Transcript Real Exchange Rate Fluctuations

Real Exchange Rate Fluctuations:
Reflections on the Uruguayan Experience
Umberto Della Mea*
Economic Policy Division
Central Bank of Uruguay
Outline
*
I.
Some simple facts about real exchange rate fluctuations in a
dollarized, multi-rest-of-the-world economy.
II.
What to do?
III.
Addendum: about the effect of real exchange rate fluctuations
on fiscal solvency
The views here expressed are those of the author and do not necessarily represent those of the Central Bank of Uruguay.
Whither rest-of-the-world?
Real effective exchange rates and exports share
100%
200
90%
180
80%
2000=100
160
70%
185%
140
60%
50%
120
40%
100
30%
80
20%
60
10%
40
0%
Global
Ene-06
Oct-05
Jul-05
Abr-05
% Rest of the World
Rest of the World
Ene-05
Oct-04
Jul-04
Abr-04
Ene-04
Oct-03
Jul-03
Abr-03
Ene-03
Oct-02
% Brazil
Brasil
Jul-02
Abr-02
Ene-02
Oct-01
Jul-01
Abr-01
Ene-01
Oct-00
Jul-00
Abr-00
Ene-00
% Argentina
Argentina
A winner’s curse?: the real exchange rate as a transmission mechanism…
… between the capital and the current account: strong RER appreciations seem mostly
explained by capital inflows. Sudden-stops and sudden-starts (not offset by similar
changes in international reserves) are usually behind sharp changes in trends.
-21%
-18%
-15%
-12%
-9%
-6%
-3%
0%
3%
6%
9%
12%
15%
190
170
150
130
110
90
2005
2004
2003
2002
2001
2000
Net Capital Outflows (right scale)
RER (extra regional)
% of GDP
RER Base 2000=100
210
Who's afraid of dollarization?: the procyclical role of the banking sector
Liabilities dollarization (currently 92%) introduces positive balance-sheet effects: non
performing assets and banks solvency improve in periods of growth, boosting
confidence in the banking system, fueling capital inflows (including repatriations) and
further appreciating the real exchange rate.
Private Banking System
200
55
RER Base 2000=100
180
45
40
160
35
140
30
25
120
20
100
15
2005
2004
2003
2002
2001
2000
Non Performing Loans
RER (extra regional)
Non Performing Loans (in %)
50
Who's afraid of dollarization?: the procyclical role of the banking sector (cont.)
Return on assets improve with the real exchange rate, further encouraging credit risk
taking and more agressive credit policies which feed back real exchange rate
appreciation through capital inflows and domestic expenditure.
Private Banking System
5
0
180
-5
160
-10
140
-15
-20
120
-25
100
-30
2005
2004
2003
2002
2001
2000
Return on Assets
RER (extra regional)
Return on Assets (in %)
RER Base 2000=100
200
The original sin in reverse…
• Governments in emerging economies often need to issue foreign currency debt
(original sin), even in domestic markets (original super-sin).
• Real exchange rate appreciations improve fiscal solvency through Debt/GDP and
Interest/Revenues reduction.
200
110
20
180
17
160
14
140
11
120
8
100
5
90
80
160
70
140
60
50
120
RER Base 2000=100
180
Debt/GDP (in %)
RER Base 2000=100
100
40
100
30
2005
Interest payments/Govt revenues
2004
2003
2002
2001
2000
RER (extra regional)
2005
2004
2003
2002
2001
2000
Debt/GDP
Interest/Revenues (in %)
200
RER (extra regional)
… and the spillover towards the private sector
210
3000
190
2500
170
2000
150
1500
130
1000
110
500
90
0
2005
2004
2003
2002
2001
2000
Sovereign Spread
RER (extra regional)
BPS over UST
RER Base 2000=100
Sovereign credit quality is normally a ceiling for private credit quality. An improval in
fiscal accounts may also trigger more investment and private capital flights.
Digression: monetary policy and inflation targeting under RER pressure
PT
 by 21%  minimum CPI inflation consistent with  N  0  13%
PN
40%
35%
30%
Likely monetary policy constraint:
“achieve inflation target s.t. N0”
25%
20%
15%
10%
5%
0%
-5%
Nov-05
Non tradables
Sep-05
Jul-05
May-05
Mar-05
Ene-05
Nov-04
Sep-04
Tradables
Jul-04
May-04
Mar-04
Ene-04
Nov-03
Sep-03
Jul-03
May-03
Mar-03
Ene-03
Nov-02
Sep-02
Jul-02
May-02
Mar-02
Ene-02
CPI, 12 rolling months
What to do? The basic approach
•
Productivity differentials, changes in preferences technology, changes in public
sector expenditure and shocks in the terms of trade don’t seem enough to
understand massive changes in this variable. Sudden-stops and sudden-starts in
capital movements seem good candidates, instead.
•
What to do, then? The usual candidates are:
 Curbing the boom-bust cycle by modifying capital and reserve requirements to
the banks: might generate problems if they are regarded as (1) a change in
the rules, in the upside or (2) a regulatory subsidy, in the downside.
 Countercyclical provisions: the advantage of having contingent rules to the
state of the nature.
 Capital controls: ¿…?
 Fiscal flexibility: always welcome.
 Sterilized FX interventions? Efficiency under discussion.
 Unsterilized FX interventions? Inflation risks.
The unlikely usefulness of nominal exchange rate policies
160%
Nominal and real FX
move in opposite
directions
(61% of total cases)
140%
120%
100%
80%
60%
40%
20%
0%
-20%
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
1980
1979
1978
Avg nominal depreciation/devaluation
Avg real depreciation/devaluation
Addendum: real exchange rate and fiscal solvency: stock vs flow approaches
The stock approach, when public debt is foreign currency denominated
The standard exercise of fiscal solvency à la Blanchard is based on the analysis of the
Debt/GDP (D/Y) ratio under a set of assumptions concerning: the primary fiscal surplus
(PFS), the interest rate (i*), the rate of growth (), the international tradable inflation (t*)
and the evolution of the RER (q) weighted by the share of tradables in the GDP deflator
():
D
d
Y 
dt


