑟𝑓 ,
(a) Suppose that an investor would like to invest his capital into 𝑁 risky assets and wish to
choose a portfolio with maximum Sharpe ratio.
(i) Prove that this optimal portfolio must be efficient.
(ii) Hence, determine this portfolio. Express your answer in terms of 𝑎, 𝑏, 𝑐, 𝜇 , Ω, 𝑟𝑓 .
(iii) Suppose that the investor also requires that the variance of portfolio return
2
cannot exceed 𝜎𝑚𝑎𝑥 , determine the corresponding optimal portfolio (if any) and
2
express your answer in terms of 𝑎, 𝑏, 𝑐, 𝜇 , Ω, 𝑟𝑓 , 𝜎𝑚𝑎𝑥 .
(Hint: You may need to consider several cases. On the other hand, you can
consider feasible region.)

(b) Another investor wishes to choose a minimum variance portfolio (with risky assets only)
with Sharpe ratio not less than 𝑆 ∗ > 0, determine this optimal portfolio (if any). Express
your answer in terms of 𝑎, 𝑏, 𝑐, 𝜇 , Ω, 𝑟𝑓 .
(Hint: Similar to (a)(iii), you may need to consider several cases.)

Solution
(a) (i) We let 𝜇𝑃 and 𝜎𝑃 be the expected return and standard deviation of return of the
optimal portfolio. Suppose that the portfolio is inefficient, there exists another
portfolio 𝑄 such that
 𝜇𝑃 ≤ 𝜇𝑄 𝑎𝑛𝑑 𝜎𝑃 ≥ 𝜎𝑄
With at least one inequality holds as strict inequality (i.e. either 𝜇𝑃 < 𝜇𝑄 or 𝜎𝑃 > 𝜎𝑄
or both). Then the Sharpe ratio of portfolio 𝑄 (denoted by 𝑆𝑄 ) is given by
𝜇𝑄 − 𝑟𝑓 𝜇𝑃 − 𝑟𝑓
𝑆𝑄 = > = 𝑆𝑃 .
𝜎𝑄 𝜎𝑃
Then the optimal portfolio P does not have the highest Sharpe ratio and there is a
contradiction.

(ii) According to the result in Example 9 of Lecture Note 3, the tangency portfolio (it
𝑏
exists and is efficient since > 𝑟𝑓 ) has the largest Sharpe ratio among all efficient
𝑎
portfolios. Therefore the required optimal portfolio is seen to be
1
⃗⃗ ∗ = 𝑤
𝑤 ⃗⃗ 𝑀 = Ω−1 (𝜇 − 𝑟𝑓 ⃗1).
𝑏 − 𝑎𝑟𝑓

(iii) Firstly, the variance of return of the tangency portfolio (the portfolio with
highest Sharpe ratio) is
1 1
𝜎𝑀2
=𝑤 𝑇
⃗⃗ 𝑀 Ω𝑤
⃗⃗ 𝑀 = 𝑤 𝑇
⃗⃗ 𝑀 (𝜇 − 𝑟𝑓 ⃗1) = (𝜇 − 𝑟𝑓 )
𝑏 − 𝑎𝑟𝑓 𝑏 − 𝑎𝑟𝑓 𝑀
1 𝑎𝑟𝑓2 − 2𝑏𝑟𝑓 + 𝑐 𝑎𝑟𝑓2 − 2𝑏𝑟𝑓 + 𝑐
= (𝑟 + − 𝑟𝑓 ) = 2 .
𝑏 − 𝑎𝑟𝑓 𝑓 𝑏 − 𝑟𝑓 𝑎 (𝑏 − 𝑟 𝑎)
𝑓
We consider the following two cases:
 If 𝜎𝑀 2
≤ 𝜎𝑃2 , then the optimal portfolio remains to be
1
⃗⃗ ∗ = 𝑤
𝑤 ⃗⃗ 𝑀 = Ω−1 (𝜇 − 𝑟𝑓 ⃗1).
𝑏 − 𝑎𝑟𝑓
 If If 𝜎𝑀 2
> 𝜎𝑃2 , we observe that the Sharpe ratio of the efficient portfolio
increases (from Example 9 of Lecture Note 3) with respect to expected return
𝜇𝑃 for 𝜇𝑃 ≤ 𝜇𝑀 . Therefore, the optimal portfolio with maximum Sharpe ratio
is the efficient minimum variance portfolio (with risky assets only) with 𝜎𝑃2 =
2
𝜎𝑚𝑎𝑥 . Tangency
𝑟̅ portfolio
The Sharpe ratio Optimal
portfolio
increases along
efficient frontier for
𝜇 𝑃 < 𝜇𝑀

𝜎
𝜎𝑚𝑎𝑥
 Recall that the equation of the frontier curve is
2
𝑎𝜇𝑃2 − 2𝑏𝜇𝑃 + 𝑐
𝜎𝑃 = .
𝑎𝑐 − 𝑏 2
By setting 𝜎𝑃2 = 𝜎𝑚𝑎𝑥
2
, the expected return of the portfolio satisfies
2 2 (𝑎𝑐
𝑎𝜇𝑃 − 2𝑏𝜇𝑃 + (𝑐 − 𝜎𝑚𝑎𝑥 − 𝑏 2 )) = 0

