Capital Asset Pricing Model (CAPM)
1.0
Introduction
The effects of risk and
uncertainty upon asset prices, upon rational decision rules for individuals and
institutions to use in selecting security portfolios, and upon the proper
selection of projects to include in corporate capital budgets, have
increasingly engaged the attention of professional economists and other
students of the capital markets and of business finance in recent years (Lintner,
1965). The
stock market is an extremely complex system with various interacting components
(Sharma
& Banerjee, 2015). One of the most important
developments in modern capital theory is the capital asset pricing model (CAPM)
developed by Sharpe (1964) and Lintner (1965) in (Wakyiku,
2010). The CAPM predicts that equilibrium prices will be
set such that expected returns in excess of the riskfree rate will be
proportional to the covariance with aggregate risk (Bossaerts
& Plott, 2002). The Sharpe (1964) and Lintner
(1965) capital asset pricing model (CAPM) is the workhorse of finance for
estimating the cost of capital for project selection (Da,
Guo, & Jagannathan, 2012). The CAPM is one the underlying
building blocks of Modern Portfolio Theory and as such is constructed on a
number of strong theoretical assumptions concerning the behaviour of financial
markets and of investors (Boďa
& Kanderová, 2014). Also called as the standard capital
asset pricing model (CAPM), or the one-factor capital asset pricing model Sharpe–Lintner–Mossin
form of a general equilibrium relationship in the capital markets (Elton,
Gruber, Brown, & Goetzmann, 2014).
Research
motivation
The overall result is a chaotic
complex system which has so far proved very difficult to analyze and predict (Sharma
& Banerjee, 2015). The movement of stock prices
are somewhat interdependent as well as dependent on a wide multitude of
external stimuli like announcement of government policies, change in interest
rates, changes in political scenario, announcement of quarterly results by the
listed companies and many others (Sharma
& Banerjee, 2015).
Research
objective
1. To determine the volatility of
the market stock using simple CAPM.
2. Revisits empirical validity of
the linear functional form of the CAPM with respect to recent data.
3. To determine the inefficiency
of beta the measure of market risk
Implication
of study
More better understanding in the
volatility of the Malaysia market stock using random selected 10 stock index to
represent each sector in the market. The revisit the validity of CAPM in
Malaysia sector. The finding of this paper is the significant of beta imply the
CAPM was hold but with low R-square show a the inefficiency of beta the measure
of market risk (Wakyiku,
2010).
2.0
Literature review
The early researcher such as Sharpe
and Cooper (1972) examined whether following alternative strategies, with
respect to risk over long periods of time, would produce returns consistent
with modern capital theory (Elton
et al., 2014). The particular model consider
is the Ross (1976) single-factor linear beta pricing model based on the stock
index portfolio. We refer to this as the CAPM for convenience, following
convention (Da
et al., 2012).
Previous researcher like Wakyiku,
(2010) refer to Lutwama (2006), Atuhairwe and Tarinyeba
(2005), Katto and Tarinyeba (2004) and a few others, carry out non-econometric
discussions of these issues. Meanwhile, Bossaerts
& Plott, (2002) that prices do move towards the
CAPM, but very slowly. This
is because subjects generally have to trade combinations of securities in order
to improve their positions, yet in thin markets, it is difficult to implement
combined trades (Bossaerts
& Plott, 2002).
Early empirical study of the CAPM
performed by Lintner and reproduced in Douglas (1968) in (Elton
et al., 2014). Lintner first estimated beta
for each of the 301 common stocks in his sample. He estimated beta by
regressing each stock’s yearly return against the average return for all stocks
in the sample using data from 1954 to 1963 (Elton
et al., 2014). Judd and Guu (2000) for
theoretical evidence that the CAPM obtains when risk is small (Bossaerts
& Plott, 2002).
Miller and Scholes (1972) in a
classic article show that the anomalous results reported by Lintner may be an
artifact of a number of statistical issues, most notably that the beta measured
in the first-pass regression is only an estimate of the true beta (Elton
et al., 2014). Hence, Welch (2008) finds that
about 75.0% of finance professors recommend using the CAPM to estimate the cost
of capital for capital budgeting. A survey of chief financial officers by
Graham and Harvey (2001) indicates that 73.5% of the respondents use the CAPM (Da
et al., 2012).
