Transcript Macro Conference IV

Chapter 18
Trade, Financial,
and Exchange Market Reforms
© Pierre-Richard Agénor and Peter J. Montiel
1



Reduction in trade barriers fosters
 adjustment in relative prices, and
 reallocation of resources toward the sector producing
exportables.
In the long term, successful trade liberalization leads to
 expansion of exports;
 contraction of activity in import-competing industries;
 transfer of resources from sectors producing
nontradables toward those producing tradables.
More open trade regime may be associated with higher
long-term rates of economic growth.
2




Trade Reforms: Some Recent Evidence.
Trade Liberalization, Wage Rigidity, and Employment.
Monetary and Financial Liberalization.
Unification of Foreign Exchange Markets.
3
Trade Reforms: Some Recent
Evidence


Papageorgiou et al. (1990): trade liberalization in 19
countries, during 36 episodes of reform undertaken in
developing countries between World War II and 1984.
Successful liberalization episodes have been
characterized by their
 comprehensive nature;
 systematic elimination of quantitative restrictions on
imports;
 political stability;
 prudent macroeconomic policies;
 maintenance of restrictions on capital movements
until trade reforms were sufficiently advanced;
 significant initial depreciation of real exchange rate.
5






There is no strong evidence suggesting that
liberalization was associated with sharp reductions in
employment and short-run contraction in output.
Balance of payments improved significantly in most
cases, as growth of exports outpaced growth of imports.
Since the mid-1980s, far-reaching trade reforms have
been implemented in the developing world.
Prior to reform, extensive barriers to trade were in place
in most of these nations.
Particularly interesting set of episodes took place in
Latin America and the Caribbean.
In relation to different macroeconomic conditions, extent
and timing of reforms differed across these countries.
6
Peru:
 Trade regime prior to reform was severely distorted.
 Import policy reform:
 dismantling nontariff barriers such as quantitative
restrictions and foreign exchange controls; and
 eliminating widespread exemptions.
 Export policy reform:
 reduction or elimination of price and quantitative
barriers to exports; and
 introduction or improvement of incentives for export
promotion and diversification.
Table 18.1:
 Indicators of trade regime before and after most recent
reforms for 15 Latin American and Caribbean countries.
7




Average tariff rates were reduced dramatically.
 In Brazil, number of tariff rates was reduced from 18
prior to reform (in 1990) to 9 by the end of 1993.
 In Colombia, it fell from 22 in 1990 to 4 in 1992.
Dispersion of tariffs was also reduced, particularly in
Colombia, Costa Rica and Ecuador.
Degree of openness increased significantly due to
expansion of both exports and imports in real terms.
Many of reforms in Latin America were introduced in
difficult macroeconomic context:
 high inflation;
 sluggish economic activity;
 balance-of-payments difficulties;
 scarce foreign exchange reserves.
8


Almost all the episodes of trade reform were preceded
by, or associated with, significant depreciation of real
exchange rate.
Figure 18.1: continuous real depreciation and
expansion of export volumes in aftermath of recent
trade liberalization in Chile.
9
F
i
g
u
r
e
1
8
.
1
C
h
i
l
e
:
T
r
a
d
e
V
o
l
u
m
e
s
a
n
d
t
h
e
R
e
a
l
E
x
c
h
a
n
g
e
R
a
t
e
,
1
9
8
0
9
6
2
1
0
0
0
1
4
0
R
e
a
l
i
m
p
o
r
t
s
(
l
e
f
t
s
c
a
l
e
)
1
8
0
0
0
1
2
0
1
5
0
0
0
R
e
a
l
e
x
c
h
a
n
g
e
r
a
t
e
(
r
i
g
h
t
s
c
a
l
e
)
1
0
0
1
2
0
0
0
R
e
a
l
e
x
p
o
r
t
s
(
l
e
f
t
s
c
a
l
e
)
8
0
9
0
0
0
6
0
6
0
0
0
4
0
3
0
0
0
0
1
9
8
0
2
0
1
9
8
2
1
9
8
4
1
9
8
6
1
9
8
8
1
9
9
0
1
9
9
2
1
9
9
4
1
9
9
6
S
o
u
r
c
e
:
I
n
t
e
r
n
a
t
i
o
n
a
l
M
o
n
e
t
a
r
y
F
u
n
d
.
N
o
t
e
s
:
R
e
a
l
e
f
f
e
c
t
i
v
e
e
x
c
h
a
n
g
e
r
a
t
e
,
b
a
s
e
1
0
0
=
1
9
9
0
(
a
r
i
s
e
i
s
a
d
e
p
r
e
c
i
a
t
i
o
n
)
.
R
e
a
l
i
m
p
o
r
t
s
a
n
d
e
x
p
o
r
t
s
a
r
e
m
e
a
s
u
r
e
d
F
O
B
,
i
n
m
i
l
l
i
o
n
s
o
f
U
.
S
.
D
o
l
l
a
r
s
,
a
t
1
9
9
0
p
r
i
c
e
s
.
10
Trade Liberalization, Wage
Rigidity, and Employment




