Fiscal Policy, Bond Prices, Credit Ratings and Original Sin

Precarious creditworthiness Pro-cyclical outcomes Electoral budget cycles Today, the first two

200

400

600

800

1000

1200

Jan

-97

May

-97

Se

p-9

7

Jan

-98

May

-98

Se

p-9

8

Jan

-99

May

-99

Se

p-9

9

Jan

-00

May

-00

Se

p-0

0

Jan

-01

May

-01

Se

p-0

1

776 pb

Market access is a problem for most EMs(LEI, Spread over US Treasuries)

Pre-Asian Crisis

Pre-Russian Crisis

Pre-Argentine Crisis

Current level

Albania

Algeria

Angola

Azerbaijan

BangladeshBolivia

Bosnia and HercegovinaBrazil

BulgariaCameroon

Chile

China

ColombiaCosta Rica

Cote d'Ivoire

Croatia

CubaDominican Republic

Ecuador

Egypt

El Salvador

Ethiopia

Gabon

Ghana

Guatemala

Honduras

IndiaIndonesia

Iran

Jamaica

Jordan

Kazakhstan

Kenya

Latvia

Lebanon

Libya

Macedonia

MalawiMalaysia

Mauritius

Mexico

Moldova

Morocco

MozambiqueNamibia

Nicaragua

Nigeria

PakistanPanama

Papua New Guinea

Paraguay Peru

Philippines

Poland

RomaniaRussia

Saudi Arabia

Senegal

South Africa

Sri Lanka

Sudan

Syria

Tanzania

Thailand

TunisiaTurkey

UgandaUkraine

Uruguay

Uzbekistan Venezuela

Vietnam

Yemen

Zambia

Zimbabwe

050

100

150

200

Pub

lic d

ebt (

% o

f GD

P)_

__E

IU

5 6 7 8 9 10LGDPPC

MaastrichtCriterion

12

24 24

8

32

1 1

05

1015

2025

Fre

que

ncy

0 50 100 150 200Public debt (% of GDP)___EIU

Public debt/GDP

ParaguayGuatemala

Rep DomHaitíEl Salvador

ColombiaUruguayCosta RicaMéxicoT&T

VenezuelaChile

ArgentinaPerúBolivia

BelicePanamá

HondurasEcuador

0 20 40 60 80 100 120 140

Debt to GDP ratios look modest

Y

Dx 1

rati

ng

fo

reig

n c

urr

ency

net_debt/gdp-.291965 1.13803

5

19

ARG

AUS

AUT

BEL

BRA

CAN

CHN

CRI

CYP

CZE

DNK

DOM

EST

FIN

DEU

GRC

HUN

ISL

IND

ISR

ITA

JPN

JOR

LVA

MEXMAR

NOR

PAK

PAN

PRY

POL

PRT

SVN

ESPSWE

TTO

TUN

TUR

GBR USA

tY

Dx 2

cred

it ra

ting

1992

-99

aver

age

DE_RE2-.579362 4.13906

5

19

ARG

AUS

AUT

BEL

BOLBRA

CAN

CHL

CHN

COL

CRI

CYP

CZE

DNK

DOM

SLV

EST

FIN

DEU

GRCHUN

ISL

INDIDN

ISR

ITA

JORKAZ

KOR

LVA

LUX

MLT

MEX

MNG

MAR

NOR

PAN

PRY

PER

POL

SGP

SVK

SVN

ZAF

ESPSWE

CHE

THA

TUN

TUR

GBR USA

tY

iDx 3

Interest payments as a share of total government expenditures 1990-94

0

5

10

15

20

25

30

OECD LAC

Argentina

Austria

Belgium

BotswanaCanada

Cyprus

Czech RepublicDenmark

Estonia

FinlandFranceGermany

GreeceHungary

Ireland

Israel

Italy

JapanNetherlandsNew Zealand

Portugal

QatarSlovakiaSloveniaSpain

Sweden

Trinidad & TobagoUnited KingdomUnited States

AlgeriaAzerbaijan

Bolivia

Brazil

Bulgaria

Chile

ColombiaCosta Rica

Croatia

Dominican Republic

Ecuador

Egypt

El Salvador

Ethiopia

India

Indonesia

Jamaica

Kazakhstan

MalaysiaMexico

Pakistan

Papua New GuineaPeru

Philippines

Poland

Romania

Russia

South Africa

Thailand

Tunisia

Turkey

Uruguay

Venezuela

Yemen

0.0

5.1

.15

PU

B_

DE

BT

_SE

RV

_G

DP

0 50 100 150 200Public debt (% of GDP)___EIU

PUB_DEBT_SERV_GDP PUB_DEBT_SERV_GDP

It must involve some risks

Markets are concerned with the solvency of the government. There is a maximum level of debt service that the government can absorb

But there is uncertainty over future flows

)Pr(_

xxi

i

tY

Dx

Figure 1

X~

Prob.

