CONSTRUCTING OPTIMAL PORTFOLIO: SHARPE’S SINGLE

INDEX MODEL

Tirthank Shah
MBA,CFA

Abstract:

This paper is an attempt to construct optimal portfolio by applying

Sharpe’s Single Index Model. Explanation is provided wherever
necessary related to design of the Single Index Model .The data taken
for the application of single index model is 50 companies part of CNX
NSE Nifty Fifty Index for the time period of Dec-08 to Dec-12.This model
generates cut off rate and only those securities which have higher
excess return to beta ratio than cut off rate are included in optimal
portfolio.

Keywords: Sharpe Single Index Model, Stock selection, Optimal

Portfolio, Beta.

Page 1

Electronic copy available at: http://ssrn.com/abstract=2459417

Biographical Note:

Author is having 7 years of financial industry experience which

comprises trading and fundamental/technical research in international
markets like USA, UK and Europe for financial products like bonds,
equity, currency, interest rate, commodity etc. and treasury
management.

He is visiting faculty to various B-Schools for subjects like Security

Analysis and Portfolio Management and Risk (Derivatives)
Management.

Objective of the Study:

To construct optimal portfolio using Sharpe’s Single Index Model

Research Methodology:

For the research purpose secondary data is used and source of the data
is www.nseindia.com. For constructing optimal portfolio 50 companies
which are part of CNX NSE Nifty Fifty Index are selected for time
duration from Dec-08 to Dec-12. Geometric mean of yearly return is
taken for every security and market index which is CNX NSE Nifty Fifty
Index for this research purpose. The single Index model formulates cut-
off rate based on data inputs and selects only those securities which
have higher excess return to beta ratio as compare to cut-off rate. Then
based on residual variance (unsystematic risk) of the security, excess
return to beta ratio, beta of the security and cut-off rate, proportion or
weightage of the investment of the selected security is computed.

Page 2

Electronic copy available at: http://ssrn.com/abstract=2459417

Various financial and statistical methodologies are used for
implementation of the model.

1. The Frame Work: Single Index Model

The simplicity is the most important feature of the Sharpe’s Single

Index Model over Markowitz‘s Model. Markowitz model is very
information demanding. For large set of securities in portfolios, working
on large set of covariance terms become unviable. Taking idea from
what Markowitz suggested that an index to which securities are related
can be used for covariance generation, William Sharpe formulated
Single Index Model. Sharpe’s single index model reflects that return on
security is function of return of the market index. The single index
model, for n securities requires only (3n+2) data inputs which are Beta
( ), Alpha ( ), Residual Variance (Unsystematic risk)( 2 ) for every
security and estimate of expected return of market index( ) and
estimate of variance of the return of the market index( 2 ).It is
observed through various research studies done in past that single
index model has performed well. It has created major advancement in
the world of portfolio management by reducing the inputs
requirements and still attaining fairly acceptable results.

1.2 Single Index Model: The Regression Equation

Regressing return of the security on return of the market index,

generates Alpha ( ) and Beta ( ). Estimation of the Beta ( ) which is
sensitivity coefficient of the security on market index is possible as
single index model is linear. Estimation of regression is done based on
Page 3
collected historical data of securities and market index. The regression
equation is

= + +
Where,

= Return on security i

= Constant term (Security‘s return when market excess return is

zero)

= Beta of the security (The slope coefficient)

= Return on market index

= Error term

The key assumptions on which single index model is based on are;

i. The error term ( ) is zero mean and has finite variance.

ii. Securities are related through common response to return of market

index only. Meaning there by error term for one security is not
correlated with error term for any other security. , =0

iii. There is no correlation between error term and return on market

index. ( , )=0

It is important to understand that from the above regression equation

return has two components: Alpha ( ) return and market related
return ( ). Same way, security‘s variance has two components:
security related risk ( 2 ) and market related risk ( 2 2 ).