There are potential channels through which trade
reforms may lead to contractionary effects in the short
run.
Such costs may have adverse effect on sustainability of
adjustment process, leading to sudden policy reversals
or complete abandonment of reform efforts.
Buffie’s (1984b) model is used to study output and
employment effects of trade liberalization.
Key implication: trade reform raises relative price of
imported inputs and, if wages are not perfectly flexible,
can lead to a contraction in output.
12

The Analytical Framework.
 Fixed Nominal Wages.
 Flexible Wages.
13
The Analytical Framework




Economy produces tradable and nontradable goods
using three factors of production: capital, labor, and
intermediate input, which is not produced domestically.
Capital stock is sector-specific and fixed.
Imported inputs are not subject to tariffs and their
domestic price PJ is equal to nominal exchange rate E.
World price of traded goods is fixed on world markets,
and domestic price is
PT = (1+)E,
 > 0 measures the extent to which import tariffs and
export subsidies have raised domestic price of tradable
14
good above its world market level.


Each sector's output is produced by competitive firms
operating with a constant-returns-to-scale technology.
Sectoral factor demands:
h, E),
(2)
Kh = yh(Lh, Jh, Kh)Ch(w, h, E),
(3)
h, E),
(4)
Lh =
Jh =
h
yh(Lh, Jh, Kh)Cw(w,
h
yh(Lh, Jh, Kh)CE(w,
h = N, T refers to production sectors;
Yh: sector h's output; Ch: sector h's unit cost function;
w: nominal wage; : rental rate on capital in sector h.
15


Let Ph denote final prices in sector h.
Zero-profit condition yields
Ph = wCwh + hCh + ECEh.



(5)
Equations (2) to (5) can be solved for output and factor
demand levels and rental rate of capital, as a function of
the nominal wage, price of imported inputs and price of
final goods.
Households consume both categories of goods and
hold only domestic money in their portfolios.
Assume homothetic preferences and utility function
separable in goods and money.
16

Relative sectoral demand functions can be defined as
homogeneous functions of degree zero in prices:
Dh/A = dh(PN, PT),

0  dh  1,
A: aggregate nominal expenditure.
It is defined by
A = PNyNs + PTysT – EJ - (M - Md),

(6)
(7)
J: total imports of intermediate goods.
(7): A is equal to difference between net factor income
and hoarding, which depends on difference between
actual (M) and desired (Md) money balances.
17

Desired balances depend on net factor income:
Md = PNysN + PTyTs - EJ.

(8)
Nominal wages are indexed to the price level, which is
defined as a geometrically weighted average of prices
of traded and nontraded goods:
w=
w(PNE1-),
~
w > 0, 0 < , < 1. (9)
~
18

Market-clearing condition for the nontraded goods
market can be written as
s
yN
= AdN(PN, PT),
(10)
which determines price of nontraded goods.
19

Logarithmically differentiating Equations (6)-(8) and
simplifying yields, in terms of rates of change,
^ ^ h h
h
h
h - h )
Lh = hK(LK - KK) + J(LJ
KJ
^ h ( h -  h ),
+ w
L LL
LK
(11)
h
 j:

share of factor j in sector h's total production costs;
ij: partial elasticity of substitution between i and j;
: devaluation rate.
Using (5) to substitute out for ^h in (11) and rearranging
yields:
+ ? _
+ _ ?
^
Lh = Lh(h, , w),
^
Jh = Jh(h, , w). (12)
20



Sign of cross-price terms is indeterminate, due to
possible divergence between output and substitution
effects.
Since factor demand functions are homogeneous of
degree zero, sum of all partial derivatives in each
equation is equal to zero.
From (9), rate of change of nominal wages:
^ = [ + (1-) ].
w
N
T

(13)
Equations (7)-(10), (12) and (13) can be solved
^ and  as a function of  or
simultaneously for L^h, J^h, w,
N
T
.
21
Fixed Nominal Wages




Assume: nominal wages are fixed ( = 0) and imported
inputs are used only in the tradables sector.
Trade liberalization: lowering of import tariffs and export
subsidies coupled with nominal devaluation.
Assume: reduction in tariffs is less than rate of
devaluation, implying increase in domestic price of
tradables and thus of imported inputs.
If nominal wages are fixed, effect of the trade reform
program on employment in traded goods sector is
indeterminate:
 net devaluation leads to fall in product wage, which
stimulates demand for labor;
22
liberalization also leads to increase in relative price of
imported inputs.
Whether net effect is positive or negative depends on
 pattern of substitution across factors, and
 effect on gross output.
If production function in traded goods sector is
separable between primary factors and imported inputs,
condition that determines whether employment in traded
goods sector rises or falls is given by



T - JT > 0,
<
JT: share of imported inputs in production costs in
tradables sector.
23



