Emerging Financial Market
6. Measuring Political Risk
Prof. J.P. Mei
Emerging markets in the 1900s
 World
Trade accounted for huge part of
GDP in many countries.
 Equity markets at the turn of the
century flourished, many markets
established.
 Internet Technology (Railroad) &
Worldwide Real Time Communication
 Foreign Investment surged (US$44
Billion in 1913 dollars)
1
Table 1. The Late 19th Century Trade Boom
Country
1870
1890
1913
44.3
33.0
11.8
31.3
25.6
13.0
42.0
34.0
11.1
50.7
87.4
32.0
34.0
35.7
57.8
54.6
149.9
47.2
43.6
48.0
81.9
59.6
249.4
42.3
50.9
61.4
134.9
na
11.7
11.3
23.7
18.3
31.7
25.0
11.1
28.3
19.3
39.8
23.9
16.1
30.9
28.7
New World
Australia
Canada
United States
Old World: Free Trade
United kingdom
The Netherlands
Sweden
Norway
Demark
Belgium
Old World: Protected
Germany
Spain
Portugal
France
Italy
Source: Williamson (NBER)
Figure1. British Overseas Investment 1854-1914 (US$ billion)
$25
$20
$15
$10
$5
$0
1854
1872
1881
1895
1905
1910
1914
Data Source: Albert Kimber, Foreign Government Securities,
1919,
A. W. Kimber & Company.
2
Table 2. Main Creditor and Debtor Countries, 1913
Source: United Nations (1949)
Major sources of Political
Risk in the past
 Two
Major Exploitations: within and across
Countries (Slavery and Child Labor) Caused
Strong Resentment.
 Communism and the Risk of Nationalization
 Colonialism and the Risk of Political Upheaval.
(Then superpower was the largest government
supported drug dealer in the world)
 World WAR I and the Russian Revolution ended
the first wave of globalization.
 Long-term Return of Emerging markets (not
glamorous due to submerged markets)
Figure 2: British Sales of Opium to China (Thousand Chests)
Source: Mark Borthwick, Pacific Century, Westview Press, 1992
25
20
15
10
5
0
1729
1790
1819
1823
1832
Cultural Clash between the Modern and Ancient
Measuring Risks
 Measurement
of political risk
 Measuring
corruption
 Measuring
the rule of Law
 Political
risk measurements can be used
in project financing. (discount rates)
 Measuring
political risks is still an art
rather than a science.
2
Political Risk Insurance
 Eligibility
& Coverage
 OPIC insurance can cover the following
three political risks: currency inconvertibility, expropriation, political violence.
 OPIC insures Business income and
assets.
 Election of Coverage & Premium Base
Rates
 Problem: Lack a systematic approach
3
Political Uncertainty and Elections
Election cycle
 a) the time leading up to an election and
the time of government transition after the
election, and
 b) the time after the transition is complete
and the next election season starts.
 In a democratic system, the election
process is a major political event for
determining future political course of a
country.
4
Why Political Risk Matters

1. The "first generation" currency crisis model represented by
Krugman (1979) and Flood and Garber (1984): Strong incentive to
engage in inconsistent policies during elections by pursuing
expansionary monetary and fiscal policies while holding exchange
rates fixed to ensure price stability or other policy objectives.

2. The "second generation" model of Obstfeld (1994). In such a
model, the cost of defending the currency increases when people
suspect that the government is leaning towards abandoning the fixed
rate. (Banking problems)

3. Self-fulfilling exchange rate crises (see, Banerjee (1992)).

4. Contingent investment or "real options": foreign capital flow to
Asia from a huge $93 billion inflow in 1996 to a $12 billion net
outflow in 1997.
5
Dependent Variables
 Financial
crisis: defined as a sharp shift
from inflow to outflow between year t-1
and t
Turkey and Venezuela in 1994, Argentina and
Mexico in 1995; and Indonesia, Korea, Malaysia,
the Philippines, and Thailand in 1997.
 78 observations (22 x 4 - 10 excluded observations)

 equity
returns and market volatility: the
IFC index.
6
Economic and Financial Variables:
the ratio of short-term debt to the foreign
exchange reserves
 total debt outstanding (long and short term)
 the change in the ratio of the financial claims on
the private sector relative to GDP over the
preceding three years.
 current account to GDP ratio
 capital flow to GDP ratio
 the percentage change in the real exchange rate
(RER) in the previous three years.
 index of corruption
7
 and Regional Market Contagion Dummy

