inﬂation stabilization in chronic inﬂation countries: the ... · the capital account improves...

Munich Personal RePEc Archive

Inflation stabilization in chronic inflation

countries: The empirical evidence

Reinhart, Carmen and Vegh, Carlos

University of Maryland, College Park, Department of Economics

August 1994

Online at https://mpra.ub.uni-muenchen.de/13689/

MPRA Paper No. 13689, posted 02 Mar 2009 05:53 UTC

JORNADAS INTERNACIONALES DE ECONOMIA

trprogramas de estabiljzaci6n:

La experiencia reciente I';;fl Amenea Latina ft

BANCO CENTRAL DEL URUGUAY Hontevideo, 8 Y 9 de agosto de 1994

ESTABILIZACION EN PAISES DE INFLACION CRONICA:

La Evidencia Empirica

CARMEN REINI~T (Departamento de Investigaciones, FMI)

CARLOS VEGH (Departamento de Investigaciones, FMI)

- T

Outline

For discussion at the conference only

July 31, 1994

Inflation Stabilization in Chror:ic Inflation .countries:

The Empirical Evidence 11

by

Carmen M. Reinhart and Carlos A. Vegh

Research Department, IMF

I. Introduction

Chronic inflation is characterized by long periods (i.e., several

decades) of high (relative to industrial countries) and persistent ,·inflation

(Pazos, 1972). The phenomenon of chronic inflation emerged in several Latin

American countries after World War II, and has been a distinguishing feature

of the economic landscape of several countries ever since. After numerous

failed attempts, some countries have succeeded to reduce inflation close to

international levels (Argentina, Chile, Israel, and Mexico).

There is a rich history of stabilizatj~ons in chronic inflation

countries, which spans more than three decades. This provides a unique

opportunity to identify the main patterns of adjustment and examine

11 This outline has been prepared to be discussed at the conference on

"Inflation Stabilization: The Recent Experience in Latin America," organized

by the Central Bank of Uruguay, to be held on August 8-9, 1994. The views

expressed her~ are those of the authors arid do not necessarily represent

those of the IMF.

r

- 2 -

econometrically some of the main features of disinflation in chronic

inflation countries. Our study is based on 17 stabilization plans from 1964

to the present in 7 countries: Argentina, Brazil, the Dominican Republic,

Israel, Mexico, Peru, and Uruguay (see Table 1). Twelve of these programs

were based on the use of the exchange rat,~ as the nominal anchor, while 5

were based on the use of a monetary aggregate.

In addition to reviewing the main "stylized facts" for these 17

stabilization plans (in the spirit of Kigne1 and Liviatan (1992), Vegh

(1992), and Calvo and Vegh (1994)), we resort to different econometric

exercises and thus are able to provide sm~e rigorous econometric basis for

several key features of disinflation in cllronic inflation countries.

II. Empirical regularities

This section reviews the main empirieal regularities of exchange rate-

based and money-based programs. These are:

1. Inflation converge only slowly to the rate of growth of the nominal

anchor (Chart 1).

2. The real exchange rate appreciates in both exchange rate-based and

money-based programs (Chart 2).

3. In exchange rate-based programs, there is an initial boom in

economic activity (real GDP and consumption) followed by a later

contraction. In money-based programs, the recession occurs at the beginning

of the programs (Table 2).

4. The external accounts (trade and current account) worsen in

exchange rate-based stabilizations.

- 3 -

5. The capital account improves sub!;tantially in exchange rate-based

stabilizations (although external factors seem to play an important role),

see Table 3. A similar story applies to stock markets (Chart 4). The boom

in asset markets is not limited to the equity market, but is also evident in

real estate prices, which tend to rise throughout the program.

6. International reserves appear to increase substantially in both

exchange rate-based and money-based stabilization, with the resulting

accumulation being more pronounced in periods of heavy capital inflows

(Chart 3 and Table 3). The accumulation of reserves in money-based

programs suggests considerably intervention in foreign exchange rate

markets, which may reflect an unwillingness to let the real exchange rate

appreciate.

7. Real interest rates have increased in the initial stages of money

based-stabilization. In exchange rate-based stabilizations, real interest

rates have generally fallen in orthodox pl~ograms (although in Chile they

remained at very high levels) and increased in heterodox programs (Table 4).

8. The fiscal stance varies considel~ably across programs. In

successful programs, there is a substantial fiscal tightening. In other

cases, however, the fiscal outcome may simply reflect the pick-up in

economic activity and the resulting increase in tax revenues (Table 5), as

emphasized by Talvi (1994).

III. Inflation Stabilization and Economic Growth

Both theory and casual evidence suggest that exchange rate-based

-----_._-_ ..-_.- •.._.~~-----------------

- 4 -

stabilization programs have been characterized by a boom-bust pattern in

economic activity, particularly private consumption. Money-based plans are

associated with and initial recession followed by recovery in economic

activity. The evolution of real GDP growth and consumption growth shown in

Table 2 highlights the marked differences in the so-called "inflation-

output" tradeoff between the two types of plans. The more systematic

econometric treatment discussed below also lends support to the "recession

now versus recession later hypothesis".

A panel of seven countries which have collectively implemented

seventeen inflation stabilization plans o,rer the last thirty years was used

to test if mean rates of growth in real GDP are statistically different

during inflation stabilization plans. The sample is 1964-1993 and the

countries are: Argentina, Brazil, Chile, Dominican Republic, Mexico, Peru,

and Uruguay. The results are robust to variations in the estimation

technique; the fixed-effects estimates are presented in Table 6.