If production function is not separable and capital and
imported inputs are better substitutes than labor and
imported inputs, condition (14) is not sufficient to
guarantee increase in labor demand and employment.
Condition for employment to rise following trade
liberalization is that expansionary, expenditure-switching
effect dominates
 “hoarding effect” induced by relative price change;
 expenditure-reducing effect induced by negative
income effect;
 potential fall in employment in traded goods sector.
If cross-price elasticity of demand for nontraded goods is
large enough, net effect of liberalization is expansion of
employment in nontraded goods sector.
24


If cross-price elasticity of demand for nontraded goods is
small, net effect of trade reform on employment is likely
to be negative.
Result: overall effect of liberalization on employment is
ambiguous in an economy with immobile capital in the
short run and nominal wage rigidity.
25
Flexible Wages




Assume: wages are free to adjust to changes in prices,
and only traded goods sector uses imported
intermediate inputs.
Effects of trade liberalization become even more
uncertain.
Product wage in traded goods sector may rise, thus
further reducing demand for labor in that sector.
In the nontraded goods sector, employment may also fall
to the extent that
 negative income effect associated with lower
employment in traded goods sector;
 possible increase in product wage compensate for
expenditure-switching effect of liberalization.
26



Result: overall decline in real consumer wage may be
associated with increased unemployment.
Buffie (1984b):
 In a large number of cases, contractionary effects on
employment may be large enough to offset the
expansionary effects.
 Real wage rigidity increases short-run contraction in
employment, unless there is very high degree of
substitutability between labor and imported inputs.
Despite of short-run adjustment costs, trade
liberalization may be beneficial in the long run, since
elasticities of substitution between inputs are higher than
in the short run.
27
Monetary and Fiscal
Liberalization



Restrictions on competition in banking industry: barriers
to entry into banking system and public ownership of
banks.
Restriction on composition of bank portfolios:
 requirements that banks engage in certain forms of
lending;
 imposition of “liquidity ratios,” requiring banks to
invest specified share of their portfolios in
government instruments;
 directed lending to specific productive sectors;
 prohibitions from acquiring other types of assets.
Park (1991):
 Draws distinction between monetary reform and
financial liberalization.
29
Monetary reform: increase in controlled interest rates
to near-equilibrium levels, with remaining set of
restrictions on behavior of banks left in place.
 Financial liberalization: more ambitious set of reforms,
directed at removing at least some of remaining
restrictions on bank behavior.
 Full financial liberalization involves
 privatization of public financial institutions;
 removal of restrictions to entry into banking;
 measures aimed at spurring competition in
financial markets;
 reduction of legal reserve requirements;
 elimination of directed lending;
 freeing of official interest rates.
30



Monetary Reform.
Financial Liberalization.
31
Monetary Reform



Arguments for monetary reform as a structural policy
conducive to a higher growth path are due to McKinnon
(1973) and Shaw (1973).
When saving instruments in formal financial system are
limited to cash, demand and time deposits, raising
interest rates to near-equilibrium levels may induce
 increase in saving rate;
 portfolio shift out of inventories, precious metals,
foreign exchange, and curb market lending into formal
financial system.
High real interest rates resulting from the reform would
increase investment either
32
because the need to accumulate funds to undertake
lumpy investments makes money and capital
complementary rather than substitute;
 or because of “credit availability” effect.
Many high-return projects not previously funded would
be undertaken after the reform.
Reason: banks have scale economies relative to
informal market in collecting and processing information
on borrowers.
Proposition:
 Raising controlled interest rates should raise demand
for domestic time and saving deposits.
 Rate of growth increases because
 increase in saving raises investment;
33
 quality of investment improves.




Vogel and Buser (1976):
 Under financial repression, higher rate of inflation should
reduce private investment, since it would be associated
with lower real interest rates.
 But, they found little evidence of this effect.
 Inflation significantly affected demand for time and
saving deposits.
 Investment level depended positively on rate of
accumulation of time and saving deposits : supportive of
complementarity and credit availability hypotheses.
Galbis (1979):
 Tested McKinnon's direct complementarity hypothesis.
 Found little support for the proposition.
34
Unable to find negative effects of inflation on investment
levels.
Lanyi and Saracoglu (1983):
 increases in real deposit interest rate increased national
saving rates;
 real deposit interest rate had positive and significant
coefficient in a regression of growth rate on deposit rate.
Fry (1996):
 Used pooled cross-section-time series regressions for
several samples of Asian countries.
 Found weak positive effect of real deposit rates on
national saving, but strong positive effect on demand for
money and on supply of credit.
 Thus, credit supply was found to have strong positive
35
effect on investment.