The expected return of the security is

Page 4
 = + +
As zero in value so,

= +
The variance of the return of the security is

= +

Covariance between any two securities can be stated as

Where,

= Covariance between security i and j

= Beta of Security i

= Beta of security j

= Variance of the return of Market index

1.3 Optimal Portfolio Construction

With the help of Single index model, portfolio managers and security
analysts can easily identify based on security’s excess return to beta
ratio, whether security should be included as part of optimal portfolio
or not. Single index model gives ‘single value’ which explains the
desirability of including any security in the portfolio. This ‘single value’
is excess return to beta ratio of that security. This excess return to beta
ratio shows how much additional return for the security is generated
for every unit of systematic risk (non-diversifiable risk).

Page 5
Symbolically, excess return to beta ratio can be shown as below;
( − )

Where,

= Expected Return on security i

= Risk free rate

= Beta of the security i

Securities are ranked based on excess return to beta ratio from Highest
to lowest, this ranking represent the desirability of including that
security in portfolio. So, if security with particular ranking is included in
portfolio, all the securities with ranking above will be included as well.
Same way if security with particular ranking is not part of the portfolio
all securities below that security in terms of ranking will also be
excluded from portfolio. Selection of the security is done based on cut-
off rate. So, all the securities having
( Ri Rf )
≥ C ∗ ( C ∗ = cut-off rate) will be part of portfolio and securities
βi
( Ri Rf )
having < C ∗ will not be part of portfolio.
βi

The steps to be followed for which security and in what proportion it is

included in optimal portfolio by applying Single index model are;
( Ri Rf )
i. To calculate excess return to beta ratio ( ) for each security
βi
under consideration.

Page 6
ii. Ranking of the all securities based on excess return to beta ratio from
highest to lowest.

iii. Calculating cut-off rate ( C ∗ ).

( Ri Rf )
iv. Finding out optimal cut-off rate up to which all securities ≥
βi
C ∗ and after that point all securities having
( Ri Rf )
< C∗ .
βi

( Ri Rf )
v. Adding all those securities having ≥ C ∗ into optimal
βi
portfolio.

vi. Calculating proportion of each selected security in the optimal

portfolio.

The proportion to be invested in each selected security is calculated in

following way;

=
∑

Where,

∗
= ( )− i= 1,2,3,……K out of N.

∗
1.4 Cut-off rate ( ) calculation:
∗
Cut-off rate ( ) is calculated in following manner;

Page 7
 ∗
∑ ( − ) ⁄
=
+ ∑

Optimal portfolio can be tested by applying Kuhn-Tucker conditions, if it

satisfies, it is optimal portfolio. Single index model also satisfy Kuhn-
Tucker conditions.

1.5 Constructing Optimal portfolio

For this research paper purpose the data is 50 companies which are
part of CNX NSE Nifty Fifty Index. Geometric mean of yearly return from
Dec-08 to Dec-12 is taken as average return for each security and for
market index. Negative beta securities are excluded for calculation
purpose. Risk free rate is taken as 5.98% (Gov. Bond yield 7.98% - India
default spread based on rating 2%).Refer Appendix 1 for the calculation
snap shot.

1.6 Inference

Based on Single Index Model run for 50 companies which are part of
CNX NSE Nifty Fifty Index with market index as CNX NSE Nifty Fifty, the
research has thrown interesting result. Only top two securities based
on ranking (excess return to beta ratio) namely 1. ITC Ltd.(1st Rank) and
2. Sun Pharmaceutical Industries Ltd. (2nd Rank) are becoming part of
optimal portfolio. Weightage of ITC Ltd is 95% and weightage of Sun

Page 8
Pharmaceutical Industries Ltd is 5%, together both companies comprise
( Ri Rf )
100% of the optimal portfolio. Excess return to beta ratio for
βi
ITC ltd is 97.01 and for Sun Pharmaceutical Industries Ltd is 85.31.

