Required
return on equity, or R, is an important building blocks for stock’s valuation.
It is used in various valuation model i.e. DCF and other relative value method
(i.e. P/E and P/B method).
Due to its
importance, I have been searching for methods that produces a reliable result
and is practical in term of availability of input data and simplicity of the
model. After review of the CFA curriculum and online materials, I came across
works done by Aswath Damodaran, a Professor of Finance at the Stern School of
Business at New York who is teaching valuation and corporate finance in MBA
programs. He was named by Business Week, in 2011 as the most popular business
school professor in US. In his “Equity Risk Premiums(ERP): Determinants,
Estimation and Implication (2013)”, he presented a number of methods to compute
the ERP, which is one of the inputs for Capital Asset Pricing Model (CAPM), a
popular model to determine the R. Besides his work on ERP, his website also
provided an extensive range of data to facilitate the computation of ERP.
Before
detailing my selected method, I will give a list of general choices in
determining R as below.
A. CAPM
Required return on stock i, Ri = Risk-free
return, Rf + βI (ERP)
B. Fama-French
Model (Multifactor Models)
Ri = Rf + βmarket RMRF + βsize SMB
+ βvalueHML
RMRF: Rm - Rf
(Return on market (i.e. FBMKLCI) – Risk-free return)
SMB: Small minus Big
(Return of SmallCap Portfolio – LargeCap)
HML: High minus Low
(High Book/Price – Low)
C. Build-up
Method
Ri = Rf + ERP ± Risk
Premia1 ± Risk Premia2 ± ……. ± Risk Premian
I will use
the CAPM to compute R because
(i)
it
requires fewer inputs compare to the method B and C
(ii)
input
data is more available (compare to, say, B that require return on SmallCap portfolio like FBMSmallCap, return data on FBMKLCI is more easily available)
(iii)
Damodaran’s
site provide updated data for the computation of ERP, which is less work for me
Hence,
Ri = Rf + βI x
ERP
where:
Ri =
Required rate of return of stock i
Rf - Risk-free rate of return. Represented
by Malaysia 5-Years or 10-Year Govt
Bonds Yield depending my selected time-horizon
ERP =Equity
Risk Premium for investing in the market(FBMKLCI), relative to the risk-free
rate
Βi – Beta of stock i against market. To obtain via
regression analysis of FBMKLCI and stock’s return data
Method for
the inputs are described above but ERP’s is following.
In his
work, Damodaran explained that ERP for a country’s market can be obtained by,
first, calculate the ERP of a matured market like US as the base. Then, a Country
Risk Premium (RP) is added to the base for the additional risk.
ERP = Base ERP for mature market + Country RP
Based ERP
is obtained by a method called Implied ERP. On a brief note, the Implied ERP is
obtained by solving the R(required return) for the latest S&P500 year-end
closing given an estimated future dividends growth. I will use the data
provided by his site as I believe his work is reliable and the computation (if I
choose to compute myself) is too data-intensive for me.
(Other that
the Implied ERP, one can also use the historical ERP. It is can be obtained by subtracting
Average Return of S&P500 with Average Return of T.Bond over a certain
period of time. However, this method is backward-looking. A forward-looking
method like the Implied ERP is more relevant)
For the
Country RP, I will use the Melded CDS approach (describe in his work).
Country RP = Malaysia CDS Spread x Relative Standard Deviation (Equity Market/Bond Market)
The first
term of the formula is the 5 years Credit Default Swap(CDS) rate of Malaysia
(or 10-years, depending on my selection). It is the required rate of return for
investors to provide insurance against the risk of sovereign debt default and
the latest rate can be obtained from the Deutsche Bank Research website. The
higher the CDS, the higher the probability debt default by Malaysia’s
government. As the CDS only represent the risk of sovereign’s bond default, we
need to covert the bond’s risk premium to equity risk premium. Hence the
second-term is used, which is the relative standard deviation (or risk) of the
equity market(FBMKLCI) vs bond market. In this case, I will used the data
provided by Damodaran, which are based on 2 years weekly data.
(Alternatively,
the CDS can be replaced by Moody’s Rating-based default spread according to
Malaysia’s sovereign rating. It is more stable and do not change frequently
compared to CDS, which is determined by the market and is fluctuating daily. As
CDS is usually higher than default spread and it imply all the market’s
expectation, I choose CDS as a conservative approach and most relevant to the
current situation. )
In
conclusion,
Ri = Rf + βI x [Implied ERPUS + (CDS SpreadMalaysia x Relative S.D.)]
ERP of US Market (1-Jan-15)
|
5.75%
|
Malaysia 5-years CDS (22-Jan-15)
|
1.3115%
|
Relative σ (2 years weekly data)
|
1.27
|
ERP
|
7.416% (5.75% + 1.27 x 1.3115%)
|
Malaysia 5Y Bond Yield (22-Jan-15)
|
3.811%
|
Hence, Ri
= 3.811% + βI x 7.416%
For
example, the R of FBMKLCI (β = 1.0) is 11.227%.
The
application of the R, for example, is as below. We can extract market
expectation on the current FBMKLCI level, based on Gordon growth model
P0 – 1803.08 (23-Jan-15)
E0
– 106.37 (estimate)
D0
– 59.354
R
– 0.11227
Solving the formula, I get g =
0.0768 (7.68%). The market is expecting FBMKLCI to have a constant growth rate
of 7.68% (if based on Gordon Growth Model).
Reference: http://pages.stern.nyu.edu/~adamodar/
Reference: http://pages.stern.nyu.edu/~adamodar/