No evidence was found for McKinnon's
“complementarity” effect.
 Found two types of evidence in support of improved
quality of investment:
 Real deposit rates were positively correlated with
ICOR (proxy for efficiency of investment).
 Real deposit interest rate had positive effect on
growth.
Gelb (1989):
 Investigates whether positive correlation between real
deposit rates and growth is consistent with causation
 from growth to interest rates, or
 from common third factor.
 Sample of 34 developing countries during 1965-85.
36






Real deposit rates had strong positive effects on both
growth and ICOR, but weaker positive effects on
investment ratio.
Inclusion of additional variables weakened effects of
interest rates on investment, but not on ICOR.
Thus, efficiency effect on investment, and not effect on
overall volume of investment, accounted for positive
relationship between real interest rates and growth.
Inclusion of other measures of distortion weakened
relationship between real interest rates and growth,
although it remained positive.
Causal chain between real deposit rates and growth
through more effective intermediation into higherproductivity investment:
37
increase in real deposit rate increased share of
domestic saving intermediated through formal
financial system;
 this share had stronger effect on growth than level of
saving itself.
Dornbusch and Reynoso (1993):
 Alternative interpretation of such results.
 In neoclassical growth models, growth of output per
capita can be written as

y^ – n = [(I/y)(K-1/y)-1 – n],
y: real output; K: capital stock; I: net investment;
n: rate of growth of the labor force; : share of capital.
38
What matters for contemporaneous growth is average,
not marginal, efficiency of capital, and this is improved
slowly through improved efficiency of investment.
 Their interpretation of correlation between real deposit
interest rates and growth is that high inflation impedes
growth through distortions induced by uncertainty.
Evidence on effects of monetary reform on growth is
inconclusive.
 Higher real deposit interest rates are unlikely to have
strong effects on saving rate.
 Even though portfolio shifts toward domestic financial
instruments are induced by monetary reform, this may
not have large effect on volume of investment.

39




While there is little evidence in favor of
“complementarity” effect, “credit availability” effect is
more strongly supported by data.
Interpretation of positive correlation between real deposit
rates and growth is problematic beacuse
 it may reflect some contribution of efficiency effect;
 real deposit rate may be serving as a proxy for more
general distortions.
Episodic evidence associated with specific-country
cases of monetary reform does not provide clearer
verdict because ceteris paribus conditions do not hold.
Consider Korean monetary reform of 1965.
40
McKinnon (1976):
 Nominal deposit and lending rates had been pegged at
low levels in Korea prior to the reform, yielding negative
real rates in 1963-64.
 Nominal rates were revised upward, in September 1965,
and directed credit restrictions were reduced, thus
qualifying this episode as monetary reform.
 Real rates of return rose subsequent to the reform.
 Ratio of broad money to GDP increased by factor of 7
between 1964 and 1969.
 Private saving increased, and growth experienced
strong acceleration.
 McKinnon interpreted this as supporting positive effect of
monetary reform on growth.
41
Giovannini (1985):
 Reached different conclusions.
 Most of the increase in national saving in Korea arose in
public sector due to fiscal correction.
 Measured increase in households' surplus after the
reform was one-shot event concentrated in 1966.
 Correlation between surplus and real interest rate was
negative after that year.
 Measured increase in saving may have been due to
recording of portfolio shift out of informal market as a
change in saving.
42
Financial Liberalization
Evidence on effects of financial liberalization is episodic.
Villanueva and Mirakhor (1990):
 Success in financial liberalization requires
macroeconomic stability and strong and effective system
of bank supervision as preconditions.
 Success is more likely if controls on interest rates are
removed gradually.
 Sri Lanka and Korea moved to financial liberalization
gradually, while requisite preconditions were put in
place.
 Without these conditions, full financial liberalization is
apt to be associated with
43
 sharp increases in real interest rates;

bankruptcy of financial institutions;
 loss of monetary control.
 Early examples of these outcomes are Southern Cone
liberalizations, Philippines and Turkey.
Argument is as follows:
 Macroeconomic instability increases variance of and
covariance among projects funded by banks.
 This increases riskiness of bank portfolios.
 If deposit insurance is absent or correctly priced, banks
would reduce interest rates and ration credit more
severely.
 But with inadequately priced deposit insurance, moral
hazard will induce banks to raise interest rates to attract
deposits and fund high-risk projects.

44
Reason is that they face one-way bet:
 if the projects pay off, bank owners reap the profits,
 if they do not, government pays off depositors, with
bank owners risking only their limited capital.
 This can be avoided when deposit insurance is priced
correctly, because
 this forces banks to pay for higher risk that their
portfolio choices impose on the government;
 So they internalize consequences of their actions.
 Same result could be ensured by adequate bank
supervision.
Villanueva and Mirakhor (1990):
 Contrasting experiences of Southern Cone countries,
Philippines, and Turkey versus that of Malaysia.
45

All of these countries moved to full liberalization of
interest rates in very short period.
Southern Cone countries:
 In Argentina, Chile, and Uruguay, removal of interest
rate ceilings and credit controls was accompanied by
 relaxation of bank supervision;
 extension of either explicit or implicit deposit
insurance;
 all in high inflation and unsatisfactory economic
performance.
 Financial liberalization measures were accompanied by
innovative macroeconomic stabilization programs in all
three countries.