Concluding Remarks

With this research paper, it is clear that constructing optimal portfolio

with Sharpe’s Single Index Model is less time consuming and more
comfortable especially when we think it in context of security analysts
and portfolio managers in real world as compare to Markowitz’s Mean-
Variance Model. As Sharpe has argued and many past studies have
shown that there is considerable similarity between optimal portfolios
generated by Single Index Model and Markowitz’s Mean-Variance
Model. This research paper is an endeavor to show that Single Index
Model is better placed for constructing optimal portfolio in context of
real world.

Page 9
Appendix 1

Mean return of market index is 18.85%.

Sr. Security Name Mean Variance Beta Variance Residual Excess Excess
No Return%( ) Security ( ) Market Var. Return % Return
( ) ( ) ( − ) To Beta
( )

1 ACC Ltd. 31.42 10.85 0.76 16.98 1.00 25.44 33.39

2 Ambuja Cements Ltd. 30.13 2.89 0.39 16.98 0.35 24.15 62.38
3 Asian Paints Ltd. 49.18 22.00 1.09 16.98 1.79 43.20 39.60
4 Axis Bank Ltd. 28.04 34.49 1.36 16.98 3.16 22.06 16.24
5 Bajaj Auto Ltd. 81.68 253.48 3.45 16.98 51.86 75.70 21.97
6 Bank of Baroda 32.63 25.03 1.02 16.98 7.28 26.65 26.07
Bharat Heavy Electricals
7 Ltd. -4.33 27.09 1.24 16.98 0.98 -10.31 -8.32
8 BPCL 17.33 19.06 0.99 16.98 2.39 11.35 11.45
9 Cairn India Ltd. 16.70 9.78 0.68 16.98 1.92 10.72 15.76
10 Cipla Ltd. 22.06 15.61 0.94 16.98 0.54 16.08 17.07
11 DLF Ltd. -4.91 10.73 0.68 16.98 2.80 -10.89 -15.93
Dr. Reddy's Laboratories
12 Ltd. 40.57 43.94 1.47 16.98 7.29 34.59 23.54
13 GAIL (India) Ltd. 14.68 30.70 1.22 16.98 5.30 8.70 7.11
14 Grasim Industries Ltd. 26.96 23.68 1.01 16.98 6.45 20.98 20.83
15 HCL Technologies Ltd. 52.20 108.81 2.40 16.98 10.67 46.22 19.23
16 HDFC Bank Ltd. 35.78 12.33 0.79 16.98 1.61 29.80 37.50
17 Hero MotoCorp Ltd. 24.01 30.97 1.17 16.98 7.65 18.03 15.38
18 Hindalco Industries Ltd. 26.23 128.05 2.61 16.98 12.61 20.25 7.77
19 HDFC 29.22 13.82 0.88 16.98 0.52 23.24 26.27
20 I T C Ltd. 35.19 1.95 0.30 16.98 0.41 29.21 97.02
21 ICICI Bank Ltd. 26.25 34.37 1.36 16.98 2.84 20.27 14.87
22 IDFC Ltd. 26.55 62.80 1.84 16.98 5.27 20.57 11.17
23 IndusInd Bank Ltd. 82.56 154.77 2.95 16.98 6.95 76.58 25.95
24 Infosys Ltd. 20.07 50.76 1.49 16.98 13.29 14.09 9.49
25 Jaiprakash Associates Ltd. 15.05 101.83 2.25 16.98 15.66 9.07 4.03
26 Jindal Steel & Power Ltd. 31.04 354.81 4.02 16.98 80.18 25.06 6.23
27 Kotak Mahindra Bank Ltd. 38.09 33.73 1.34 16.98 3.17 32.11 23.94
28 Larsen & Toubro Ltd. 20.05 49.54 1.68 16.98 1.63 14.07 8.38
29 Lupin Ltd. 49.29 37.37 1.42 16.98 3.18 43.31 30.52
Mahindra & Mahindra
30 Ltd. 61.41 189.87 3.06 16.98 30.48 55.43 18.09
31 Maruti Suzuki India Ltd. 30.09 111.01 2.42 16.98 11.83 24.11 9.98
32 NMDC Ltd. -0.10 85.01 2.02 16.98 15.46 -6.08 -3.00
33 NTPC Ltd. -3.53 5.13 0.52 16.98 0.62 -9.51 -18.45
34 ONGC 12.59 17.21 0.96 16.98 1.48 6.61 6.87
35 Power Grid Corporation 8.39 3.41 0.33 16.98 1.51 2.41 7.19

