Estimation of Discount

Rates in Emerging Markets

 Learning Objectives
 To know the main features of Latin American
emerging markets for the valuation of investment
projects
 To understand the implications for the valuation of
real investments in Latin America.
 To understand the different cases for the estimation
of discount rates in emerging capital markets
 To know the different models to estimate discount
rates for well-diversified investors (WDI), non-well
diversified investors (NWDI) and non-diversified
investors (NDI).
 Content
• What is an emerging economy?
• What is an emerging capital market?
• Types of investors in emerging markets
• Features of Latin American emerging markets
• Implications for project valuation
• Estimation of discount rates in emerging markets
• Cases for the estimation of discount rates in emerging
markets
• Discount rates for Well-Diversified Investors (WDI)
• Discount rates for Non-well-Diversified Investors (NWDI).
• Discount rates for Non-Diversified Investors (NDI)
• Comparison between the different models
• Case Trinity
What is an emerging economy?
• In 1981, Antoine W. van Agtmael from the IFC
(International Finance Corporation) used for the
first time the label of emerging country and
emerging market.

• According to the IFC (a branch of the World

Bank), an emerging country is the one with a
gross national income (GNI) per capita below the
average of all countries. The threshold varies
through time, but it was around 11,500 USD in
2019
What is an emerging economy?
These countries share the following features:

•Low-to-mid per capita income.

•The brisk pace of economic growth.
•Commodity and currency swings.
•High market volatility—possibly caused by natural
disasters, external price shocks, or domestic policy
instability.
•Huge growth potential.
What is an emerging market (EM)?
• In the world there are 195 countries, from them 81
are developed countries (high income countries) and
114 developing countries according to the World
Bank)

• As April 2019 there are 70 members of the World

Federation of Exchanges (WFE)

• There are many classifications of stock markets,

among them the most important ones are the S&P
and MSCI
S&P Classification, 2018
MSCI Classification, 2020
MSCI Market Classification
 Types of investors in
emerging markets
• Global well-diversified investors.
• Non-well diversified local institutional investors
• Non-diversified local entrepreneurs.
Features of Latin American
Emerging Markets
• Distributions of stock and bond returns
• Availability and quality of historical data
• Degree of market integration
• Degree of market completeness
• Degree of investors’ diversification
Distribution of monthly stock
returns Period: 1993-2005
Statistic (%) Argentina Brazil Colombia Chile Mexico Peru Venezuela
Mean 0.41% 1.32% 0.84% 0.53% 0.62% 0.91% -0.03%
Mode 2.02% 3.23% 3.60% 4.10% 6.00% 0.22% 0.74%
Median 1.22% 2.45% 1.47% 0.32% 2.01% 1.18% -0.24%
Maximum 42.47% 31.12% 26.48% 18.28% 17.42% 30.44% 48.04%
Minimum -38.43% -49.44% -27.59% -34.40% -41.95% -40.98% -63.77%
Variance 125.83 147.16 87.53 51.11 94.15 80.16 214.02
Skewness -0.46 -1.14 -0.33 -1.09 -1.87 -0.82 -1.06
Kurtosis 4.81 5.31 3.65 5.84 6.44 6.17 6.67

Source: Mongrut S. (2007) Valoración de proyectos de inversión en

economías emergentes latinoamericanas: el caso de los inversionistas no
diversificados. Tesis Doctoral inédita. Universidad del Barcelona .
 Availability and quality of the
historical market information
• Availability: There is scarce data about expectations
from investors. Nonetheless, there are different
providers for historical market information (EMDB,
MSCI, Economática, Bloomberg, Reuters) their offers
span a short time-period.

• Relevance: The relevant data only belongs to the post-

liberalization period (1990 - ...)

• Quality: Historical data is highly volatile and lacks

liquidity.
Degree of market integration

• In integrated capital markets financial assets with

equal risk must command the same expected
returns regardless of their domicile.

• Indicators of integration: Openness ratio,

financial services ratio, transaction costs, number of
American Depositary Receipts (ADRs), ownership
concentration, home-country bias and country risk.
 Ownership concentration to the first
stockholder. Period: 1995-2005
70%

60%

50%

40%

30%

20%

10%

0%

Trimestres

Brasil Chile Colombia Perú Venezuela

Degree of market completeness
• A capital market is complete when it is possible to
replicate the firm or project’s risk in every future state of
nature and at any moment by using a portfolio of traded
and liquid financial assets.

• A liquid stock is the one with a trading frequency equal or

higher than 75% where the trading frequency is estimated
as the number of trading days with the stock divided by
the total number of trading days in the stock market
during the year.