46
Bank portfolios included unusual number of bad loans,
 impairing bank capital, and
 increasing moral hazard problems created by deposit
insurance.
 Lending rates rose to high real levels, distress borrowing
by firms ensued, and bankruptcies became common.
 Liberalization and stabilization programs collapsed in
midst of a financial crisis.
 Philippines and Turkey: Liberalizations in the 1980s
were carried out under similar circumstances and in
similar fashion and they produced similar results.
Malaysia:
 Although financial liberalization was carried out rapidly,
there were macroeconomic stability and banking
47
supervision.

Transition to liberalized financial system was smoother,
with mild increase in real interest rates and no
widespread bankruptcies.
Sri Lanka and Korea:
 Both undertook liberalization from initially unsatisfactory
macroeconomic performance.
 But Asian countries removed restrictions on interest
rates gradually while pursuing macroeconomic stability
and stronger prudential regulation over banks.
 Greater flexibility was permitted only after supervisory
mechanism strengthened and macroeconomic stability
was achieved.

48
Unification of Foreign
Exchange Market




Attempts at imposing exchange and trade restrictions in
developing countries have led to emergence of illegal
markets for foreign exchange.
Costs of this:
 high volatility of exchange rates and prices;
 creation of incentives to engage in rent-seeking
activities or divert export remittances from official to
parallel market;
 loss in tax revenue.
These costs have led policymakers to seek ways to unify
official and parallel markets for foreign exchange.
Exchange-rate unification has objectives as
 absorption and legalization of parallel market;
50
elimination of inefficiencies and market fragmentation
associated with quasi-illegal activity;
 thus permitting efficient allocation of foreign-currency
earnings.
Unification attempts have often taken form of adopting
uniform floating exchange rate.
Impact of such a policy shift on short- and long-run
behavior of exchange rate and inflation is ambiguous.



51



Short-Run Dynamics of Unification.
 The Pre-Reform Dual Rate Regime.
 The Post-Reform Flexible Rate Regime.
 Dynamics in Anticipation of Reform.
Longer-Run effects of Unification.
Evidence on Unification Attempts.
52
Short-Run Dynamics of
Unification

Lizondo (1987), Kiguel and Lizondo (1990), and Agénor
and Flood (1992): short-run effects of preannounced
future adoption of unified flexible-exchange-rate.
53
The Pre-Reform Dual Rate Regime






Small open economy operating informal dual exchangerate regime: official, pegged exchange rate coexists with
freely determined parallel rate.
Official rate applies to current account transactions
authorized by the authorities.
Parallel rate is used for capital account transactions and
remainder of current account items.
Agents are endowed with perfect foresight and hold
domestic- and foreign- currency balances.
Domestic output is exogenous single exportable good.
In each period, exporters
 surrender given proportion of their foreign exchange
earnings at official exchange rate, and
54

repatriate remaining proceeds via parallel market.
55

Model is described by the following log-linear equations:
.
m – p = -s,
m = R + (1-)d,
(15)
0 <  < 1,
(16)
p = s + (1-)e,
0 <  < 1,
.
.
R = -(s-e), (18)
d = 0,
(17)
(19)
m: nominal money stock; d: domestic credit;
R: stock of net foreign assets held by central bank;
p: domestic price level; e: official exchange rate;
s: parallel exchange rate.
56





(15): money market equilibrium.
(16): log-linear approximation that defines domestic
money stock as a weighted average of domestic credit
and foreign reserves.
(17): price level depends on official and parallel
exchange rates.
(18): behavior of net foreign assets.
 Negative effect of premium on behavior of reserves
results from its impact on under-invoicing of exports.
 The higher the parallel exchange rate is, the greater
will be the incentive to falsify export invoices and to
divert export proceeds to the unofficial market.
(19): stock of credit is constant over time.
57



In pre-reform dual exchange-rate system:
 forward-looking parallel rate s and predetermined
level of official reserves R are endogenous variables;
 official exchange rate e is set by the authorities.
In post-reform, unified flexible-rate regime s = e =  and
reserves remain constant.
With d constant, solving (15)-(19) yields
.
s
.
R
/ -/
=
0
-
s
R
+