Fama and French (1993) conjecture
that two additional risk factors beyond the stock market factor used in
empirical implementations of the CAPM are necessary to fully characterize
economy wide pervasive risk in stocks (Da
et al., 2012). The positive value of the
intercept that emerges is evidence in support of the two-factor model (Elton
et al., 2014).
Fama and French (1992) come to
similar conclusions using portfolios organized by size, book to market, as well
as beta and conclude that the relation between beta and aver- age return is
flat, even when beta is the only explanatory variable (Elton
et al., 2014). Roll and Ross (1994) in Elton
et al., (2014) argue that this is an artifact
of using ordinary least squares in the cross-sectional second-pass regression.
They argue that the relationship between average returns and beta is retrieved
once heteroskedasticity and cross-sectional residual correlation is accounted
for using generalized least squares instead of the more usual ordinary least
squares in the second-pass cross-sectional regression
3.0
Method
By follow method use by Boďa
& Kanderová, (2014), ten risky asset, stock,
included in the KLSE index were selected to participate in the study in a
random design. From each of the 10 different GICS sector represent in the KLSE
index one stock was picked random so as to enable a variety of stocks
stratified across various industries. The KLSE index is take as a proxy of the
market portfolio, and the riskless rate is proxied by the nominal 3 month
interest rate Malaysia government securities. The data was obtain using data
stream from 2012 until 2015 with weekly frequency enable the 157 observation in
the analysis.
|
GICS Sector Stock (&
Ticker)
|
Consumer Discretionary UMW HOLDINGS (UMWH)
|
Consumer Staples BRIT.AMER.TOB.(MALAYSIA) (BAT)
|
Energy PETRONAS DAGANGAN (PDAG)
|
Financials PUBLIC BANK (PBK)
|
Health Care BIOSIS GROUP (BISS)
|
|
GICS Sector Stock (&
Ticker)
|
Industrials GAMUDA (GAM)
|
Information Technology MESINIAGA (MESI)
|
Materials PETRONAS CHEMICALS GP. (PCHEM)
|
Telecommunication
Services DIGI.COM (DIGI)
|
Utilities YTL POWER INTERNATIONAL
(YTLP)
|
Certain hypotheses can be
formulated that should hold whether one believes in the simple CAPM or the
two-factor general equilibrium model (Elton
et al., 2014).
• The first is that higher risk
(beta) should be associated with a higher level of return.
• The second is that return is
linearly related to beta; that is, for every unit increase in beta, there is
the same increase in return.
• The third is that there should
be no added return for bearing nonmarket risk.
3.1
Model
Theoretical model
E (γi) = β E (γm) (1)
γi = Ri – Rf represents the
return on risky asset i in excess of the riskless rate, and;
γm = Rm – Rf signifies the excess
return on the market portfolio (often referred to as market premium or risk
premium) (Boďa
& Kanderová, 2014).
Econometric model
Where,
4.0
Result
A β of 1 indicates that the
security’s price will move with the market. A β of less than 1 means that the
security will be less volatile than the market (Sharma
& Banerjee, 2015). A β of greater than 1 indicates
that the security’s price will be more volatile than the market.
The estimated CAPM betas for
individual stocks
|
GICS Sector Stock (&
Ticker)
|
Consumer Discretionary UMW HOLDINGS (UMWH)
|
Consumer Staples BRIT.AMER.TOB.(MALAYSIA) (BAT)
|
Energy PETRONAS DAGANGAN (PDAG)
|
Financials PUBLIC BANK (PBK)
|
Health Care BIOSIS GROUP (BISS)
|
|
BETA
|
1.14907
|
0.971523
|
0.883707
|
0.691022
|
-0.267932
|
|
GICS Sector Stock (&
Ticker)
|
Industrials GAMUDA (GAM)
|
Information Technology MESINIAGA (MESI)
|
Materials PETRONAS CHEMICALS GP. (PCHEM)
|
Telecommunication
Services DIGI.COM (DIGI)
|
Utilities YTL POWER INTERNATIONAL
(YTLP)
|
|
BETA
|
1.141312
|
0.579111
|
1.106843
|
1.024636
|
1.124951
|
A β of greater than 1 indicates that the security’s price will be more volatile than the market. As an example, from table above we can see that the β of consumer discretionary sector is 1.14907 (15% more volatile than the market) while that of the financial sector is 0.691022 (30% less volatile than the market).