X OECD

X LA X

Same expected value

Greater variance

i

x

_x

i*

itY

Dx

)Pr(_

xxi

Sound

Fragile

Bankrupt

i

x

_x

i*

High variance

Less variance

High risk causes high interest rates Fat tails raise the interest rate

High interest rates causes high risk High interest rates increases the expected

value of debt service BUT WHAT RISKS ARE WE TALKING ABOUT?

tY

iDx 3

Table 1: Volatility of GDP Growth (1980-1999)

All Countries

Industrial Countries

Developing countries

Real GDP Growth 5.0% 2.7% 5.8%

N. Countries 43 11 32

Fiscal revenue and GDP volatility(stdev of real rates of growth)

Revenue GDP

OECD (1970-1994) 5.2 2.2

LAC (1970-1994) 15.2 4.7

LAC (1994-2000) 7.9 3.4

Venezuela 21.9 4.8Ecuador 17.5 3.5Mexico 9.9 3.7Peru 6.6 4.3Uruguay 6.4 3.9Chile 6.4 3.7Colombia 5.0 2.9Argentina 4.0 5.0Costa Rica 3.0 3.0Panama 3.0 2.5Bolivia 3.0 1.4

Volatiliy

Stress test on the debt service capacity: an illustration.OECD LAC

4.0% 10.0%30.0% 20.0%40.0% 40.0%

133.3% 200.0%1.6% 4.0%5.3% 20.0%

Real Interest rate (local)Fiscal revenue to GDPDebt to GDPDebt to revenue Debt service to GDPDebt service to revenue

Impact of 1 standard deviation shock to revenues on the the debt service to tax ratio

Shock Impact OECD -5.0 0.3 LAC -8.0 1.7

Its part of the problem, but only part.

Definition: You cannot borrow abroad in your own currency

…if in addition you do not have long-term fixed-debt markets in local currency

…you are condemned to choose between short term debt in pesos, or long term debt in dollars.

This makes debt service sensitive to the real exchange rate and the real interest rate

t

ttlt

lt

st

st

tY

eDiDiDix

*1

*1111

4

Volatility of real interest rates Volatility of real exchange rates

Volatility of revenue

Table 1: Volatility of GDP Growth (1980-1999)

All Countries

Industrial Countries

Developing countries

Real GDP Growth 5.0% 2.7% 5.8%

Real Dollar GDP Growth

12.3% 10.3% 13.0%

N. Countries 43 11 32

Country YearChange in Dollar

GDPChange in Real

GDP

Nigeria 1999 -74% -2%

Uruguay 1984 -67% -8%

Egypt, Arab Rep. 1991 -63% 4%

Indonesia 1998 -60% 7%

Sierra Leone 1986 -57% -10%

Mexico 1983 -56% -9%

Uruguay 1983 -55% -17%

Costa Rica 1982 -54% -10%

Nigeria 1986 -52% 1%

Syrian Arab Republic 1989 -48% 9%

Jamaica 1985 -46% 4%

Honduras 1991 -46% -4%

Country YearChange in Dollar

GDPChange in Real

GDP

Jordan 1990 -40% -19%

Guatemala 1987 -40% 3%

Syrian Arab Republic 1988 -40% -13%

Trinidad and Tobago 1987 -38% -20%

Togo 1982 -38% -15%

Mexico 1982 -38% 8%

South Africa 1985 -38% 4%

Ecuador 1987 -38% 1%

Egypt, Arab Rep. 1992 -37% 6%

Indonesia 1999 -37% -7%

Egypt, Arab Rep. 1990 -36% 10%

Trinidad and Tobago 1986 -36% -13%

Swaziland 1985 -36% 2%

Namibia 1985 -35% 15%

Country YearChange in

Dollar GDPChange in Real

GDP

Paraguay 1985 -35% 13%

Ecuador 1999 -33% -2%

Jamaica 1984 -33% 12%

Papua New Guinea 1999 -33% -5%

Mexico 1995 -33% 1%

Sierra Leone 1998 -31% -22%

Sweden 1982 -31% -1%

Papua New Guinea 1998 -31% -4%

Madagascar 1988 -31% 7%

Jamaica 1992 -30% -10%

Morocco 1982 -30% 1%

Venezuela 1984 -30% 4%

AVERAGE -46% -2%

Industrialized countries Developing countries

Volatility of Real Exchange rate over 5-year periods(Cross-Country Average Normalized to 1)