e
(20)
where  = -(1-)d/ + (1-)e/.
58


Because determinant of the matrix of coefficients is
-/ < 0, system (20) is saddlepoint stable, with one
negative root (denoted by 1) and one positive root, 2.
Solving for the particular solutions yields
s = s~ + C1e-1t + C2e-2 t,
~
R = R + 1C1e-1t + 2C2e-2t,
(21)
1 > 0, 2 < 0,
(22)
C1 and C2: yet undetermined coefficients;
~
~
s = e and R = [e - (1-)d]/ > 0: steady-state values of
parallel rate and foreign exchange reserves.
59
If existing dual-rate system is expected to last forever:
 Stability would require setting C2 = 0 in (21) and (22).
 Using initial condition on reserves R0 allows
determination of C1.
 Economy's equilibrium path is unique nonexplosive path
SS depicted in Figure 18.2.
 For positive (negative) premium, reserves are falling
(increasing), as indicated by the arrows pointing west
(east) in the figure.
 Saddlepath SS has positive slope (-1) and is flatter
.
than the curve [s = 0].
.
 Increase in  rotates [s = 0] and SS clockwise.
.
 Increase in d shifts both [s = 0] and equilibrium point E
to the left.
60
F
i
g
u
r
e
1
8
.
2
S
t
e
a
d
y
S
t
a
t
e
E
q
u
i
l
i
b
r
i
u
m
P
r
i
o
r
t
o
R
e
f
o
r
m
s
.
s
=
0
S
E
_
.
R
=
0
s
*
=
e
S
R
*
S
o
u
r
c
e
:
A
g
é
n
o
r
a
n
d
F
l
o
o
d
(
1
9
9
2
,
p
.
9
3
0
)
.
R
61
Rise in  translates into clockwise rotation of SS.
 Devaluation of official exchange rate leads to upward
.
.
shift of [R = 0] and rightward shift of [s = 0].
 In the long run, devaluation leads to equiproportional
depreciation of parallel exchange rate and increase in
reserves.
If policymakers announce their intention to switch to
floating-rate arrangement at a well-defined date in
the future:
 Forward-looking agents will anticipate abandonment of
dual-rate system.
 Coefficient C2 will not be zero.
 C1 and C2 will be determined satisfy constraints imposed
by perfectly anticipated transition to post reform regime.

62
The Post-Reform Flexible Rate Regime




T > 0: future transition date announced at period t = 0.
In post-reform flexible-rate
. system  = e = s, condition
that yields, from (18), R = 0.
Therefore, reserves remain constant beyond t  T at
level RT+.
Under these assumptions, unified flexible exchange rate
is determined by
.
 -  = -mT+,
t  T,
(23)
where
mT+ = RT+ + (1-)d,
t  T.
63

(23): linear differential equation in , whose solution is
 = Cet/ + m+T,


Ruling out speculative bubbles requires setting C = 0.
Exchange rate in effect at the instant the economy
switches to unified floating-rate regime is
T+

t  T.
=
+
mT,
t  T,
(26)
which depends on reform date, since it is linearly related
to terminal stock of reserves at T.
Relationship between reserves and unified exchange
rate at T defines “terminal curve,” (denoted CC), whose
slope is equal to .
64
Dynamics in Anticipation of Reforms


How will reserves and premium evolve when agents
perfectly anticipate transition from dual-rate regime to
flexible system?
Two requirements to “connect” two regimes:
 initial condition on reserves R0;
 price continuity condition to prevent a jump in parallel
exchange rate at the moment reform is implemented.
~
R 0 = R 0,

sT = +T,
t  T.
(27)
(27) allow us to determine C1 and C2 in solutions for
reserves and parallel exchange rate obtained for dualrate regime.
65

Setting t = 0 in (22), t = T in (21) and (22) and using (26)
and (27) yields
~
~
1C1 + 2C2 = R0 – R,
~
R +  1C 1

e-1T
+  2C 2
e-2T
(28)
~
= s + C1e-1T + C2e-2T.
(29)
Solving this system yields
C1 = [(1-2
~
~
)e2T(R
~
0
~
- R) - 2(R - e)]/, (30)
C2 = [1(R - e) - (1-1
~
)e 1T(R
where  = 1(1-2)e2 T - 2(1-1)e1T.
0-
~
R)]/, (31)
66





Substituting (30) and (31) in (21) and (22) yields
solutions for parallel exchange rate and foreign reserves
prior to reform.
These solutions depend on
~
 relation between initial value of reserves R0 and its
~
steady-state value in dual rate regime R, and
 initial level of domestic credit,
~
since –R – e = -(1-)(d-e)/ .
Realistic case for developing countries: initial situation in
which positive premium exists up to an instant before
~
future reform is announced (s0 > e = s).
~
~
Such a condition is satisfied for R0 > R.
Assuming that this condition holds, two cases must be
considered, depending on initial positions of saddlepath
67
SS and terminal curve CC.
Figure 18.3: case where CC is steeper than SS.
+.
 In such a situation, e < s0 < T
+
 Announcement at t = 0 of future reform at T leads to
immediate depreciation of parallel exchange rate and
gradual fall in reserves.
 Dynamic behavior in anticipation of reform is illustrated
in Figure 18.3.
~
 Economy before reform announcement is such that R0 >
~
R and is located on a point such as A on saddlepath SS.
 Position of steady-state equilibrium in post-unification
regime depends on length of T.
If reform occurs where T  0+:
 Upon announcement, economy moves immediately to its
post-reform steady state.
68
F
i
g
u
r
e
1
8
.
3
D
y
n
a
m
i
c
s
u
p
o
n
U
n
i
f
i
c
a
t
i
o
n
:
C
a
s
e
I
.
s
=
0
s