The main results are:

1. The mean rate of growth increases in the early stages of an

exchange rate-based plan: the estimated irLcrease is about 2.3 percent and

statistically significant.

2. During the last year of the program, a recession ensues. Note

that these estimates do not imply an output decline of 5.4 percent, but a

decline of 1.8 percent (5.4 below the mean growth rate of 3.6). The

difference in growth in the final year of the program is also statistically

significant.

- 5 -

3. The mean rate of real GDP growth falls by about 5.4 percent in

the early phases of money-based programs. There is also evidence of a later

rebound in economic activity, but (unlike the early recession) this

difference is not significant.

4. The cumulative effect on output of an exchange rate-based

stabilization plan varies considerably as the duration of the plan varies.

As Table 7 illustrates, given the magnitude of the late recession, the level

of output ends up being lower for plans lasting three years or less. Short-

lived plans are costly.

5. More generally, the evidence suggests that output is not

invariant to th~ policy regime, as suggested by some of the recent

endogenous growth literature.

IV. How effective are the nominal anchors?

The speed and magnitude of the decline in inflation depends crucially

on a multitude of institutional factors as well as on macroeconomic and

microeconomic policies. The analysis that follows does not attempt to trace

out the various transmission mechanisms through which the nominal anchor

(the exchange rate or money) affect prices. Rather, the emphasis is on

examining the "reduced-form" dynamic interaction between these two variables

during selected stabilization plans. The goal is to assess how quickly a

policy shock is transmitted to.inflation, and thus make statements about the

presence or absence of inflation inertia and the effectiveness of policy.

A total of 11 plans are examined (nine exchange-rate based and two

money-based) .. A bivariate vector autoregi:ession (VAR) using monthly

- 6 -

inflation rates and devaluation rates is estimated for the exchange rate-

based plans; money growth (Ml) and inflation define the system for the

money-based plans. The period of estimation covers only the duration of the

plan, hence it is free from the parameter inconstancy that is at the heart

of the Lucas critique. We evaluate, the effectiveness of policy at two

levels.

First, we determine if there is a systematic and significant causal

relationship between the nominal anchor and the target variable, inflation.

Significant causality is necessary but not sufficient for policy to be

effective, since the impact may be significant but quantitatively small.

"For instance, one would expect the pass-through from exchange rates to

prices to depend on the openness of the economy, the degree of indexation,

the extent of credibility, and so on. Secondly, to assess the issue of

quantitative importance as well as speed of adjustment we examine the impact

on inflation of a hypothetical tempo.rary s:hock to the nominal anchor.

The main results can be summarized as foll,ows:

1. In eight out of the nine exchange rate-based programs there is

a significant causality from the rate of devaluation to the rate inflation

(Table 8). The exception is the Argentine 1967 plan, where the fixity of

the rate resulted in serious collinearity problems. The recent

Convertibility plan also shows some of the same problems. As one would

expect with an exogenous policy instrument, the exchange rate is not

affected by past (or current) inflation with the exception of Israel. Money

also appears exogenous during the money-based plans considered. However,

significant causality from money growth to inflation was detected in the

- 7 -

Chilean 1975 plan. Thus, causal patterns appear to be constant across

countries and across time.

2. In six of the nine episodes examined, past inflation

significantly affected current inflation, indicating systematic inflation

inertia. The inertial behavior was most pronounced in Chile, Mexico, and

Israel.

3. Charts 5-8 plot the response of monthly inflation to a

reduction in the rate of devaluation (or money growth) at an annual rate of

20 percent lasting 12 months. Note that:

a) The smallest impact on inflal:ion is seen in the plans of the

1960s, where inflation falls by 13.6 perceIlt for Argentina and only 10

percent for Uruguay (this is the lowest response in our sample). It may

possibly be reflecting the relatively closed nature of the economy at that

time.

b) The maximum effectiveness is reached in the Tablita plans,

where the decline in inflation ranges from 113.5 percent for Argentina to 22

percent for Uruguay, with Chile in the middle. Further, the decline in

inflation for the cases of Argentina and Uruguay is fairly sudden. These

two observations possibly suggest that the m)sence of more significant

declines in inflation in these programs may l:lavehad more to do with "policy

inertia" rather than with "inflation inertia". The inflation inertia

argument appears to find more support in the Chilean data, as it takes a

year for inflation to decline to its minimum level.

c) The exchange rate pass-through appears to decline in the 1980s

and 1990s. For the four exchange rate-based programs, the maximum decline

----------------------------------------------------

- 8 -

in inflation in response to a twenty percent reduction in the rate of

devaluation ranges from 11 percent for Mexico to 15 percent for the

Argentine Convertibility plan. This result remains puzzling. Perhaps

greater indexation and/or lower credibility may account for this change.

The time (in months) it takes to reach the minimum ranges from 3 months for

Uruguay to 14 months for Israel.

d) With regard to the money-based plans, the response to the

shock for both Chile and the Dominican Republic is gradual, taking 12-15

months to reach its most pronounced effect~ The inertia is also evident in

the Chilean 1975 money-based plan, where It takes a year for inflation to

decline by about 14 percent. Possibly, the more delayed effects may reflect

the inherent difficulties in monetary con1:rol.

e) Based on three indicators of inertia (the significance of past

inflation, the response to a temporary policy shock, and the speed of

adjustment) for the more recent peri?d, I~:rael and Mexico and, to a lesser

degree Chile, stand out as the countries in which inflation inertia has been

most pervasive. On the other hand, in Argentina, the Dominican Republic and

to a lesser extent, Uruguay, inflation inertia does not appear to be that

much of a problem. The results also highlight that the interaction between

instrument and target, not only vary across countries (which is obvious),

but can change markedly across time.