• Indicators: ratio of liquid stocks divided by the total

number of stocks traded during the year and restrictions
to trade with financial assets.
 Liquidity in Latin American
Stock Markets
Year Argentina Brazil Chile Colombia México Perú Venezuela
1995 49% 22% 42% 23% 44% 30% 35%
1996 55% 24% 41% 18% 52% 29% 44%
1997 60% 27% 36% 22% 55% 26% 56%
1998 53% 20% 28% 20% 45% 23% 42%
1999 47% 29% 33% 16% 40% 21% 27%
2000 37% 29% 30% 7% 38% 15% 28%
2001 32% 27% 28% 12% 36% 11% 20%
2002 43% 27% 24% 28% 41% 12% 13%
2003 58% 30% 28% 40% 39% 18% 18%
2004 61% 32% 31% 43% 41% 16% 21%
2005 59% 33% 32% 44% 42% 18% 21%
Average 49% 27% 32% 22% 43% 19% 27%
Degree of investors’ diversification
• The main institutional investors in Latin American
stock exchanges are: Pension Funds, Mutual Funds,
Insurance Companies and local banks. Most
importantly, each one of them holds a non-well
diversified investment portfolio.
• From these, Pension Funds are the most important
institutional investors Latin American stock exchanges.
• Outside stock exchanges, non-diversified
entrepreneurs are the most important investors in
Latin American emerging countries.
• Indicators: Number of firms by size, proportion of
local financial assets in the investment portfolio and
foreign investment limits.
 Number of Firms as 2011
Eme rging countrie s
Country Micro % Small % Medium % Large % Total
Argentina 421,014 70% 138,218 23% 33,420 6% 11,051 2% 603,703
Brazil 4,518,832 88% 518,627 10% 71,617 1% 20,129 0% 5,129,205
Chile 627,998 76% 161,225 20% 23,994 3% 12,149 1% 825,366
Philippines 743,250 91% 70,222 9% 3,287 0% 3,496 0% 820,255
Poland 1,452,022 95% 53,104 3% 15,278 1% 3,014 0% 1,523,418
Romania 356,008 87% 43,241 11% 8,161 2% 1,541 0% 408,951
Rwanda 114,329 93% 8,548 7% 513 0% 106 0% 123,496
Serbia 72,995 86% 8,984 11% 2,103 2% 488 1% 84,570
Slovenia 110,343 94% 5,739 5% 1,174 1% 225 0% 117,481
Turkey 2,522,011 97% 45,900 2% 19,408 1% 3,763 0% 2,591,082
Ukraine 295,815 44% 354,283 53% 20,753 3% 659 0% 671,510
Average 83.7% 13.8% 1.9% 0.5%
Developped countries
Country Micro % Small % Medium % Large % Total
Spain 2,103,390 94% 120,940 5% 15,484 1% 2,728 0% 2,242,542
France 2,013,398 94% 100,875 5% 18,417 1% 4,048 0% 2,136,738
United Kingdom 1,518,211 89% 147,024 9% 25,495 2% 5,858 0% 1,696,588
Netherlands 759,086 94% 42,154 5% 8,418 1% 1,497 0% 811,155
Norway 249,704 92% 19,171 7% 2,808 1% 581 0% 272,264
New Zealand 421,823 90% 19,571 4% 15,980 3% 12,972 3% 470,346
Portugal 789,947 95% 35,681 4% 5226 1% 801 0% 831,655
Sweden 616,132 95% 29,265 4% 5,027 1% 994 0% 651,418
United States 22,110,600 79% 5,160,400 19% 556,900 2% 17200 0% 27,845,100
Average 91% 7% 1.3% 0.5%

Source: World Bank. Own elaboration. Company size is defined according to the
number of employees
 Estimation of discount
rates in emerging markets
Discount Rate

Degree of diversification
0 Max
Total Risk Systematic Risk
Cases for the estimation of discount
rates in emerging markets

Type of investor / degree of integration Segmented Partially integrated Integrated

Global WDI Do not operate World CAPM with CR World CAPM
Local NWDI Model with relative volatility and CR
Local NDI Model with a required return based upon total risk
Estimation of discount rates
Global WDI
• What matters is to estimate the value of the project
as it were traded (quoted) in the capital market.
• The key variable is the degree of integration of the
EM (country risk is being used as the degree of
integration).
• Three models: total segmentation (Local CAPM),
total integration (World CAPM) and models of partial
integration.
• Models of partial integration: Mariscal and Lee
(1993), Lessard (1996), Damodaran (1999), Estrada
(2002) and Bodnar, Dumas and Marston (2003).
 Model of total segmentation
Local CAPM
 Local CAPM (Sharpe, 1964)

E K SL  L
 Kf L
 i L
KM L
 Kf 
K Lf : Local risk free rate
Li : Security’s beta with respect to the local
market index
Model of complete integration
Global or World CAPM
 World CAPM (Solnik, 1974)

E K SL   K fw  iw K m
w

 K fw 
K fw  K US 
KUS  K US

f W
KM  K fW  
M
w
us
f

w
K
: Global
f risk free rate
w
:Security’s
i beta with respect to the global market index
 Models of partial integration
 Mariscal and Lee or Goldman Sachs’s Model (1993):