C
B

+
S
B
' D
0
'

s
0
_
E
s
*
=
e
.
A
R
=
0
S
C
R
*
R
0
R
69
Parallel exchange rate jumps from A to B, with no
change in the initial stock of reserves, and unified
exchange rate equal to 0+.
If T is positive:
 At the moment future reform is announced, parallel
exchange rate depreciates instantaneously.
 It appreciates continuously during transition period
toward B’ on curve CC, which is reached at T.
 Unified exchange rate is equal to ’, which is more
appreciated than 0+.
If T  :
 Announcement will have no effect on parallel exchange
rate and reserves.

70



Economy will remain at initial position A upon
announcement, and then move along original
saddlepath SS toward E.
There are therefore three possible paths associated with
unification when terminal curve CC is steeper than SS.
While reserves fall continuously, parallel market
exchange can either
 jump instantaneously to more depreciated equilibrium
level, appreciate continuously after initial depreciation
 or appreciate continuously with no initial jump toward
unified exchange rate, which may be
 above initial parallel exchange rate, or
 below it,
depending on T and parameter values.
71
Figure 18.4: CC is flatter than SS.
+
 This case corresponds to s0 > T > e.
 If the reform occurs overnight: parallel exchange rate will
appreciate immediately and jump from A to B.
 If reform is preannounced: initial downward jump is from
A to D, followed by movement along unstable trajectory
DB’ during transition period, which is characterized by
 falling reserves;
 continuous appreciation of parallel exchange rate with
or without initial downward jump.
Summary:
 Short-run behavior of parallel exchange rate in
anticipation of reform depends on
 state of expectations about timing of reform;
72
F
i
g
u
r
e
1
8
.
4
D
y
n
a
m
i
c
s
u
p
o
n
U
n
i
f
i
c
a
t
i
o
n
:
C
a
s
e
I
I
s



.
s
=
0S
A
C
s
0
D

+
.
R
=
0
B
'B
E
0
_
s
*
=
e
C
S
R
*
R
0
R
73
initial position of the economy;
 length of transition period between announcement
and implementation of reform;
 macroeconomic policy stance that agents expect
policymakers to adopt in post-reform regime.
If unification attempt is fully anticipated, agents will
adjust their portfolios toward
 foreign-currency-denominated assets if uniform
floating exchange rate is expected to be more
depreciated than existing parallel rate; or
 domestic-currency-denominated assets if floating
exchange rate is expected to be more appreciated.
Due to this adjustment, parallel market rate will
depreciate or appreciate immediately.



74
After initial jump, and if the reform is known to take place
in the near future, parallel market rate will steadily
depreciate or appreciate toward unified exchange rate.
Reason for jump in parallel exchange rate upon
announcement of the reform:
 If unified exchange rate is more depreciated than initial
level of parallel exchange rate and official exchange
rate, agents realize that future reform may imply
 depreciation of official exchange rate;
 rise in prices; and thus
 reduction in real money balances.
 Under perfect foresight, these effects are reflected
immediately in the expected and actual rate of
depreciation of parallel rate.

75
This leads agents to reduce their demand for domestic
currency.
 But since the initial money stock is constant, equilibrium
in money market can be maintained only if parallel
exchange rate depreciates.
Uncertainty about reform date or nature of posttransition regime:
 Expectations of reform would cause
 jump in parallel exchange rate at the moment the
reform is implemented;
 volatile exchange-rate movements prior to transition if
this type of uncertainty varies over time.
 It may be difficult to interpret short-run movements in
parallel exchange rate.

76

Erratic fluctuations in exchange rates induced by this
uncertainty may
 distort portfolio decisions; and
 have adverse inflationary effect.
77
Longer-Run Effects of Unification
Pinto (1989, 1991a):
 These effects depend on fiscal implications of exchangerate reform.
 He models government behavior and assumes that
rationing occurs in official market for foreign exchange.
 Government buys imported goods and uses foreign
exchange surrendered by exporters to pay for them.
 Excess foreign exchange is resold to private sector at
official exchange rate.
 Thus, reserves remain constant over time and dynamics
of money supply are determined by credit policy.
 Production for exports is partly smuggled out and partly
sold through official market.
78




Thus, parallel market premium acts as implicit tax on
exports taking place through official channels.
Since exporters face rising marginal costs to smuggling,
allocation of output between official and parallel markets
is determined by equating marginal returns.
In equilibrium proportion of smuggled exports is
positively related to the parallel market premium.
Private agents spend a fixed fraction of their wealth on
imported consumption goods.
79

Government finances its deficit by credit from central
bank:
D = E(G - ),

E: official exchange rate; G: government imports;
: revenue from conventional, explicit taxes.
(32) implies that
.
m = G -  - m,

(32)
(33)
m: real money stock in terms of official exchange rate;
: constant rate of official depreciation.
.
(33): in steady state (m = 0) budget deficit is financed by
~
80
inflation tax m.