Table 1. Major Inflation Stabilization Plans

Plan Pori 0 d

Exchange Rate-Based Stabilization Plans

Orthodox plans

Argentine tablita

Chilean tablita

Uruguayan tablita

Convertibility Plan (Argentina)

Uruguay 1991

1978:12 - 1981:2

1978:2 - 1982:6

1978: 10 - 1982: 10

1991:4 - present

1991: 1 - present

Heterodox plans

Argentina 1967

Austral (Argentina)

Brazil 1964

Cruzado (Brazil)

Israel 1985 11 .

Mexico 19871/

Uruguay 1968

1967:3 - 1970:10

1985:6 - 1986:9

1964:3 - 1968:8

1986:2 1986: 11

1985:7 - 1990:6

1987: 12 - 1992:2

1968:6 -1971:12

Money-Based Stabilization Plans

BONEX (Argentina)

Collor (Brazil)

Chile 1975

Dominican Republic

Peru

1989:12 - 1991:1

1990:3 -1991:1

1975:4 - 1977: 12

1990:8 - present

1990:8 - present

1 /These a re success fu l p rog ram s, hence the te rm Ina l da te has been a rb itra rily

se t to five ye a rs .

Table 2. Inflation Stabilization:Real GDP

( •.nou- .•.l rAt. of growth, in peroent)

Proc; r.a.Dl

thr ••

yeAr. Year

belor. Firat. Second Third proqr •.•

ergST"" Iter Year refT ended

Exchange Rate-Based Plans

Argentina 1967Argen tine Tablita

Argen tina (Austral)Arg en tin e Coo vertib ilityBrazil 1964

Brazil (Cruzado)O1ilean Tablita

Israel 1985MeXico 1987

Uruguay 1968Uruguayan Tablita

Uruguay 1991Average

6.5-4.0

-0.4

-2.7

Money-based Plans

Argentina (Bono:)Brazil (Collor)

O1ile 1975Dominican Rep.1blic 1990Peru 1990Average

-1.8

24-1.9

4.6-3.9

-0.1

3.30.2201.1

0.23.50.70.9

0.1-4.4

-12.9-5.4-4.4-5.4

0.87.1

7.38.9

26

7.68.2

4.01.21.1

6.20.04.8

8.7

23.1

8.3

3.63.3

6.36.05.37.1

3.30.7

4.9 26-5.7

26

Growth and Consumption CyclesReal Private Consumption

CAnna •. l rat. of qro'ttth, in percent)

'1'h r ••

Progra.JIl,

,. ••. ra r •• r

before Fir.t Second Third proqra.m

pro9£tm YeAr rear year ended

Exchange Rate- Based Plans

Argen tina 1967Argentine TabIita

Argentina (Austral)Argentine Coovertibility

Brazil 1964

Brazil (Cruzado)O1ilean Tablita

Israel 1985Mexico 1987

Uruguay 1968Uruguayan TablitaUruguay 1991Average

6.8-4.21.2

-2.1

3.6

281.0

0.60.3

0.50.2

-5.3

0.5

3.5

-0.9

261.7

5.0

3.5 10.8

3.67.8 -14.16.1 1.3

4.4 264.8 -1.0

20 -9.65.04.6 -1.2

7.7-2.8

25

Money-based Plans

0.9

9.9

Argentina (Bono:)Brazil (Collor)O1ile 1975Dominican Rep.1blic 1990Peru 1990Average

-1.2

-0.5

-6.3

-0.31.5

-1.45.4

Hot •• J Dot_ lndi.c~t. the dAtA Are not .,..i.l.bl •• D•• h •• J.ndJ.oAt. d.t. do not .pp1 r.

r11••• Are .ucce •• ful p rog rA II., h'.nee the eer1!in.l d,.t. b,.. b •• n .rbitr.ri.ly •• t to fi,... y •• r •.

O n. r •• r b .fo r ••

loure • ., Buf",a" and z,.J.d.r",a" 0 ")1 , Fav.ro and s.".J.on 0 ")1 , /Ugu.l and z,J..,J.otan {l"'I,

z,u.tJ.g 0 "21 , x.d.lro. 11 ")/, ral.,J., { l"41 , Vlano (1 "01 , Int.rnatJ.onal FlnanoJ.al ItatJ..tJ.ca IIXFI,

world S.'nJc t.bl •• , Fund~ci.on H.dit.zr~n ••. , nAtio'n.l .ourc •• , .nd Fund .e.ff •• e1".tt •••

2614.4

7.915.73.36.4

7.514.81.8

8.2

9-1.5

7.5

-1.8

-2.5

-11.4

-12.9-15.3

-8.8

10.10.7

6.59

6.3

6.4

510.46.4

0.37.5

10.86.2

4.05.6

6.4 4.4

-3.6

0.7

4.0

4.3 10.2

-0.96.8 -12.1

4.3 0

5.7 4.9

24 -9.7

13

5.9 -0.5

3.9

16

-1.1

10.0

:pro9r~l'll

T a b le 3 . T h e E x te r n a l A c c o u n t s

Current account balance excluding offk:ial transfers

T h r c o

yeAr. Year

b e fo r e r ir .t S e o o n d T h ir d p r o 9 r a .m

p r o q r am Y .a .r r ." r r ." r e n d e d


Argentina 1967Argentine TablitaArgentina (Austral)Argentine ConvertibilityBrazil 1964