Heath care sector is -0.267932
indicates investment in the health sector tent to go down when market go up. Negative
betas are possible for investments that tend to go down when the market goes
up, and vice versa.
|
Stock beta estimates
obtained using monthly stock returns
|
|||||
|
Stock name
|
Estimated beta
|
t-value
|
Standar error
|
R-squared
|
other factor
|
|
UMWH
|
1.14907
|
7.377338
|
0.155757
|
0.259879
|
74.01%
|
|
BAT
|
0.971523
|
5.234624
|
0.185596
|
0.150225
|
84.98%
|
|
PETD
|
0.883707
|
4.109046
|
0.215064
|
0.09823
|
90.18%
|
|
PBK
|
0.691022
|
6.637867
|
0.104103
|
0.221345
|
77.87%
|
|
BISS
|
-0.267932
|
-0.320462
|
0.836079
|
0.000662
|
99.93%
|
|
GAM
|
1.141312
|
7.459311
|
0.153005
|
0.264152
|
73.58%
|
|
MESI
|
0.579111
|
3.147779
|
0.183974
|
0.060085
|
93.99%
|
|
PCHEM
|
1.106843
|
8.893994
|
0.124448
|
0.337899
|
66.21%
|
|
DIGI
|
1.024636
|
7.965626
|
0.128632
|
0.290459
|
70.95%
|
|
YTLP
|
1.124951
|
6.421298
|
0.175191
|
0.210123
|
78.99%
|
The estimated beta coefficients
range from 0.267932
to 1.14907. Nine out of the ten stocks’ beta coeffecients are positive and
statistically significant at the 5% level. CAPM predicts that higher risk is
associated with higher return, and therefore we expect positive beta estimates
for CAPM to hold, and yet one stocks have beta coeffecients that are not
statistically different from zero, even at the 10% level.
The R-squared values, as seen in
table above, are also generally very low. It is NONE the stocks that have their
variation in excess return fairly explained by the excess return on the market
index. This is equivalent to having betas that are fairly inefficient in
explaining the relation between market risk and return, for only three of the
nine stocks with positive statistically significant beta. The R-squared value
is the ratio of market risk to the sum of market and firm-specific risk, and as
such a low value points to the inefficiency of beta the measure of market risk (Wakyiku,
2010).
5.0 Conclusion
The movement of stock prices are
somewhat interdependent as well as dependent on a wide multitude of external
stimulation like announcement of government policies, change in interest rates,
changes in political scenario, announcement of quarterly results by the listed
companies and many others (Sharma & Banerjee, 2015). The view that the CAPM
provides a reasonable estimate of a project’s
cost of capital, provided that any embedded real options associated with the
project are evaluated separately for capital budgeting purposes. The CAPM beta
enable the measure of exposure of the stock toward the risk.
The movement of stock prices are
somewhat interdependent as well as dependent on a wide multitude of external
stimuli like announcement of government policies, change in interest rates,
changes in political scenario, announcement of quarterly results by the listed
companies and many others (Sharma & Banerjee, 2015). The finding support
the CAPM were the high risk imply with the high return does hold. The nine out
of ten stock were significant which is support the previous study done by (Bossaerts
& Plott, 2002) where positive value of the
intercept that emerges is evidence in support of the two-factor model (Elton
et al., 2014). However, the R-squared values are
also generally very low. It is none the stocks that have their variation in
excess return fairly explained by the excess return on the market index. This
is equivalent to having betas that are fairly inefficient in explaining the
relation between market risk and return, for only three of the nine stocks with
positive statistically significant beta. The R-squared value is the ratio of
market risk to the sum of market and firm-specific risk, and as such a low
value points to the inefficiency of beta the measure of market risk (Wakyiku,
2010). We can suggest that the beta inefficient measure
of market risk and the model such as FF 3 and 5 factor model (Fama
& French, 2015) model are available to the
future researcher.
Reference
Boďa, M.,
& Kanderová, M. (2014). Linearity of the Sharpe-Lintner version of the
Capital Asset Pricing Model, 110, 1136–1147.
doi:10.1016/j.sbspro.2013.12.960
Bossaerts, P., & Plott, C.
(2002). The {CAPM} in Thin Experimental Markets. Journal of Economic
Dynamics {&} Control, 26(7--8), 1093–1112.