To identify steady-state revenue from implicit tax on
exports requires moving from
 “recorded” deficit measured in terms of official
exchange rate (33) to
 “true” deficit measured in terms of parallel exchange
rate.
Let  denote 1 plus parallel market premium.
“True” long-run capital loss incurred by private sector
~ ~
due to inflation tax is m/.
Officially recorded taxes are equal to E.
~
Real tax burden is equal to /.
~
If  > 1, “true” fiscal burden imposed by inflation tax and
conventional taxes is lower than “recorded” one.
81



But since real government spending does not depend on
premium, difference between “recorded” and “true” tax
~
~ ~
burdens is implicit tax on exports: G - /
- m/.
~
Since G -  = m, this tax is equal to G(1-1/).
Steady-state government budget constraint:
~
~
 m

1
m
G =  + m = ( ~ + ~ ) = G(1 - ~ ) + ~ + ~ ,





~


~
This shows how implicit tax on exports serves to finance
government expenditure.
If elasticity of demand for domestic money is less than
unity, there is trade-off between inflation tax rate and
premium for given level of fiscal deficit.
82
By unifying, government loses tax revenue implicit in
~
premium, G(1 - 1/).
 The larger implicit tax on exports prior to reform, the
larger the jump in inflation upon unification.
 Pinto's analysis neglects another important source of
implicit tax or subsidy associated with informal dual
exchange-rate regimes.
Agénor and Ucer (1998):
 Taxes on imports are important source of revenue for
developing countries.
 In many countries, official rather than parallel market
exchange rate is used for customs valuation purposes.
 Use of official exchange rate for valuation of imports for
duty purposes provides implicit subsidy to importers.

83


If subsidies are large relative to revenue generated from
implicit tax on exports, net long-run effect of exchange
market unification may be a fall in domestic inflation rate.
Evidence: prior to unification significant net quasi-fiscal
losses were registered in many countries operating
informal currency markets.
84
Evidence on Unification Attempts
Experience of sub-Saharan African countries in the
mid-1980s.
 While some of them chose to follow gradual path to
unification, most of them opted for “overnight” approach.
 Indications:
 In some countries exchange-rate unification led to a
surge in inflation.
 Parallel market premium rose substantially in the
months preceding unification attempt and fell sharply
upon implementation of reform.
 Unified floating exchange rate was in some cases
very close to pre-reform parallel rate.
85
This implies that drop in premium resulted from
sharp depreciation of official exchange rate.
 Case most notably in Nigeria.
 Despite sharp drop on impact, significant premium
reemerged in some countries.
Other countries:
 Unified their foreign exchange markets by adopting a
floating exchange-rate arrangement: Egypt (early 1991),
Guyana (March 1990), India (March 1993), Iran (March
1993), Jamaica (September 1991), Madagascar (May
1994), Peru (August 1990), Sri Lanka (August 1990),
Trinidad and Tobago (April 1993), and Venezuela (March
1989).
 Prior to floating, these countries had widespread trade
and exchange restrictions.
86

These restrictions led to substantial diversion of foreign
exchange to parallel market, creating in some cases
severe balance-of-payments problems.
Experience of Jamaica, Guyana, and Sri Lanka:
 In addition to balance-of-payments difficulties, pre
reform period was marked by rising inflation.
 Parallel market premium also displayed a rising trend.
 Adoption of floating exchange-rate system was
accompanied by relaxation of trade and capital account
restrictions.
 After the reform was implemented, parallel market
premium dropped considerably, but, import/reserves
ratio recovered sharply only in Guyana.
 Overall balance of payments deteriorated.

87





Inflation is accelerated in Guyana and Jamaica, but fell
in Sri Lanka.
Sharp acceleration of money growth occurred in Guyana
and Jamaica.
Possible reason for inflationary burst: inability of
policymakers to control primary budget deficit and rate of
expansion of money supply (Agénor and Ucer, 1998).
This may be the case in several unification experiments
attempted in sub-Saharan African countries.
Increase in premium prior to reform can be a result of
expectations about
 timing of reform process, and
 size and direction of movements in official and parallel
exchange rates upon implementation.
88





Parallel market exchange rate in some cases did not
undergo large “jump” when reform was implemented.
Reason: prediction of the reform with high degree of
accuracy by private agents.
At the time of implementation, only official exchange rate
adjusts in a discrete fashion.
Reason: most of the effects of the change in exchangerate regime have already been discounted through
parallel market by forward-looking agents.
Reemergence of significant premium subsequent to
reform occurred in countries where
 money growth was not kept under control;
 foreign exchange controls were retained or
reintroduced;
89
 inflation rose substantially.