Brazil (Cruzado)Chilean Tablita

Israe11985Mexico 1987

Uruguay 1968Uruguayan Tablita

Uruguay 1991Average

1.4

0.5-2.1-1.4

-1.2

-2.3

-2.7-3.3

0.71.7

-3.7

0.5-1.3

Money- based Plans

Argentina (Bonex)

Brazil (Coilor)Chile 1975

Dominican Republic 1990Peru 1990Average

-2.4

-0.3-1.8

-5.9-4.1

-2.9

1.0

-0.4

-2.7-1.5

0.1

-2.1

-7.1

4.7-2.2

1.4

-5.2

0.7-2.1

1.3

-0.7

-5.8

-3.7

-4.3

-2.6

-0.3

-2.3

-3.70.4

-5.7

5.5-2.9

0.9-7.5-1.8

-3.3

1.5

-3.3

-5.1

-2.3

-1.2

-3.3-0.1

-7.1

-2.5

-3.2

-2.7-4.5

1.1

-3.2

-2.6

-5.2

-3.9

-0.7

-2.8

-3.9

-0.8

-0.5-9.5

-7.0

-2.8

-5.2

-4.7

-0.3-4.1

-2.2

Capital balance including errors and omissions(a.. a peroent of aOP)

proqra.JIl

T h r o .

y e a r . ~ear

b e fo r e T ir .t S .c o n d T h ir d p r o q r a J l

p r o g r a .m Y .a r Y e a r Y e a r e n d e d


Argentina 1967Argentine TablitaArgentina (Austral)Argentine ConvertibilityBrazil 1964

Brazil (Cruzado)Chilean Tablita

Israel 1985Mexico 1987

Uruguay 1968Uruguayan TablitaUruguay 1991Average

-1.0

-2.9

-6.0

-0.1

1.0

-0.7

3.1

-3.5-1.3

Money- based Plans

Argentina (Bonex)Brazil (Coli or)Chile 1975Dominican Republic 1990Peru 1990Average

-2.3

-2.8-1.3

-0.0-5.4

-2.3

Hot •• ' Dot. iDdloAt. tb. dAtA ~r. n o t AVA1JAbJ •• D •• h •• indicAt. d.tA do not Apply.

l/~h •• e Are .uoo ••• lul progrA~.J h.noe th. t.r.ln.J d.te b •• been .rbitrArily .et to lip. y •• r •.

$ourc •• , tnt~rnAtion.l FinAnoi.l 6tAti.tlo., And world Econnomic outlookJ tHF.

4.6-1.6

-2.8

-6.4

22.4

-3.4

12.0-2.82.4

-1.8

-2.0-1.0

-13.6

-4.2

-4.5

-4.4

21.2

11.8

-7.2

4.3

3.813.0

7.0

7.9

10.2

3.4

- 3.0

2.6

-0.2

-6.8

-3.2

14.6

24.0

-6.0

7.0

5.4 15.2

18.4

7.4

3.6

-8.0

1.1

-1.6

12.0

5.2

Table 4. Real Interest Rates i.l Selected Stabilization Programs.

Real interest rates b

(in percent per year)

Four

Quarters First Last

before Four Four

Program Quarters Quarters

Programs Period a ~verage) (average) (average)

Exchange Rate- Based

Argentina 1967 1967-1970

Brazil 1964 1964-1968

Uruguay 1968 1969-1971

Argentine tablita 1979-1981 0.7 -2.8 5.9

Chilean tablita 1978-1982 70.9 43.0 46.4

Uruguayan tablita 1979-1982 18.2 -7.2 24.9

Austral (Argentina) C 1986 20.0 48.0 -7.5

Cruzado (Brazil) C 1986 -4.5 8.5 -9.5

Israel 1985 d e 1986-1990 -2.0 21.2 11.0

Mexico 1987d 1988-1992 -2.9 292 2.0

Convertibility (Arg.) f 1991-1992 38.1 -2.0 4.0

Money-based

Chile 19759 1975-1977 127.2 58.0

Bonex (Argentina) 1990 -7.4 112.7

Collar (Brazil) h 1990 -8.1 -2.4

Dominican Rep. 1990 f i 1990-1992 15.1 13.7

Peru 1990 f 1990-1992 --17.3 235.0 48.1

.• Calendaryears during which the program was taken to be in effect for the purpOses

ofcurrentaccountfigure~ .

b Quarterly real lending rates unless otherwise indicated. Periods specified in

Tables 1 and 4 apply. Dots indicate data are not available. Dashes indicate data do not apply.

C Real interest rates are reported for two-quarter periods, and exclude the initial

price shock.

d Duration of program has been arbitrarily set to five years ..

•• Real interest rate before the program refers to two quarters before.

f Program in progress.

g Annualreal interest rates.

/, Monthly averages of overnight interest rates on government securities. Real interest

rate after the program refers to first three quarters.

i Real interest rates for 1991.3 and 1991.4. Before January 1991, interest rates were

subje<:t to controls.

Sources: Bali no (1991), Barkai (1990), Bufman and Leider.nan (1993),

Castro and Rond (1991), Cukierman (1988), Kiguel and Liviatan (1989),

Perez-Campaneroand Leone (1991), International Enand:d Statistics (IMF),

and national sources.