 Table
1: eight out of nine financial crises
happened within one year before or after the
election.
 Table 2 presents some summary
financial crisis: 23% in political years vs 2% in
non-political years
a significant difference in market volatility in
political years
high correlation between the political dummy
and financial crisis.
negatively correlated with changes in currency
8
value.
Table 1: Summary of Financial Crisis Year, Election Date,
Political System, and Election Cycle
Country
Crisis Year Election Date Presidential or
Election
Parliamental
Cycle
System*
(Years)
May-95
1
4
Brazil
Nov-94
1
4
Chile
Dec-93
1
6
Colombia
May-94
1
4
Hungary
May-94
0
4
India
Apr-96
0
5
1997
Mar-98
1
5
Korea
1997
Dec-97
1
5
Malaysia
1997
Apr-95
0
4
Mexico
1995
Aug-94
1
6
Apr-95
1
5
Argentina
Indonesia
1995
Jordan
Monarch
Peru
Philippines
May-98
1
6
Poland
Nov-95
1
5
Russia
Jul-96
1
4
South Africa
May-94
1
5
Sri Lanka
Nov-94
1
6
Taiwan
Mar-96
1
4
1997
Thailand
1997
Nov-96
0
4
Turkey*
1994
93 & 95
0
5
Venezuela
1994
Dec-93
1
5
Mar-96
1
6
Zimbabwe
Notes
Note: 1=Presidential System. * Turkey has a parliamental system with a strong president.
Data Source: Microsoft Encarta Encyclopedia and CIA Factbook (obtained at WEB
http://www.odci.gov/cia/publications/factbook/country-frame.html)
Table 2: Summary Statistics of Crisis Variables (by Political Dummy)
Financial Current Capital Corrupt 3 year
3 year % ShortTotal
Equity Change Equity Contagion
Crisis Account Inflow to Index Change change in term debt debt to Return $
in
Market
to GDP
GDP
in Credit Real FX
to GDP reserve
Currency Volatility
to GDP
rate
ratio
Value
Political
Mean
St. Dev
0.23
0.43
-0.03
0.04
0.06
0.09
3.47
0.81
0.07
0.13
-17.09
26.74
1.15
1.16
2.06
2.21
0.02
0.53
-0.17
0.22
0.11
0.04
0.34
0.48
NonPolitical
Mean
St. Dev
0.02
0.15
-0.02
0.09
0.01
0.21
3.63
0.95
0.04
0.26
-14.95
25.13
1.04
1.04
2.29
3.37
0.12
0.73
-0.09
0.14
0.08
0.06
0.14
0.35
T-stat.
2.71
-0.97
1.36
-0.78
0.55
-0.36
0.45
-0.36
-0.69
-1.83
2.62
2.09
Correla tions
C. Acc
-0.15
Cap. FL
0.07
Corrupt
-0.14
Credit
0.20
Real FX
0.00
ST Debt
0.24
Debt/Res
0.01
Eq. Ret$
-0.31
Devalue
-0.53
Volatility
0.31
Political
0.32
Contagion
0.28
-0.9
-0.06
0.31
-0.26
0.34
0.38
0.73
-0.24
0.53
-0.10
-0.10
0.05
-0.33
0.31
-0.37
-0.38
-0.73
0.29
-0.54
0.14
0.08
-0.05
0.31
-0.03
0.07
-0.03
0.15
-0.08
-0.09
-0.26
0.31
-0.07
-0.07
0.3
-0.34
0.29
0.06
0.16
-0.25
-0.21
-0.29
0.04
-0.31
-0.04
-0.09
0.78
0.13
-0.26
0.27
0.05
-0.12
0.16
-0.23
0.2
-0.04
-0.06
-0.10
0.46
-0.08
-0.15
-0.65
-0.21
-0.24
0.28
0.08
0.24
Data Sources: The political dummy variables are based on election information provided in
World Factbook published by CIA and confirmed by Microsoft's Encarta Encyclopedia.
Radelet and Sachs (1998) provided the economic variables and crisis definition.
Note: The current account to GDP ratio, the capital Inflow to GDP ratio, 3 year Change in
Credit to GDP ratio, 3 year % change in Real FX rate, Short-term debt to GDP ratio, and
total debt to reserve ratio are measured at the end of last year.
 Table
3 presents some summary statistics
according to financial crisis.
significantly higher current account deficit,
 higher capital inflows,
 larger change in bank credit in the past three years,
 and higher short-term debt to GDP ratios.