Da, Z., Guo, R. J., &
Jagannathan, R. (2012). CAPM for estimating the cost of equity capital:
Interpreting the empirical evidence. Journal of Financial Economics, 103(1),
204–220. doi:10.1016/j.jfineco.2011.08.011
Elton, E. J., Gruber, M. J., Brown,
S. J., & Goetzmann, W. N. (2014). MODERN PORTFOLIO THEORY AND INVESTMENT
ANALYSIS (9th ed.).
Fama, E. F., & French, K. R.
(2015). A five-factor asset pricing model $, 116, 1–22.
Lintner, J. (1965). The valuation
of risk assets and the selection of risky investments in stock portfolios and
capital budgets. The Review of Economics and Statistics, 47(1),
13–37. doi:10.2307/1924119
Sharma, C., & Banerjee, K.
(2015). A Study of Correlations in the Stock Market, (April 2006), 1–12.
Wakyiku, D. (2010). Testing the
Capital Asset Pricing Model (CAPM) on the Uganda Stock Exchange. arXiv
Preprint arXiv:1101.0184, (1964). Retrieved from
http://arxiv.org/abs/1101.0184
.
APPENDIX
UMWH
Dependent
Variable: UMWH
|
||||
Method:
Least Squares
|
||||
Date:
05/31/15 Time: 02:46
|
||||
Sample:
3/16/2012 3/13/2015
|
||||
Included
observations: 157
|
||||
Variable
|
Coefficient
|
Std.
Error
|
t-Statistic
|
Prob.
|
FBMKLCI
|
1.149070
|
0.155757
|
7.377338
|
0.0000
|
C
|
0.002320
|
0.001811
|
1.280922
|
0.2021
|
R-squared
|
0.259879
|
Mean dependent var
|
0.000313
|
|
Adjusted
R-squared
|
0.255104
|
S.D. dependent var
|
0.025992
|
|
S.E. of
regression
|
0.022433
|
Akaike info criterion
|
-4.743939
|
|
Sum
squared resid
|
0.078000
|
Schwarz criterion
|
-4.705006
|
|
Log
likelihood
|
374.3992
|
Hannan-Quinn criter.
|
-4.728127
|
|
F-statistic
|
54.42511
|
Durbin-Watson stat
|
1.838461
|
|
Prob(F-statistic)
|
0.000000
|
|||
BAT
Dependent
Variable: BAT
|
||||
Method:
Least Squares
|
||||
Date:
05/31/15 Time: 02:53
|
||||
Sample:
3/16/2012 3/13/2015
|
||||
Included
observations: 157
|
||||
Variable
|
Coefficient
|
Std.
Error
|
t-Statistic
|
Prob.
|
FBMKLCI
|
0.971523
|
0.185596
|
5.234624
|
0.0000
|
C
|
0.001189
|
0.002158
|
0.551083
|
0.5824
|
R-squared
|
0.150225
|
Mean dependent var
|
-0.000508
|
|
Adjusted
R-squared
|
0.144743
|
S.D. dependent var
|
0.028904
|
|
S.E. of
regression
|
0.026730
|
Akaike info criterion
|
-4.393388
|
|
Sum
squared resid
|
0.110748
|
Schwarz criterion
|
-4.354455
|
|
Log
likelihood
|
346.8810
|
Hannan-Quinn criter.
|
-4.377576
|
|
F-statistic
|
27.40129
|
Durbin-Watson stat
|
2.580097
|
|
Prob(F-statistic)
|
0.000001
|
|||
PETD
Dependent
Variable: PETD
|
||||
Method:
Least Squares
|
||||
Date: 05/31/15 Time: 02:54
|
||||
Sample:
3/16/2012 3/13/2015
|
||||
Included
observations: 157
|
||||
Variable
|
Coefficient
|
Std.