Table 5. Public Sector Balance in Se!ec,ed Stabilization Program$ 1/

(as percentofGCP)

Three

Yean

before

Program Fint Second Third Fourth Fifth

Program$ Period 11 (average) Year Year Year Year Year

Exchange Rate-Ba$ed

Argentina 1967 1967-70 -4.1 -2.0 -2.0 -1.5 -1.6 -4.1

Brazil 196431 1964-68 -4.0 -3.2 -1.6 -1.1 -1.7 -1.2

Uruguay 1968 1%9-71 -1.8 -2.5 -1.3 -5.8 -2.6

Argentine tablita 1979-81 -5.9 -4.7 -6.1 -8.5 -5.3

Chilean tabIita 1978-81 0.8 1.5 4.8 5.4 0.3 -4.0

Uruguayan tablita 1979-82 -1.4 (1.0 -0.3 -1.5 -9.1 -3.9

Austral (Argentina) 1986 -7.1 -2.0 -4.3

Cruzado (Bram) 1986 -3.9 -3.6 -5.5

Israel 19854/ 1986-90 -4.9 ~.3 0.0 -0.3 -4.0 -2.7

Mexican 198741 51 1988-92 -0.3 -4.5 -1.7 2.6 2.3 3.3

<Anvertibility (Arg.) 61 1991-93 -6.0 -0.9 0.6

Uruguay 1991 1991-93 -4.8 -0.9 -0.0 -1.6

Money-based

Chile 1975 1975-77 -20.1 -2.0 3.9 0.4 1.5

Bonex (Argen tina) 1990 -9.3 -2.3 -1.2

<Allor (Brazil) 51 1990 -5.8 1.4 1.5

Dominiean Rep. 199061 1991-92 -5.9 (1.1 1.6

Peru 19906/ 1991-92 -7.5 -2.6

Souces : B u tm an and Le lde rm an (1993), C orbo and So Ilm ana (1,~1), 01 Te lla and D om busch (1989),

K lgue landLM a tan (1989 ). Lem grw e r(1977 ). M ede Iros (1993), R W lO S (1986), Wana (1990 ),

na tiona l sou rces , and Fw d s ta ff es tim a les .

1 / O Ve ra ll ba lance o f the non -financ Ia l pub lIc sec to r, u r"le ss o thuvJse Ind Ica ted . A m Inus s Ign

Ind ica tes a de fic it.

21 C a lenda r yea rs duJ ing v .h lch p rog ram w as In e ffed . If a p rogm m sta rted ea rty

In the yea r. tha t yea r Is a lso Induded . R gu res repo rted up to one yea r a fte r

the p rog ram ended . D o ts Ind ica te da ta a re no t avaH ab le .

3 / Fede ra l govem m eri .

.4/ D ura tion o f the p rog ram has been a rb ltra rftiy se t to five yea rs .

Table 8. Causality and Exchanpe Rate Regimes

Plan and Sample Period

Inflation EqualJon

F-statistic Probability value

EXchange Rate EqualJon

F-statistic Probability value

Exchange Rate-Based Stabilization Plans

Argentina

1967 Plan: 1965:1 - 1972 :6 "'v

Inflation 1.257 (0.287) 1.039 (0.407)

Exchange Rate 0.980 (0.444) 0.380 (0.889)

Tablita: 1978:12 - 1981:2

Inflation 0.772 (0.926) 0.255 (0.975)

Exchange Rate 3.398 (0.051) 0.026 (0.399)

Convertibility Plan: 1991:4 - 1994 :2v

Inflation 8.021 (0.001) 0.547 (0.654)

Exchange Rate 2.160 (0.115) 4.761 (0.008)

Chile

Tablita: 1978:2 - 1982:5

Inllation 26.283 (0.000) 1.480 (0.232)

Exchange Rate 4.765 (0.006) 3.706 (0.018)

Israel

1985:7 -1994:2

Inflation 10.509 (0.000) 6.193 (0.000)

Exchange Rate 2.655 (0.067) 3.449 (0.011)

Mexico

1987:12 - 1994:2

Inflation 6.988 (0.000) 1.324 (0.274)

Exchange Rate 4.533 (0.006) 4.544 (0.006)

~1968 Plan: 1966:1 -1973 :12 11

Inflation 1.565 (0.203) 1.579 (0.200)

Exchange Rate 5.635 (0.001) 2.337 (0.079)

Tabrrta: 1978:10 -1982:10

Inllation 7.264 (0.000) 0.723 (0.634)Exchange Rate 2.278 (0.057) 44.479 (0.000)

1991:1-1994:2

Inllation 0.568 (0.641) 2.139 (0.115)Exchange Rate 2_349 (0.092) 1.250 (0.309)

Memo item:

Chile

Real Exchange Rate Target; 1985:7 -1991:12

Inllation 1.968

Exchange Rate 0.248

Foreign Inllation 1.054

(0.127)

(0.862)

(0.374)

3.880

0.632

2.266

(0.013)

(0.597)

(0.088)

Money-Based Stabilization Plans

Chile

1975:4 -1977:12

Inllation

Money (M1)

28.036

2.793

(0.000)

(0.060)

0.515

1.334

(0.676)

(0.284)

Dominican Republic

1990:8 - present

. Inflation

Money (Ml)

0.367

0.843

(0.587)

(0.938)

7.0661.158

(0.000)

(0.371)

Notes: The optimal lag length was chosen according to the Schwarz: cJiteria Details are given in the appendix.