Page 10
36 Punjab National Bank 13.41 20.49 1.04 16.98 2.10 7.43 7.14
37 Ranbaxy Laboratories Ltd. 18.82 32.60 1.37 16.98 0.58 12.84 9.35
38 Reliance Industries Ltd. 8.03 22.16 1.13 16.98 0.56 2.05 1.82
39 Sesa Goa Ltd. 22.84 399.04 4.35 16.98 77.48 16.86 3.87
40 State Bank of India 16.64 25.49 1.18 16.98 1.86 10.66 9.04
41 Sun Pharma 36.38 6.27 0.36 16.98 4.11 30.40 85.32
Tata Consultancy Services
42 Ltd. 51.41 99.07 2.11 16.98 23.62 45.43 21.55
43 Tata Motors Ltd. 76.84 345.41 4.28 16.98 34.95 70.86 16.57
44 Tata Power Co. Ltd. 10.17 25.91 1.22 16.98 0.64 4.19 3.43
45 Tata Steel Ltd. 18.51 100.28 2.36 16.98 5.76 12.53 5.31
46
UltraTech Cement Ltd. 50.66 35.03 1.34 16.98 4.48 44.68 33.31
47 Wipro Ltd. 29.55 93.94 2.09 16.98 19.64 23.57 11.27

Applying Single index model:

Ranking Security ( − ) ⁄ {( − ) ⁄ } ⁄ ⁄ ∗ Status
based on Name
Excess
return to
beta
ratio
1 I T C Ltd. 21.62 21.62 0.22 0.22 76.74 IN
2 Sun Pharma 2.63 24.25 0.03 0.25 77.58 IN
Ambuja
3 Cements Ltd. 26.96 51.21 0.43 0.69 68.76 OUT
Asian Paints
4 Ltd. 26.34 77.55 0.67 1.35 55.01 OUT
HDFC Bank
5 Ltd. 14.67 92.22 0.39 1.74 51.21 OUT
6 ACC Ltd. 19.41 111.63 0.58 2.32 46.86 OUT
UltraTech
7 Cement Ltd. 13.37 125.01 0.40 2.72 44.91 OUT
8 Lupin Ltd. 19.33 144.34 0.63 3.36 42.24 OUT
9 HDFC 39.23 183.57 1.49 4.85 37.38 OUT
Bank of
10 Baroda 3.74 187.31 0.14 5.00 37.06 OUT
IndusInd Bank
11 Ltd. 32.50 219.82 1.25 6.25 34.85 OUT
Kotak
Mahindra
12 Bank Ltd. 13.58 233.39 0.57 6.82 33.95 OUT
Dr. Reddy's
Laboratories
13 Ltd. 6.97 240.36 0.30 7.11 33.52 OUT