2𝑏 ± √4𝑏 2 − 4𝑎(𝑐 − 𝜎𝑚𝑎𝑥

2 (𝑎𝑐 − 𝑏 2 ))

⇒ 𝜇𝑃 =
2𝑎
𝑏
Since the portfolio is efficient and 𝜇𝑃 > 𝑎, then

2𝑏 − √4𝑏 2 − 4𝑎(𝑐 − 𝜎𝑚𝑎𝑥

2 (𝑎𝑐 − 𝑏 2 ))

𝜇𝑃 = .
2𝑎
Thus the optimal portfolio can be expressed as
𝑐 − 𝑏𝜇𝑃 𝑎𝜇 − 𝑏
⃗⃗ ∗ = Ω−1 (
𝑤 ⃗1 + 𝑃 𝜇 ).
𝑎𝑐 − 𝑏 2 𝑎𝑐 − 𝑏 2

(b) To facilitate the analysis, we let 𝑆𝑔 and 𝑆𝑚 be the Sharpe ratio of global minimum
𝑏
variance portfolio and tangency portfolio respectively. Since > 𝑟𝑓 and tangency
𝑎
𝑟̅𝑖 −𝑟𝑓
portfolio is efficient, it follows that 𝑆𝑚 > 𝑆𝑔 > 0 since 𝑆 = and tangency
𝜎𝑖
portfolio has the highest Sharpe ratio after all efficient portfolios.

We consider the following three cases:

Case 1: 𝑆𝑔 > 𝑆 ∗ > 0

Since global minimum variance portfolio has the lowest variance among all feasible
portfolios, it follows that the required portfolio is the global minimum variance
Ω−1 ⃗1
⃗⃗ ∗ = 𝑤
portfolio and 𝑤 ⃗⃗ 𝑔 = .
𝑎

Case 2: 𝑆𝑚 > 𝑆 ∗ > 𝑆𝑔 > 0

𝑟̅ −𝑟𝑓
𝑟̅ Slope = = 𝑆∗
𝜎

Required
portfolio
The region above
the bold dash line
Feasible
𝑏/𝑎 represents the
region
portfolios with
𝑟𝑓
Sharpe ratio > 𝑺∗

𝜎
One can observe from the above diagram that the optimal portfolio is the minimum
variance portfolio with Sharpe ratio 𝑆 = 𝑆 ∗ and lowest expected return (*Note: For
the case if there are more than 1 minimum variance portfolios with Sharpe ratio 𝑆 ∗ .)

It remains to find the expected return of this optimal portfolio (denoted by 𝜇𝑃 ). Since
the portfolio lies on frontier curve and have Sharpe ratio 𝑆 ∗ , 𝜇𝑃 must satisfy
𝜇𝑃 − 𝑟𝑓
= 𝑆∗
𝜎𝑃 𝜇𝑃 − 𝑟𝑓 2 𝑎𝜇𝑃2 − 2𝑏𝜇𝑃 + 𝑐
⇒ ( ) =
2
𝑎𝜇𝑃2 − 𝑏𝜇𝑃 + 𝑐 𝑆∗ 𝑎𝑐 − 𝑏 2
{ 𝜎 𝑃 =
𝑎𝑐 − 𝑏 2
𝑎 1 2
2𝑟𝑓 2𝑏 𝑐 𝑟𝑓2
⇒( − ) 𝜇 𝑃 + ( − ) 𝜇 𝑃 + − =0
⏟𝑎𝑐 − 𝑏 2 (𝑆 ∗ )2 ⏟(𝑆 ∗ )2 𝑎𝑐 − 𝑏 2 ⏟ − 𝑏 2 (𝑆 ∗ )2
𝑎𝑐
𝑎′ 𝑏′ 𝑐′

−𝑏 ′ − √𝑏 ′ 2 − 4𝑎′ 𝑐 ′
⇒ 𝜇𝑃 = .
2𝑎′
Hence, the required portfolio is found to be
𝑐 − 𝑏𝜇𝑃 𝑎𝜇 − 𝑏
⃗⃗ ∗ = Ω−1 (
𝑤 ⃗1 + 𝑃 𝜇)
𝑎𝑐 − 𝑏 2 𝑎𝑐 − 𝑏 2
Case 3: 𝑆 ∗ > 𝑆𝑚 > 𝑆𝑔 > 0.
Since the tangency portfolio has the highest Sharpe ratio among all feasible
portfolio, so there is no portfolio which has Sharpe ratio greater than 𝑆 ∗ . Therefore
such optimal portfolio does not exist
𝑟̅ −𝑟𝑓
Slope = = 𝑆∗
𝑟̅ 𝜎

The region above

the bold dash line
Feasible
𝑏/𝑎 represents the
region
portfolios with
𝑟𝑓
Sharpe ratio > 𝑺∗