A
Probit Analysis of Emerging Market
Crises
Yi  a  bX i  cDi   i
9
Table 3: Summary Statistics of Crisis Variables (by Crisis Countries)
Current Capital Corrupt 3 year
Account Inflow Index Change
to GDP to GDP
in Credit
to GDP
3 year % Short- Total Equity Change in Equity Political Contagion Year
change term debt to Return Currency Market
(0 for
Crisis
in Real debt to reserve
$
Value
Volatility
nonOccur
FX rate GDP
ratio
political
years)
-25.66
1.57
2.74
12.7%
0.1%
10.6%
1
1
1995
-8.22
1.70
2.89 -73.7% -46.0%
16.5%
1
1
1997
-6.65
2.06
3.04 -68.7% -50.1%
14.3%
1
1
1997
-13.78
0.61
1.08 -71.7% -35.1%
11.0%
0
1
1997
-30.74
5.28
3.40 -26.0% -36.2%
14.6%
1
0
1995
-22.64
0.85
1.44 -61.9% -34.1%
9.6%
1
1
1997
-11.24
1.45
2.21 -79.3% -46.6%
13.0%
1
0
1997
-11.32
2.06
2.26 -40.2% -62.2%
20.2%
1
0
1994
-12.09
0.81
1.70 -25.7% -41.5%
14.8%
1
0
1994
Argentina
Indonesia
Korea
Malaysia
Mexico
Philippines
Thailand
Turkey
Venezuela
-0.04
-0.04
-0.05
-0.08
-0.08
-0.05
-0.08
-0.04
-0.03
0.07
0.06
0.05
0.04
0.07
0.10
0.09
0.09
0.06
3.00
2.00
4.00
4.00
3.00
3.00
3.00
4.00
3.00
0.06
0.07
0.08
0.69
0.20
0.22
0.20
0.01
-0.01
Crisis
Mean
St. Dev
-0.05
0.02
0.07
0.02
3.22
0.67
0.17
0.21
-15.82
8.42
1.82
1.40
2.31
0.78
-0.48
0.31
-0.39
0.17
0.14
0.03
0.89
0.33
0.56
0.53
-
Non-crisis
Mean
St. Dev
-0.02
0.07
0.03
0.18
3.60
0.91
0.04
0.20
-15.92
27.21
0.99
1.01
2.17
3.07
0.15
0.64
-0.09
0.15
0.09
0.05
0.39
0.49
0.19
0.39
-
T-stat.
-2.80
1.72
-1.53
1.70
0.02
1.71
0.30
-4.92
-5.01
3.92
3.95
2.02
-
Data Sources: The political dummy variables are based on election information provided in World
Factbook published by CIA and confirmed by Microsoft's Encarta Encyclopedia. Radelet and
Sachs (1998) provided the economic variables and crisis definition.
Note: The current account to GDP ratio, the capital Inflow to GDP ratio, 3 year Change in Credit to
GDP ratio, 3 year % change in Real FX rate, Short-term debt to GDP ratio, and total debt to reserve
ratio are measured at the end of last year.
Table 4: Probit Analysis
 the political dummy turns out to be quite significant
even after adjusting
 pseudo R-square increase from 0.37 with six
independent variables to 0.63 with only four
independent variables.
 a higher ratio of short-term debt to reserves (liquidity)
 a rapid buildup in the claims of the banking sector
 a larger current account deficit or capital flows
(weakly)
 real exchange rate overvaluation: close to zero
 corruption not significant
 contagion appear to be less important than political risk
10
Table 4. Probit Results for Financial Crisis
__________________________________________________________________________________________________
Output
Probit results
I
II
III
Independent variable
IV
V
VI
0.543
(2.12)**
2.501
(1.96)**
0.538
(2.04)**
0.590
(2.05)**
2.201
(1.70)*
0.353
(1.76)*
Private credit/GDP
3.774
(2.51)**
4.147
(2.51)**
4.152
(2.51)**
3.967
(2.55)**
5.178
(2.35)**
2.145
(2.03)**
Total debt/reserves
-1.071
(-1.55)
3.203
(1.63)
1.923
(0.55)
3.653
(1.82)*
3.357
(1.81)*
-0.008
(-0.65)
Corruption
3.174
(2.03)**
-6.108
(-0.98)
1.486
(2.11)**
1.589
(2.16)**
1.506
(2.01)**
-0.709
(-0.05)
-0.363
(-1.06)
-0.621
(-1.27)
1.420
(2.01)**
2.287
(1.93)*
1.574
(2.19)**
Contagion
1.308
(1.78)*
1.009
(2.08)**
Polticalrisk*Parliamentary
Pseudo Rsquared
2.610
(1.76)*
-26.911
(-1.67)*
Real exchange rate
Constant
0.639
0.579
(2.66)*** (2.23)**
-0.820
(-1.08)
Current acct surplus/GDP
Political risk
VIII
Coefficient (Z stat)
Short term debt/ Reserves
Capital inflow/GDP
VII
0.742
(1.41)
0.197
(0.34)
-3.483
-3.662
-3.675
-2.302
(-3.98)*** (-3.68)*** (-3.77)*** (-1.74)*
0.63
0.86
0.64
0.65
-3.041
-2.997
-2.809
-3.599
(-1.57) (-3.83)*** (-4.86)*** (-4.02)***
0.88
0.56
0.51
0.62
No. of obs
78
78
78
78
78
78
78
78
***
1% significance
**
5% significance
*
10% significance
__________________________________________________________________________________________________
Data Sources: The political dummy variables are based on information provided in World Factbook
published by CIA and confirmed by Microsoft's Encarta World Encyclopedia. Radelet and Sachs (1998)
provided the economic variables and crisis definition.
 1. Changes in the currency value (in dollars):
si  a  bX i  cDi   i