Norw

ay

Neth

erl

ands

Irela

nd

Denm

ark

Aust

ria

Baham

as,

The

Fra

nce

Isra

el

Canada

Sw

itze

rland

Germ

any

Taiw

an

Cypru

sG

reece

Papua N

ew

Guin

ea

Sw

eden

St.

Vin

c&G

renadin

es

Malt

aIc

ela

nd

St.

Luci

aG

am

bia

, The

Belg

ium

Leso

tho

Kore

a,

Rep.

Hong K

ong,

Chin

aSpain

Fin

land

Unit

ed K

ingdom

Italy

Moro

cco

Aust

ralia

Phili

ppin

es

Thaila

nd

Cost

a R

ica

Beliz

eJa

pan

New

Zeala

nd

Togo

Tunis

iaH

ungary

Unit

ed S

tate

sSin

gapore

Fiji

Turk

ey

India

Mala

ysi

aPort

ugal

Pakis

tan

South

Afr

ica

Bahra

inPeru

Mexic

oB

razi

lSaudi A

rabia

Tri

nid

ad a

nd T

obago

Chile

Kuw

ait

Para

guay

Cote

d'Iv

oir

eD

om

inic

an R

epublic

Cam

ero

on

Colo

mbia

Arg

enti

na

Indonesi

aB

uru

ndi

Uru

guay

Chin

aV

enezu

ela

Ecu

ador

Zam

bia

Rom

ania

Boliv

iaN

igeri

a

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

1YR

Vol.

5YR

Vol.

All Countries 0.089 0.083

Developing 0.112 0.103

Industrial 0.044 0.041

Difference 0.068 0.062

T-statistics 4.262 4.818

P(Dev>Ind) 1.000 1.000

Real exchange rates: volatility and cyclical properties

volatility elasticity t-stat

United States 9.10% -0.03 -0.9Latin America 21.40% 0.18 8.9Peru 28.40% 0.15 3.1Ecuador 25.70% 0.27 3.7Venezuela 23.60% -1.04 -7.7Argentina 21.10% 0.02 1.2Colombia 19.30% 0.26 6.2Brazil 18.80% 0.42 10.4Mexico 18.30% 0.61 10.7Chile 16.00% 0.32 4.5

Elasticity of real exchange rate to the import gap in the 1990s

Note: excludes periods in which inflation exceeded 40 percent

Volatility elasticity t-stat

United States 0.9 -3.3 -4.1Latin America 10.5 -126.3 -10.9

Mexico 23.0 -73.3 -13.2Venezuela 17.6 0.1 0.0Brazil 17.2 -451.6 -3.4Ecuador 12.2 -2.4 -0.5Uruguay 11.8 2.6 0.4Peru 11.2 -151.4 -1.7Colombia 7.8 -16.6 -2.3Chile 5.4 -8.8 -1.0Costa Rica 5.0 -19.7 -5.0Argentina 4.0 -221.9 -10.3Panama 0.6 -0.4 -0.6

Note: excludes periods in which inflation exceeded 40 percent

(monthly data, 1990-1999)

Real exchange rates and real interest rates

(monthly data, 1990-1999)

Country elasticity t-statUnited States (1990s) -4.5 -1.8Latin America -153.7 -6.2

Argentina -174.5 -0.9Brazil -1430.5 -7.2Chile -66.6 -7.4Colombia -4.7 -0.4Ecuador -26 -5.2Mexico -93.6 -16.2Peru -555.3 -3.4Venezuela -28.8 -4.3

exceeded 40 percent

Elasticity of real interest rates vs. RER

Note: excludes periods in which inflation

Stress test on the debt service capacity: an illustration.OECD LAC

4.0% 10.0%4.0% 10.0%

100.0% 100.0%30.0% 20.0%

GDP 100.0% 100.0%40.0% 40.0%0.0% 20.0%

30.0% 0.0%10.0% 20.0%

133.3% 200.0%1.6% 4.0%5.3% 20.0%

Real Interest rate ($)Real Interest rate (local)Real exchange rateFiscal revenue to GDP