!/]hese plans fixed the exchange rate, hence for the duration of the plan it mimics the constant temJ of the regression

and a serious co/linearity problem emerges. To attemt to capture (; period were the exchange rate eJdJibitssome

variability,we broadened the sample adding about two years of obmvations at both ends. This was not

feasible for the case of the convertibility plan, since the period imm<1diatelyduring preceeding it a money-bBsed

plan was In effect (sea Table 1)

'11-1lentho sample Is extended further back (to 1960:1), there is a ~ignificant causal relationship from the exchange

Table 6. Inflation Stabilization and Growth:

Evidence from Panel Data

(1964 - 1993)

Dependent

Variable:

Constant Stabilization Plans Dum mies

(M aan) Exchange rate

E a,r1y Lata

Money

Early Late

Annual Growth Rate

in Real G DP

3.645

(0.518)

2.331

(1.037)

-5.454

(1.631)

-5.403

(1.931 )

1.822

(2.909)

N o tes : The coun trie s inc luded in · the sam p le a re A rgen tina . B raz il. C h ile . D om in ican R epub lic .

M ex ico . P e ru , and U ruguay . hence the re a re 210 obse rva tions . S tanda rd e rro rs a re in

pa ren theses . The random e ffec ts es tim a tes a re repo rted .

Table 7. Output Effects of Exchange Rate-Based

Stabilization Plans as the Duratic,o' of the Plan Varies

Level of Real G DP at the end of year:

Beginning

of th e

End

of th e

program 2 3 4 5 pro gra m

100

100

100

100

100

102.331

102.331

102.331

102.331

102.331

104.716

104.716

104.716

104.716

96.7498

107.157

107.157

107.157

99.0051

109.655

109.655

101.312

112.211

103.674

106.091

N o tes : These ca lcu la tions re fle c t the d iffe rence in ou tpu t due to the s tab iliza tion p lan and

a re based on the es tim a tes repo rted In Tab le 6 .

CHA ..R~T 1

Exchange Rate-Based StabilizationOrthodox Programs 12-Month Inflation Rates

ARGENTINE TAB UTA CHILEAN TABUTADecember 1978 February 1978

180 60

170

160

150 ~O140

130 W

120

110W

100 lO90

BOO o .~4 8 12 16 20 24 0 4 8 12 16 20 24 28 32 36 40 44 48 52

Months after beginning oC stabilization Months after beginning of stabilization

URUGUAY 1968 URUGUAYAN TABLITAJune 1968 October 1978

200 "0

180 m160

~O140

120({)

100 ~Of80'0

60::0

40

20 :10

°0 36]0 -

12 16 20 24 28 32 36 40 444 8 12 16 20 24 28 32 40 0 4 8

Months after beginning oC stabilization Months after beginning oC stabilization

ARGENTINE CONVERTIBILITYApril 1991

URUGUAYAN PLANJanuary 1991

300

4 8 U ~ ~ ~ nMonths af!er beginniot: of stabilization

1'0 --

1:0 \

1;0

110

](0

~O

10

,0

(0

~O

400 7 ~V n· ~

Months aCter ber:innint: of stabilization

Notes: The date for the beginning of the plan is giver above each graph. The 12-month inflation

. rates are plotted for the duration of the progr am.

CHART

Exchange Rate-Based StabilizationHeterodox Programs: 12-Month Inflation Rates

AUSTRAL PLAN

June 1985

1200

1000

800

600

400

200

o 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Months after ~inning of slabilization

CRUZADOFebruary 1986

250

200

150

100

50 01 2 3 4 567 8 9

Months after beginning of slabilization

ISRAELI PLAN MEXICAN PLANJuly 1985 Deccmbc:r 1987

450 180

400 160

350 140

300 120

250 100

200 80

150 60

100 40

50 20

°0 7 14 21 280

35 42 49 56 0 7 14 21 28 35 42 49 56

Months after beginning of stabilization Months after beginning of stabilization

Notes: The date for the beginning of the plan is given abov(: each graph. The 12-monlh inflation

rates are plotted for the duration of the program. F<'r the ongoing plans (Le. the Israeli

and Mexican plans) an arbitrary cut off of five years .vas chosen as the "end" of the plan.

CHART

Money-Based Stabilization12-Month Inflatior Rates

BONEX (ARGENTINA) COLLOR (BRAZIL)December 1989 March 1990

25 7000

206000

~5000.•

-a15c

'" 4000.•'"0

10E,3000

5 2000

010

1000(J 1 2 3 4 5 6 7 8 9 100 2 4 6 8 12


DOMINICAN REPUBLIC PERUAugust 1990 August 1990

120 14

100 12 \

~ 10.•-ac 8~'"0 6

40 t:4

20 2

°0 5 10 15 2(J 25 30 35 00 5 10 15 2(J 25 30 35 40

Months after beginning of stabilization Months after N:ginning of stabilization

Notes: The date for the beginning of the plan is given abo're each graph. The I2-month inflation

rates are ploUed for the duration of the program.

CHART 1

Exchange Rate- Based ~tabilizationOrthodox Programs: Real Effecti rc Exchange Rates

ARGENllNE TABUT A

December 1978=100

Cf-IILEAN TABLITA

February 1978=100

4 8 12 16 20 24

Months after beginning or stabilization

110

105/100

95

90

85

80

75

70

650 4 3 12 16 20 24 2S 32 36 40 44 48 52

Me nlhs after beginning or stabilization

100

95

90

85

SO

75

70

65

60

55

500

URUGUAYANTABLITA

October 1978=100

AP GENTINE CO NVERTIB rrnYApril 1991=100

110 100

95

4 8 12 16 20 24 2S 32

Me nths arter beginning or stabilization

90

85

60 80

500 4 8 12 16 20 24 2S 32 36 40 44 48


URUGUAYAN PLlli

January 1991 =100

100

95

90

85

80

75

700 7 14 2t 28

Months .ftef beginning of sub liution

Notes: A decrease in the index indicates real appreciation. Tlle date for the beginning of

the plan is given below each graph. The real exchange ~ates are plotted for the

duration' of the program.