 change in bank credit has a very significant negative
impact on currency value.
 the political dummy a strong negative impact on
currency
 foreign capital inflows positive
2. Equity market returns in dollars.
Ri  a  bX i  cDi   i
 high current account (surplus)
 high capital flow to GDP ratio (lower)
 Warning: information lags
11
 3.
Volatility of equity market returns in
dollars.  i  a  bX i  cDi   i
 bank credit has a very significant impact
 changes in real exchange rates (currency appreciation)
 political risk has significant impact
 why volatility differs across countries and why
volatility shifts through time

Implication for Risk Management
 investors and government should increase protection
against devaluation and crisis
 Political risk premium should adjust according to
political risk cycles.
12
Table 5: Regression Analysis of Determinants of Currency
Devaluation, Equity Returns (in Dollars) and Market Volatility
________________________________________________________________________________
Dependent variable
% Change in
Currency
Value ($)
Equity
Return ($)
Equity
Market
Volatility ($)
Constant
-0.092
(-1.06)
0.134
(0.60)
0.058
(2.20)**
Current Account to GDP
0.541
(0.99)
3.972
(2.94)***
0.177
(1.19)
Capital Inflow to GDP
0.364
(1.64)
-1.114
(-1.89)*
-0.068
(-1.16)
Corrupt Index
0.018
(0.89)
0.036
(0.56)
0.005
(0.66)
3 year change in credit to GDP
-0.220
(-3.02)***
0.442
(1.01)
0.049
(2.66)***
3 year %change in real FX rate
-0.000
(-0.31)
-0.005
(-2.19)**
-0.000
(-2.04)**
Short-term debt to GDP
-0.026
(-1.15)
-0.091
(-1.94)*
0.006
(1.07)
Total debt to reserve ratio
-0.007
(-1.22)
-0.009
(-0.77)
-0.002
(-1.33)
Political Dummy
-0.064
(-1.87)*
0.045
(0.47)
0.031
(3.74)***
Contagion
-0.072
(1.38)
-0.215
(-1.91)*
0.003
(0.33)
Adjusted R-square
0.196
0.571
0.403
78
78
78
No. of obs
***
1% significance
**
5% significance
*
10% significance
_______________________________________________________________________________
Data Sources: The political dummy variables are based on information provided in World
Factbook published by CIA and confirmed by Microsoft's Encarta World Encyclopedia.
Radelet and Sachs (1998) provided the economic and financial variables. . The tstatistics have been adjusted for heteroscadaticity using the White-matrix.