 PFS  1  i *

Y
1  *t






1  q 

  1 D
Y
1 


.

Addendum: real exchange rate and fiscal solvency: stock vs flow approaches
The flow approach
The fiscal deficit at constant prices FD  depends on the real evolution of revenues and
expenditures, but also on the evolution of the terms of trade of the government:
 IPT 
d

IPG 
d FD
 d G  T  

 0 
 ...
  1


dt
dt
 dt




GOVT '
REAL



EFFECT
ON G & T
TERMS OF
EXCHANGE
EFFECT
d
where:
IPT
IPG

dq
T  G q  
T
G 1
0 
T  G   0
If Tradables weigh more in the revenues basket (T) than in expenditures (G), -eg,
because taxes are more based on consumption goods while expenditures are more
concentrated in nontradables (e.g, wages and pensions), then a real depreciation
improves the fiscal accounts.
Addendum: real exchange rate and fiscal solvency: stock vs flow approaches
Fiscal impact of a % change in RER, Uruguay 1991-2004 (preliminary)
The RER is highly correlated with
government’s terms of exchange:
increase in PT over PN increases
domestic purchasing power of
revenues
120
110
100
90
80
70
60
50
105
100
95
90
phi
q
RER defined as q=PT/PN vs
TE (phi=PR/PE)
85
80
2003
2001
1999
1997
1995
1993
1991
q
phi
TE effect as a % of real revenues,
as a function of % changes in RER (q=PT/PN)
40%
20%
0%
-20%
-40%
-60%
-80%
40%
30%
% changes in RER
20%
10%
0%
-10%
-20%
-30%
This correlation sustained a significant
improvement in the fiscal flows in 2002,
measured at constant prices. But during
appreciations, it is countercyclical and
mitigates the stock effect!
the
an
the
tax
Addendum: real exchange rate and fiscal solvency: stock vs flow approaches
Decomposition of changes in the fiscal deficit (preliminary)
6000
180
4000
160
150
2000
140
0
130
120
-2000
110
100
-4000
90
-6000
80
2004
2003
2002
2001
2000
1999
Quantum effect
Other
Terms of Exchange Effect
RER (extra region, right scale)
RER Base 2000=100
millions of UYP, 1991 prices
170