CHARTz

Exchange Rate-Based StabilizationHctcrodox Programs: Real Effcci ivc Exchange Rates

AUSTRAL PLAN CRUZADO

June 1985=100 February 1986 = 100

104 112

110

r100

108

98106

96

10494

92 102

901 2 3 4 5 6 7 8 9 1011 121314 15

1000 12" 3 4 5 6 7 8 90

Months after beginning or stabilization Months after beginning of stabilization

ISRAELI PLAN

July 1985=100

104

102

100

98

96

94

92

90

88

860 7 14 21 28 35 42 49 56


MEXICAN PlAN

December 1987=100

100

95

90

85

80

75

·70

65

600 7 14 21 28 35 42 49 56

Months after beginning of stabilization

Notes: A decrease in the index indicates real appreciatillO. The date for the beginning of

the plan is given above each graph. The real excl:1ange rates are plotted for the

duration of the program. For the ongoing plans ~i.e. the Israeli and Mexican plans)

an arbitrary cut off of five years was chosen as th.: "end" of the plan.

CHART 2

Money-Based Stabili~:ation ProgramsReal Effective Exchange Rates

BONEX (ARGENTINA) COLLaR (BRAZIL)December 1989=100 March 1990=100

130 lGO

120

no 150

100 14090

80 130

70

60120

50 llO40

306 8 10 12

1000 1 2 3 4 5 6 7 8 90 2 4

Months a(ter beginning o( wbilizalion Months after beginning o( stabilization

DOMINICAN REPUBLIC PERUAugust 1990=100 August 1990=100

100 140

98

1~ r%

94 120

92 11090

88 100

86 90

84

8280

800 5 10 15 20 25 30 35 40

700 5 10 15 20 25 30 35 40

Months a(ter beginning o( stabilization Months after beginning o( stabilization

Notes: A decrease in the index indicates real appreciati on. The date for the beginning of

the plan is given above each graph. The real excllange rates are ploued for the

duration of the program.

CHAF~T J,

Exchange R a te -B a s e d S ta b i l iz a t io n

Orthodox Programs: Tola! Reserves Minus Gold

ARGENTINA (1967)March 1967= 100

450

400

4 8 U M m M m ~ ~M on th s a C ta - b e g in n in g o f s ta b i l iz a t io n

URUGUAY (1968)June 1968=100

CHILEAN TABLITAFebruary 1918=100

8 U U m M m II ~ ~ « • IIMonths .C,er beginning of stabtlization

ARGENTINE CONVERTIBILITYApril 1991=100

U 16 m 24 m 32

M on th s a f t e r ~gillnin&oC s 1 3 b t l iz .3 t . io a

BRAZIL (1964)March 1964= 100

550

500

450

400

350

300

250200

150100 •..•..

~O 4 8 U U m M m II ~ ~ « • IIM on th s a f t e r b e g in n in g o f s ta b i l iz a t io n

ARGENTINE TABLITADecember 1918=100

200

180

160

1~

1m

100 ..-

80

600 8 U M m M

M on th s a f t e r b e g in n in g o f s ta b i l iz a t io n

URUGUAYANTABLITAJanuary 1918=100

1~

1m

100 V80

60

~

mo 4 8 U H m ~ m II ~ ~

Months aCta beginning of st.abtlization

URUGUAYAN PLANJanuary 1991=100

180

160

1~

1m

14 21 mMonth •• C'e< beginning of stabilization

35

Notes: The date for the beginning of the plan is given abov( each graph. Total Reserves Minus

Gold data is plotted for the duration of the program.

----------------------------------------

.'

CHAR,T'~,3

Exchange Rate-Based StabilizationHeterodox Programs: Total Reserves Minus Gold

AUSTRAL PLAN CRUZADOJune 1985=100 February 1986=100

260 11

240 105

220 100

200 95

180 90

160 85

140 80

120 75

100o 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

700 1 2 3 4 5 6 7 S 9

Months after beginning of stabilization Months after beginning of stabi1iz:ation

ISRAELI PLAN MEXICAN PLAN

July 1985=100 December 1987=100

260 1

240 140220

200120

180 tOO

160 80

14060

120

100 40

80 2000 7 14 21 28 35 42 49 56 7 14 21 28 35 42 49 56


Notes: The date for the beginning of the plan is given aX>ve each graph. The Total Reserves

Minus Gold data is plotted for the duration of tlIe program. For the ongoing plans

(i.e. the Israeli and Mexican plans) an arbitrary CJt off of five years was chosen as

the "end" of the plan.

CHARrrl3

Money-Based Stabilization ProgramsTotal Reserves Minus Gold

CHILEAN PLANApril 1975= 100

700

600

500

400

300

200

100

°02468WUMMWWll~~W~TIMonths after beginning of stabilization

coLLaR (BRAZIL)March 1990=100

170

160

150

140

1~

120

110

1001 2 3 4 5 6 7 8 9


500

450

400

350

300

250

200

150

100 0

BONEX (ARGENTINA)December 1989=100

350

250

30(i

2O<J

150

100

50 0 2 4 6 8 10 12


DOMINICAN REPUBLICAugust 1990=100

450

400

350

300

250

200

150-

100-

50-

o·o 5 10 15 20 25 30 35 40


PERUAugust 1990=100

5 10 15 20 'B' 30' 35


,I

40

Notes: The ~ate for the beginning of the plan is given above each graph. The Total Reserves

Minus Gold data is plotted for the duration of the program.