Debt to GDP -foreign currency - domestic long term -domestic short termDebt to revenue Debt service to GDPDebt service to revenue

OECD LAC

SHOCK

Real Interest rate (local) 1.0% 10.0%Real exchange rate -9.0% -21.0%Fiscal revenue to GDP -5.0% -8.0%GDP -2.0% -4.0%

IMPACT

Debt to GDP 0.8% 6.0% -foreign currency 0.0% 5.2% - domestic long term 0.6% 0.0% -domestic short term 0.2% 0.8%Debt to revenue 9.9% 50.2%Debt service to GDP 0.1% 2.7%Debt service to revenue 0.7% 16.3%

Why do countries get into trouble at levels of debt that are manageable for other countries?

Because debt structure (original sin) makes real exchange rates and real interest rates matter for debt service

And these are very volatile and have the “wrong” cyclical properties, making debt service much riskier

(1) (2) (3) (4) RATING1 RATING1 RATING1 RATING1 DE_GDP2 -1.553 -1.815 (1.91)* (2.19)** DE_RE2 -0.599 -0.665 (1.40) (1.52) LGDP_PC 3.189 3.051 2.884 2.764 (8.54)*** (7.59)*** (6.47)*** (5.68)*** OSIN3 -3.429 -3.324 -4.883 -4.435 (3.85)*** (3.49)*** (3.49)*** (3.11)***

Constant -12.369 -11.059 -8.751 -7.889 (3.16)*** (2.60)** (1.89)* (1.57)

Observations 56 49 51 44 R-squared 0.82 0.81 0.81 0.80

Total Debt Total Debt in Total Debt in Own Currency SIN1 SIN3Own Currency Issued by the Country

Financial Centers 5324 77.0% 6698 96.8% 3425 96.8% 0.357 -0.258

Other developed 627 9.1% 121 1.7% 58 1.6% 0.907 0.807

Developing 592 8.6% 97 1.4% 55 1.6% 0.907 0.835

International Org. 373 5.4% 0 0.0% 0 0.0% 1.000 1.000

TOTAL 6916 100.0% 6916 100.0% 3538 100.0% 0.488 0.000

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

United States EUROLAND Japan U.K Switzerland Canada Australia

Debt by Country

Debt byCurrency

(0.9857)

(0.8859)

Major FinancialCenters (61 %)

Major FinancialCenters (45 %)

Euroland(37%)

Euroland(33%)

Other Developed (<2%)

Other Developed (8%) Developing

(8%)

InternationalOrganizations (7%)

Total Debt issued in own currency (99-01)

Total Debt issued by residents (99-01)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Financial Centers Euroland Other Developed Developing

OSIN1 OSIN2 OSIN3

00.10.20.30.40.50.60.70.80.9

GoldClauses

MixedClauses

DomesticCurrency

OSIN3 and Flandreau-Sussman classification circa 1850

Why are countries trapped in a pro-cyclical fiscal response in bad times?

Because in bad times they need to finance not just the decline in tax revenues but also the jump in debt service

This makes it harder to maintain spending and access to lenders

Targets on the overall deficit are not credible because governments cannot control debt service Primary balances are more credible but

may be less relevant Reduction of the absolute level of debt

or even the debt to GDP ratio may not be the most efficient way to achieve fiscal consolidation Working on debt denomination may be

more effective

Governments should target a risk-weighted debt level Risk weights should depend on the volatility

and cyclicality of its determinants In the previous examples, short term domestic

currency debt should get the highest weight. Followed by foreign currency debt

A risk-weighted debt target should create incentives for countries to optimize the trade-off between cheaper and safer debt A deficit target favors cheap, risky debt

Long-term fixed-rate domestic-currency debt is best but it is hard to develop Bravo Mexico

Inflation-indexed, long-term, fixed-rate domestic debt are second best and are easier to develop Chile dixit

Working on debt structure may be a more efficient way of achieving fiscal consolidation Virtuous circle

Inflation-indexed fixed-rate long-term domestic- currency debt may be a practical way to go

Target adjusted primary deficits …and risk-weighted debt levels