Error
|
t-Statistic
|
Prob.
|
FBMKLCI
|
0.883707
|
0.215064
|
4.109046
|
0.0001
|
C
|
-0.000464
|
0.002500
|
-0.185417
|
0.8531
|
R-squared
|
0.098230
|
Mean dependent var
|
-0.002007
|
|
Adjusted
R-squared
|
0.092413
|
S.D. dependent var
|
0.032513
|
|
S.E. of
regression
|
0.030974
|
Akaike info criterion
|
-4.098660
|
|
Sum
squared resid
|
0.148708
|
Schwarz criterion
|
-4.059727
|
|
Log
likelihood
|
323.7448
|
Hannan-Quinn criter.
|
-4.082848
|
|
F-statistic
|
16.88426
|
Durbin-Watson stat
|
2.199891
|
|
Prob(F-statistic)
|
0.000064
|
|||
PBK
Dependent
Variable: PBK
|
||||
Method:
Least Squares
|
||||
Date:
05/31/15 Time: 03:00
|
||||
Sample:
3/16/2012 3/13/2015
|
||||
Included
observations: 157
|
||||
Variable
|
Coefficient
|
Std.
Error
|
t-Statistic
|
Prob.
|
FBMKLCI
|
0.691022
|
0.104103
|
6.637867
|
0.0000
|
C
|
0.000804
|
0.001210
|
0.664545
|
0.5073
|
R-squared
|
0.221345
|
Mean dependent var
|
-0.000403
|
|
Adjusted
R-squared
|
0.216322
|
S.D. dependent var
|
0.016937
|
|
S.E. of
regression
|
0.014993
|
Akaike info criterion
|
-5.549767
|
|
Sum
squared resid
|
0.034844
|
Schwarz criterion
|
-5.510834
|
|
Log
likelihood
|
437.6567
|
Hannan-Quinn criter.
|
-5.533955
|
|
F-statistic
|
44.06128
|
Durbin-Watson stat
|
2.466226
|
|
Prob(F-statistic)
|
0.000000
|
|||
BISS
Dependent
Variable: BISS
|
||||
Method:
Least Squares
|
||||
Date:
05/31/15 Time: 03:02
|
||||
Sample:
3/16/2012 3/13/2015
|
||||
Included
observations: 157
|
||||
Variable
|
Coefficient
|
Std.
Error
|
t-Statistic
|
Prob.
|
FBMKLCI
|
-0.267932
|
0.836079
|
-0.320462
|
0.7490
|
C
|
-0.001961
|
0.009720
|
-0.201732
|
0.8404
|
R-squared
|
0.000662
|
Mean dependent var
|
-0.001493
|
|
Adjusted
R-squared
|
-0.005785
|
S.D. dependent var
|
0.120069
|
|
S.E. of
regression
|
0.120415
|
Akaike info criterion
|
-1.383082
|
|
Sum
squared resid
|
2.247478
|
Schwarz criterion
|
-1.344149
|
|
Log
likelihood
|
110.5720
|
Hannan-Quinn criter.
|
-1.367270
|
|
F-statistic
|
0.102696
|
Durbin-Watson stat
|
2.406164
|
|
Prob(F-statistic)
|
0.749050
|
|||
GAM
Dependent
Variable: GAM
|
||||
Method:
Least Squares
|
||||
Date:
05/31/15 Time: 03:38
|
||||
Sample:
3/16/2012 3/13/2015
|
||||
Included
observations: 157
|
||||
Variable
|
Coefficient
|
Std.
Error
|
t-Statistic
|
Prob.
|
FBMKLCI
|
1.141312
|
0.153005
|
7.459311
|
0.0000
|
C
|
0.001900
|
0.001779
|
1.068259
|
0.2871
|
R-squared
|
0.264152
|
Mean dependent var
|
-9.30E-05
|
|
Adjusted
R-squared
|
0.259405
|
S.D. dependent var
|
0.025606
|
|
S.E. of regression
|
0.022036
|
Akaike info criterion
|
-4.779587
|
|
Sum
squared resid
|
0.075268
|
Schwarz criterion
|
-4.740654
|
|
Log
likelihood
|
377.1976
|
Hannan-Quinn criter.
|
-4.763775
|
|
F-statistic
|
55.64132
|
Durbin-Watson stat
|
2.254536
|
|
Prob(F-statistic)
|
0.000000
|
|||
MESI
Dependent
Variable: MESI
|
||||
Method:
Least Squares
|
||||
Date:
05/31/15 Time: 03:40
|
||||
Sample:
3/16/2012 3/13/2015
|
||||
Included
observations: 157
|
||||
Variable
|
Coefficient
|
Std.