CHART', 4

Exchange Rate-Based S ta b i l iz a t io n

Orthodox Programs: Stock Market Price Indices

ARGENTINE TABLITADec<:mbcr 1978=100

700

600

500

400

300

200

<4 & 12 16 20 24

Months after beginning of 6tabilization

CHILEAN TABLITAFebruuy 1978 = 100

350

300

250

200

150

l00 ....•~

500 7"7'12 16 20 24 28 32 36 .w « 4S 52

Months after beginning of ltabilization

ARGENTINE C O N V E R T IB IL IT Y

April 1991=100

350

300

250

200

150

100

50 0 4 8 12 16 '"~ 28 32

Months after beginning <:>fstabilization

Notes: The date for the beginning of the plan is given above each graph. Ibe stock market price

indi~ are plotted for the duration of the program.

CHART 4

Exchange Rate-Based StabilizationHeterodox Programs: .

AUSlRAL PLAN CRUZADOJune 1985=100 Febru~ry 1986=100

350 200

190

300 1&0

170

250 160

150

200 140

no150 120

110

1000 1 234 5 6 7 8 9 10 11 1213 14 15

1001 2 3 4 5 6 7 8 9

Months after beginning of stabilization Moc,ths after beginning of stabiliutioo

MEXICAN PLANDecember 1987=100

1100

1000

900

SOO

700

600

500

400

300

200

100 0 4 8 12 16 20 24 28 32 36 40 « 48 52 56

Months after beginning of stabili~tion

Notes: The date for the beginning of the plan is given above each graph. The stock man::et price

indi~ are plottod for the duration of the program. For ahe ongoillg plans an arbitrary cut off

of five years was chosen as the "end" of the plan.

CHART 4

Money- Based StabilizationStock Market Price Indices

BONEX (ARGENTINA) COLLaR (BRAZIL)December 1989=100 Much 1990=100

100 1~

15090

1<W

to 1~

12070

110

~ 100

9050

to

<W0 4 IS 10 12

704 IS

Month& after bePnnin& or nabiliu.tion Month& after bePnnin& of mbiliu.tion

PERU (1990 PLAN)August 19'90 = 100

lto

170

1~

150

1<W

1~

120

110100)

90

to 0 4 ,-. 12 1 IS 20 24 21~ n 36 <W 44

Month& after be&i=in& or nabifuation

Notes: The date foc the beginning of the plan is given above each graph. The stock market price

indices an: plotted for the duration of the proram.

Chart 5. Exchango Ratos and Inflation: Programs of tho 1960s·

Effect of Temporary 20% Reduction in tho Rate of Devaluation

Percent

Ar entina 1967 Plan2

o-2

-4

-6

-6

-10

-12

-14

·16

-16

-20

Sample period:

1965:1-1972:6

o 4 10 22 246 12 14 16 202 6 16

Number of Mon:ths

Percent

2

o

-2

-4

-6

-B

-10

-12

-14

-16

-16

-20

Uru uay 1968 Plan

Sample period:

1966:1 -1973:12

o 2 4 6 6 10 12 '14 16 18 2~ 22 24 26 28 30 32 34 36

Number of Months

Notes: Temporary refers to a chl'lnge lasting 12 months. Results of the estimated

equations aro summarized in Tables 8-9.

Sources: International Financial Statistics, IMF and the authors.

Chart 7. Exchange Rates and Inflation: Programs of the 19808 and 19908

Effect of Temporary 20% Reduction In the Rate of Devaluation

Percent Percent

Israeli Plan Mexican Plan4

2

o·2-4

-$

.a·10

·12

·14

·16

·18

·20o

Sampleperiod:

1985:7 ·1994:2

Sample period:

1987:12 ·1994:2

2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 o 2 4 6 8 10 12 14 16 18 20 22 24

Number of Months Number of Months

P e r c e n t Percent

Argentine Convertlbillt Plan Uru 1991 Plan

Sampleperiod:

1991:4 ·1994:2

Sampie period:

1991:1 ·1994:2

o 2 4 6 8 10 12 14 16 18 20 22 24 o 2 4 6 8 10 12 14 16 18 20 22 24

Number of Months Number of Months

N o tes : Tem po ra ry re fe rs to a change las tIng 12 m on ths . R esu lts o f the es tim a ted

equa tions a re sum m arized In Tab les 8 ·9 .

S ou rces : In te rna tIona l F inanc Ia l S ta tis tics , IM F and the au tho rs .

Chart 8_ Monoy and Inflation

Effoct of Tomporary 20% Reduction in tho Rate of Growth of Ml

Percent

2

o

-2

--4

-{l

-{l

-10

-12

-14

-16

-18

-20

Sample period:

1975:4 -1977:12

o 2 10 12 14 16 18 20 22 24 26 2S6 e

Number" of Monlhs

"2o-2-4

-{l

-{l

-10

-12

-14

-16

-18

-20

Sam~pcriod:

1990:8 - 1993:10

Number of Months

Notes: Temporary refers to a change/asting 12 months. Results of the estimated

equations are summarized in Tables 8-9.

Sources: International Financial Statistics, /MF slid the authors.

inﬂation stabilization in chronic inﬂation countries: the ... · the capital account improves...

Documents