Exchange Rate Policies in Latin America: Discussion

1
Exchange Rate Policies in Latin America
Discussion

• Roberto Chang
• Rutgers University and NBER

2
Summary of the Papers

• The three papers in this session discuss recent
  experience with inflation targeting in Colombia,
  Chile and Peru, with emphasis on exchange rate
  management issues.
• The three cases share several common features,
  but also differ in substantial ways.

3
Inflation Targeting Experiences

• Transition to a fully fledged inflation
  targeting regime has been recent (1999 in Col and
  Chi, 2002 in Peru).
• Announced explicit targets for inflation over
  some horizon (2.5 - 3 /- 1 in Chi and Pe, 5
  in Col).
• Communication with the public has been
  institutionalized (i.e. Inflation Reports).

4
Policy Implementation

• All three central banks have switched to using a
  short term interest rate as their main (but not
  the only) policy instrument.
• And they claim to have flexible exchange rates,
  consistent with the inflation targeting regime.

5
Favorable Results

• By and large, the three countries have been very
  successful at controlling inflation (to about 2-3
  percent in Chile and Peru, 5 percent in Col).
• Output growth in Chile and Peru rather
  satisfactory for the region, less so in Colombia.

6
Impact on Credibility

• Not only inflation has fallen to low levels, but
  policy appears to have gained substantial
  credibility.
• In Chile and Peru, surveys of inflation
  expectations have converged to inflation targets.
  
• In Colombia, expectations of inflation remain
  half a percent above target range.

7
Departing from the IT playbook

• As noted, the three central banks claim to have
  flexible rates.
• But they also note that they reserve the right to
  depart from flexible rates under some exceptional
  circumstances.

8
Cited reasons for deviating from flexible rates

• Excessive exchange rate movements not justified
  by fundamentals (Colombia, Chile).
• Large devaluations that may jeopardize financial
  and payments systems (Peru) .
• Forex intervention may complement monetary policy
  (Colombia).

9
What the deviations have looked like

• Interest rate has been adjusted upwards to
  prevent an excessive devaluation (Peru in
  September 2002).
• Foreign Exchange intervention, with or without
  sterilization, announced and limited or
  discretionary.

10
Should we worry about these peccadilloes?

• Are there good reasons to engage in active
  exchange rate management within an inflation
  targeting regime?
• If so, what is an appropriate way to do so?

11
What conventional IT gurus would say

• Svensson (1998, 2000) examines the optimal choice
  of a central bank that chooses its instrument
  path maximizes expected discounted loss of a (LQ)
  function of inflation and real activity.
• A central result (in the presence of control
  lags) is that the instrument should be set so as
  to minimize deviations of inflation forecast from
  target (as in Chile and Peru).

12

• A corollary is that the current instrument
  setting should react to any variable that
  provides news with respect to expected inflation.
• Foreign variables and exchange rates are clearly
  in the latter category.

13

• the more open the economy, the more the
  reaction function that determines the central
  banks instrument will depend on foreign
  variables, for instance foreign inflation,
  output, and interest rates (Svensson 1998)

14

• So, even from the now widely popular theory of
  inflation targeting it should not be surprising
  that implementation of that regime should make
  the instrument (overnight interest rates) respond
  to news in exchange rates.
• But of course this does not seem to be the main
  reason why our central bankers have engaged in
  active exchange rate management .

15
What conventional IT theory misses and its
implications

• In practice, our central bankers express strong
  concern for two possibilities that exchange
  rates movements may not be justified by
  fundamentals, and that large devaluations may
  destabilize financial and payments systems.

16
On non-fundamental exchange rate gyrations

• It is very hard to dispute the contention that
  exchange rate movements at times (and perhaps
  most of the time) are not justified by
  fundamentals.
• But it is not clear what kind of policy such a
  contention justifies.

17
Exchange Rates and Financial Stability

• We are now on much firmer ground to discuss the
  effects of exchange rate movements on financial
  stability.
• An active literature on liability dollarization,
  currency mismatches, and financial frictions
  emphasizes how a devaluation may result in a
  contraction in aggregate demand through net
  worth effects .

18

• Therefore, countries with high degrees of
  liability dollarization and less developed
  financial systems (e.g. Peru) are well advised to
  pay close attention to the balance sheet effect
  of devaluation.
• And it could be argued that a (Peruvian like)
  strong effort at containing dollarization,
  reducing mismatches, and the like, is warranted.

19

• On the other hand, it is not obvious in this case
  how exactly the central bank should adjust its
  policy instrument to news in exchange rates.
• It would be desirable to derive optimal policy in
  the context of a Svensson-type framework,
  extended to allow for balance sheet effects, etc.
  

20
How to manage the exchange rate

• The three central banks state that they have used
  foreign exchange intervention, as opposed to
  adjusting the interest rate, to influence
  exchange rates.
• This appears much more problematic.

21
Sterilized or not?

• The existing literature on intervention fails to
  provide clear evidence on the effectiveness of
  sterilized intervention (see Disyatat and Galati
  2005 for a recent survey).
• Most studies find a weak and short lived effect
  of intervention on exchange rates.
• The signaling effect is controversial.

22

• A non sterilized intervention can be seen as a
  sterilized one plus an adjustment in the supply
  of money.
• Seen in this way, it amounts to changing the
  policy instrument (money instead of interest
  rate).
• This entails no gain and instead impairs
  transparency and causes confusion.

23

• So it is not surprising that, in the Colombian
  case, it is now harder for the public to verify
  and for the central bank to communicate the
  monetary policy stance, since the central bank is
  not only moving the interest rate, but also
  intervening in the for-ex market with substantial
  amounts, closing/opening repo facilities, and
  selling government securities in the secondary
  market. (Vargas, page 10)

24

• The conclusion is that it does not appear
  advisable to include foreign exchange
  intervention, whether sterilized or not, as part
  of the overall monetary policy framework.
• The central bank should adjust interest rates
  instead.

25

• Indeed, in his review of the Central Bank of New
  Zealand, Svensson (2001) concludes I see no
  reason why a transparent inflation-targeter
  should undertake foreign exchange interventions.

26

• This is not to say that accumulating foreign
  reserves when they are cheap is a bad idea.
• But the rationale for doing so (such as enhancing
  the ability of the central bank to act as a
  lender of last resort in dollars) is about
  preserving the stability of the financial system
  and prevent crises, not about implementing
  inflation targeting.

27
Final Remarks

• The recent progress in monetary management in the
  three countries has been remarkable.
• Surely, it has been aided by sound fiscal policy.
• It is fair to claim success.
• Careful in tinkering with what is not broken.