Page 11
14 Bajaj Auto Ltd. 5.03 245.39 0.23 7.34 33.17 OUT
Tata
Consultancy
15 Services Ltd. 4.06 249.45 0.19 7.53 32.88 OUT
Grasim
16 Industries Ltd. 3.28 252.73 0.16 7.69 32.63 OUT
HCL
Technologies
17 Ltd. 10.41 263.14 0.54 8.23 31.76 OUT
Mahindra &
18 Mahindra Ltd. 5.57 268.71 0.31 8.54 31.27 OUT
19 Cipla Ltd. 28.29 297.00 1.66 10.19 28.97 OUT
Tata Motors
20 Ltd. 8.67 305.67 0.52 10.72 28.37 OUT
21 Axis Bank Ltd. 9.49 315.16 0.58 11.30 27.75 OUT
Cairn India
22 Ltd. 3.80 318.96 0.24 11.54 27.50 OUT
Hero
23 MotoCorp Ltd. 2.76 321.72 0.18 11.72 27.31 OUT
24 ICICI Bank Ltd. 9.72 331.44 0.65 12.37 26.66 OUT
25 BPCL 4.70 336.14 0.41 12.78 26.17 OUT
26 Wipro Ltd. 2.51 338.65 0.22 13.01 25.92 OUT
27 IDFC Ltd. 7.19 345.84 0.64 13.65 25.23 OUT
Maruti Suzuki
28 India Ltd. 4.93 350.76 0.49 14.14 24.70 OUT
29 Infosys Ltd. 1.57 352.34 0.17 14.31 24.52 OUT
Ranbaxy
Laboratories
30 Ltd. 30.39 382.73 3.25 17.56 21.72 OUT
State Bank of
31 India 6.77 389.49 0.75 18.31 21.21 OUT
Larsen &
32 Toubro Ltd. 14.46 403.96 1.73 20.04 20.10 OUT
Hindalco
33 Industries Ltd. 4.19 408.14 0.54 20.57 19.78 OUT
Power Grid
34 Corporation 0.53 408.68 0.07 20.65 19.74 OUT
Punjab
35 National Bank 3.68 412.35 0.51 21.16 19.43 OUT
GAIL (India)
36 Ltd. 2.01 414.36 0.28 21.45 19.27 OUT
37 ONGC 4.29 418.65 0.62 22.07 18.92 OUT
Jindal Steel &
38 Power Ltd. 1.26 419.91 0.20 22.27 18.80 OUT
39 Tata Steel Ltd. 5.13 425.04 0.97 23.24 18.24 OUT
Jaiprakash
40 Associates Ltd. 1.31 426.35 0.32 23.56 18.05 OUT
41 Sesa Goa Ltd. 0.95 427.29 0.24 23.81 17.90 OUT
Tata Power
42 Co. Ltd. 8.01 435.30 2.33 26.14 16.61 OUT
43 Reliance 4.15 439.45 2.28 28.42 15.43 OUT

Page 12
 Industries Ltd.
44 NMDC Ltd. -0.80 438.65 0.27 28.69 15.26 OUT
Bharat Heavy
45 Electricals Ltd. -13.08 425.58 1.57 30.26 14.04 OUT
46 DLF Ltd. -2.66 422.92 0.17 30.43 13.87 OUT
47 NTPC Ltd. -7.84 415.08 0.42 30.85 13.43 OUT

Weightage of selected securities in optimal portfolio

Ranking Security ∗ = ( )− ∗
=
based on Name ∑
Excess return Weightage%
to beta ratio
1 I T C Ltd. 76.74 4.33 95
2 Sun Pharma 77.58 0.24 5

Portfolio return = 1 ∗ 1 + 2 ∗ 2

= (0.95) * (35.19) + (0.05) * (36.38)

Portfolio return = 35.25%

Covariance between return of the ITC Ltd and Sun Pharma = 0.022
2 2 2 2
Portfolio Risk = 1 ∗ 1 + 2 ∗ 2 + 1 2 12 1 2

= (0.95)2 ∗ ( .39)2 + (0.05)2 ∗ ( .50)2 + (0.95) ∗ (0.05) ∗ (0.0 )

Portfolio Risk (Standard Deviation) = 1.32%

Page 13
Reference:

Chandra, P. (2009). Investment Analysis and Portfolio

Management McGraw-Hill Publishing Company Ltd.

Fischer, D.E. and Jordan, R.J. (2000). Security analysis and Portfolio
Management. NY. Prentice Hall.

Elton,E and Gruber,M(2006).Modern Portfolio Theory and

Investment Analysis. NY. John Wiley & Sons

Reilly, F.K. and Brown, K.C. (2006). Investment Analysis and

Portfolio Management. New Delhi: CENGAGE Learning.

Sen, T. “Constructing an Optimal Portfolio With and Without Short

Selling Using Single Index Model”. Research Paper at SSRN.

www.nseindia.com

www.moneycontrol.com

Page 14