Error
|
t-Statistic
|
Prob.
|
FBMKLCI
|
0.579111
|
0.183974
|
3.147779
|
0.0020
|
C
|
-0.003794
|
0.002139
|
-1.773627
|
0.0781
|
R-squared
|
0.060085
|
Mean dependent var
|
-0.004805
|
|
Adjusted
R-squared
|
0.054021
|
S.D. dependent var
|
0.027243
|
|
S.E. of
regression
|
0.026497
|
Akaike info criterion
|
-4.410935
|
|
Sum
squared resid
|
0.108822
|
Schwarz criterion
|
-4.372002
|
|
Log
likelihood
|
348.2584
|
Hannan-Quinn criter.
|
-4.395123
|
|
F-statistic
|
9.908514
|
Durbin-Watson stat
|
2.371045
|
|
Prob(F-statistic)
|
0.001974
|
|||
PCHEM
Dependent
Variable: PCHEM
|
||||
Method:
Least Squares
|
||||
Date:
05/31/15 Time: 03:41
|
||||
Sample:
3/16/2012 3/13/2015
|
||||
Included
observations: 157
|
||||
Variable
|
Coefficient
|
Std.
Error
|
t-Statistic
|
Prob.
|
FBMKLCI
|
1.106843
|
0.124448
|
8.893994
|
0.0000
|
C
|
-0.002331
|
0.001447
|
-1.611222
|
0.1092
|
R-squared
|
0.337899
|
Mean dependent var
|
-0.004264
|
|
Adjusted
R-squared
|
0.333627
|
S.D. dependent var
|
0.021957
|
|
S.E. of
regression
|
0.017924
|
Akaike info criterion
|
-5.192747
|
|
Sum
squared resid
|
0.049794
|
Schwarz criterion
|
-5.153814
|
|
Log
likelihood
|
409.6306
|
Hannan-Quinn criter.
|
-5.176935
|
|
F-statistic
|
79.10313
|
Durbin-Watson stat
|
2.337326
|
|
Prob(F-statistic)
|
0.000000
|
|||
DIGI
Dependent
Variable: DIGI
|
||||
Method:
Least Squares
|
||||
Date:
05/31/15 Time: 03:42
|
||||
Sample:
3/16/2012 3/13/2015
|
||||
Included
observations: 157
|
||||
Variable
|
Coefficient
|
Std.
Error
|
t-Statistic
|
Prob.
|
FBMKLCI
|
1.024636
|
0.128632
|
7.965626
|
0.0000
|
C
|
0.002222
|
0.001496
|
1.485782
|
0.1394
|
R-squared
|
0.290459
|
Mean dependent var
|
0.000432
|
|
Adjusted
R-squared
|
0.285882
|
S.D. dependent var
|
0.021923
|
|
S.E. of
regression
|
0.018526
|
Akaike info criterion
|
-5.126615
|
|
Sum
squared resid
|
0.053199
|
Schwarz criterion
|
-5.087682
|
|
Log
likelihood
|
404.4393
|
Hannan-Quinn criter.
|
-5.110803
|
|
F-statistic
|
63.45119
|
Durbin-Watson stat
|
1.895091
|
|
Prob(F-statistic)
|
0.000000
|
|||
YTLP
Dependent
Variable: YTLP
|
||||
Method:
Least Squares
|
||||
Date:
05/31/15 Time: 03:44
|
||||
Sample:
3/16/2012 3/13/2015
|
||||
Included
observations: 157
|
||||
Variable
|
Coefficient
|
Std.
Error
|
t-Statistic
|
Prob.
|
FBMKLCI
|
1.124951
|
0.175191
|
6.421298
|
0.0000
|
C
|
-0.001207
|
0.002037
|
-0.592715
|
0.5542
|
R-squared
|
0.210123
|
Mean dependent var
|
-0.003172
|
|
Adjusted
R-squared
|
0.205027
|
S.D. dependent var
|
0.028299
|
|
S.E. of
regression
|
0.025232
|
Akaike info criterion
|
-4.508780
|
|
Sum
squared resid
|
0.098678
|
Schwarz criterion
|
-4.469847
|
|
Log
likelihood
|
355.9392
|
Hannan-Quinn criter.
|
-4.492968
|
|
F-statistic
|
41.23307
|
Durbin-Watson stat
|
2.103193
|
|
Prob(F-statistic)
|
0.000000
|
|||
No comments:
Post a Comment