E K SL   K US
f K L
B  K US
B 
 US
i KS&P
m  K US
f 
: Sovereign yield spread

:KRisk
L
B  
K Premium estimated with the Standard and Poor's
US
B
market index
 
K MS&P  K US
f
Models of parcial integration
 Lessard’s Model (1996)

EK SL   K US
f  β US
i K US
m  K US
f 
Where :
β US
i  β L US
i βL

: Security’s beta with respect to the local market

Li : Local market’s beta with respect to the US market
US
L
 Models of parcial integration
 Estrada’s D-CAPM Model (2002)

E K SL   K US D US
 US
f  i K M  K f 
Di 

     
E Min  K i  K ,0  Min  K M
US
 KMUS
 
,0 

2



 US US
E Min K M  K M ,0  
 

: US risk free rate

K US
f
: Downside riskD beta
i
: US market risk premium
K US
M  KUS
f 
Models of parcial integration
 Bodnar, Dumas y Marston’s Model or Hybrid
Model (2003)

E K SL   K US
f  L
i K 
L
M  K L
f  US
i K M
US
 K US
f 


: Local

K LM  market
K Lf risk premium
KUS
M  Kf 
: US US market risk premium
 Damodaran’s assumption
Local stock and bond markets are perfect substitutes:
𝐾𝑀𝐿 − 𝐾𝐹𝐿 𝐾𝐵𝐿 − 𝐾𝐹𝐿
=
𝜎𝑀𝐿 𝜎𝐵𝐿
𝐾𝑀𝐿 − 𝐾𝐹𝑈𝑆 𝐾𝐵𝐿 − 𝐾𝐹𝑈𝑆
=
𝜎𝑀𝐿 𝜎𝐵𝐿
𝜎𝐿
𝑈𝑆 ሻ 𝑀
ሺ𝐾𝑀𝐿 − 𝐾𝐹𝑈𝑆 ሻ = ሺ𝐾𝐵𝐿 − 𝐾𝐹 ቆ 𝐿 ቇ
𝜎𝐵
𝜎𝑀𝐿
ሺ𝐾𝑀𝐿 − 𝐾𝐹𝐿 ሻ= ሺEMBIሻቆ 𝐿 ቇ
𝜎𝐵
Finally plugging this into Bodnar et.al (2003) model:
𝜎𝑀𝐿
𝐾𝑖𝐿 = 𝐾𝐹𝑈𝑆 + 𝛽𝑖𝐿,𝑈𝑆 ሺ𝐾𝑀𝑈𝑆 − 𝐾𝐹𝑈𝑆 ሻ+ 𝜆𝑖 ሺEMBIሻቆ 𝐿 ቇ
𝜎𝐵
𝑋
Lambda: 𝜆𝑖 = 1 −
𝑆

𝐾𝑖 = 𝛼𝑖+ 𝜆𝑖 ሺ𝑌𝑇𝑀𝑖 ሻ+ 𝑒𝑖
 Models of parcial integration
 Damodaran’s Model (2002, 2003)

 L  LM  
E K SL   K US
f  US
i K US
M  K US
f  
  i  K B  K US
 B   L  
  B  

: Sovereign
US yield spread
K K 
L
B B
: Exposition of security’s “i” to the country risk
i
 Discount rates for NWDI
• Again, what matters is to estimate de value of the project as it were traded (quoted) in
the capital market.
• In EM the estimation of betas still contain specific risk for each asset due to the lack of
market transparency.
• Hence, one must consider that the estimation of betas in fact also involve specific risk.
• Damodaran (2002) suggested the following adjusment for betas to include this fact (total
beta):

• Models: Godfrey and Espinosa (1996) and Estrada (2000).

σi σ β
βi  ρ i,M  i  i  β iT
σM σ M ρ i,M
 Model of parcial integration
 Godfrey and Espinosa’s Model (1996)

 i 
E K SL   K US
f  
K BL  K BUS 
K US
m  K US
f  0.6   US 
 

: Relative volatility with respect to the US market

i
US
 Model of total downside risk
 Estrada’s Model- Relative volatility (2000)

iSD
E K SL   K US
f  K US
M K US
f  MSD

SD

: Ratio of
i downside volatility
SD
M
 Discount rates for NDI
• What it matters is to estimate the value of the project
according to the total risk assumed by the investor.
Then, the estimated value will not be a market value,
but a subjective (consistent and unbiased) value.
• The discount rate in this case becomes a required
return
• The variable of interest is the degree of investor’s
diversification and the degree of emerging market’s
integration (country risk).
• Model: Erb, Harvey and Viskanta (1996).
 Model of total credit risk
 Erb, Harvey y Viskanta’s Model (1996)

EK SL,t 1  K L
f  α  βLnCCR t 1 

:Ln CCR Logarithm of the Country Credit Rating

Natural