The Open Economy

1
The Open Economy

• Part 1. Adding the Exchange Rate to the
  3-EquationMacro Model

2
Why are exchange rates important?

• In open economy, exchange rate is key component
  of monetary transmission mechanism
• Channel through which overseas developments may
  affect domestic economy. Eg are flexible exchange
  rates shock absorbers or additional sources of
  volatility?
• Real exchange rate affects terms on which
  companies compete in international markets

3
Real and nominal exchange rates

• ? price of foreign goods in domestic currency
  price of domestic goods ?
  Pe P
• where P is foreign price level P is domestic
  price level e units of domestic currency
  one unit of foreign currencyor in logs
  ? e p - p
• e? ? exchange rate depreciation

4
Price setting in the open economy

• Two alternative pricing rules(1) Home-cost
  pricing firms set export prices in same way as
  goods sold at home, ie based on domestic
  costs(2) World pricing firms set export prices
  based on prices of similar products produced
  abroad
• Suppose increase in costs in UK but not abroad.
  Under (1) price competitiveness deteriorates.
  Under (2) non-price competitiveness deteriorates.
  Both captured by RULC foreign unit labour
  costs in domestic currency
  domestic unit labour costs
  ULCe ULC

5
What determines exchange rate movements?

• Currency fluctuations should reflect
  macroeconomic variables such as domestic and
  foreign real interest rates, money supplies,
  current accounts and relative domestic demand
• But there is no well specified theoretical
  exchange rate model that encompasses all the
  relevant variables and performs well empirically
• Nothing can systematically explain exchange rate
  movements between major currencies

6
Purchasing power parity (PPP)

• Law of one price (LOOP) implies common currency
  price of a traded good is identical in different
  countries. PPP is exchange rate that equalises
  price levels expressed in domestic currency
  between two countries (when LOOP holds for all
  traded goods)
• Assumed to hold in long run. In short run, real
  exchange rate variability is almost perfect
  reflection of nominal exchange rate variability
• By definition?? ?e p p
• ?? 0 when PPP holds

7
Uncovered interest parity (UIP)

• If UIP holds, risk averse investors are
  indifferent between investing in domestic and
  foreign assets

If i rises by 1 relative to i for a year, e
would be expected to depreciate by 1 over the
next year To rule out arbitrage opportunities, e
must appreciate by 1 immediately
8
Exchange rate as a random walk

• On a daily basis, changes in floating foreign
  exchange rates are largely unpredictable
• On a month to month basis, over 90 of exchange
  rate movements are unexpected, and less than 10
  are predictable
• Empirically, it is hard to beat the following
  model?e ex where ex is white noise
• Bank of England MPC assumes ½UIP and ½RW when
  producing its forecasts

9
Dornbusch (1976) overshooting model

• Dornbusch developed a simple model that captured
  two stylised facts of fx markets(i) high
  exchange rate volatility(ii) high correlation
  between nominal and real exchange rates
• Dornbusch showed that with sluggish adjustment in
  goods market and rapid adjustment in asset
  markets, exchange rate volatility is needed to
  equilibriate economy in response to monetary
  shocks
• Two relationships at heart of Dornbusch model(1)
  UIP ?e i i(2) Money demand m p ?i
  fy ?,fgt0

10
Dornbusch model (continued)

• Need to combine equations (1) and (2) with three
  assumptions(i) domestic price level does not
  move instantaneously to unanticipated monetary
  disturbances but adjusts slowly over time(ii)
  output moves slowly in response to monetary
  shocks(iii) money is neutral in long run, ie a
  permanent fall in m leads to a proportionate fall
  in e and p in long run
• Suppose m falls. With p fixed in short run, (m-p)
  falls. To equilibriate system, demand for real
  balances must fall. With y fixed in short run,
  (2) ? i must rise
• From (1) when i rises, domestic currency expected
  to depreciate. But how is this possible if
  long-run effect of higher m is proportionate
  appreciation of exchange rate?

11
Exchange rate overshooting

• Only way of resolving conundrum is for initial
  appreciation, on impact, to be larger than
  long-run appreciation, ie exchange rate must
  overshoot
• Whole result driven by assumed rigidity of
  domestic prices. No overshooting if all prices
  perfectly flexible

12
Weaknesses of Dornbusch model

• Considers flows of income etc but not potential
  stock implications of these flows (subsequently
  considered by portfolio balance models)
• Ignores role of current account in exchange rate
  determination (subsequently considered by
  intertemporal approach)
• Neither microfounded nor general equilibrium in
  nature
• Model has difficulty rationalising high
  persistence of real exchange rate movements
• Model not proved good at forecasting future
  currency movements

13
Fundamental equilibrium exchange rate (FEER)

• FEER is exchange rate consistent with internal
  and external balance)
• Wide range of estimates of equilibrium exchange
  rate for /

Selected estimates of / nominal equilibrium
exchange rate
14
Monetary policy under flexible exchange
ratesCase 1 PPP and UIP hold simultaneously

• When PPP and UIP hold simultaneously (in
  long-run)PPP ?? ?e p p 0UIP ?e i
  i
• Note that irp and irp?? r r
• When PPP holds, r r
• When PPP and UIP hold simultaneously, no room for
  independent real interest rate policy under
  flexible exchange rates. Central bank cant deal
  with demand shocks
• But since irp, central bank can set nominal
  interest rates to achieve inflation target

15
New Keynesian open economy model

• IS curvey a ßr ?? eD
• Phillips curvep pT dy eS
• Assume companies price to market, so they leave
  prices unchanged in local market even if nominal
  exchange rate changes

16
Monetary policy under flexible exchange
ratesCase 2 UIP holds but not PPP (short run)

• Minimise loss function L (p pT)2 ?y2
• ? y d eS
• d2?
• From IS curverOPT a 1 eD d eS
  ? ? ß ß ß(d2?) ß
• Central bank responds to demand and supply shocks
  as well as to the real exchange rate
• Note real exchange rate is function of r (?? r
  r)

17
Approximation to UIP

• Assume real exchange rate adjusts to its long-run
  level ? asymptotically?1 ? g(?-?)
• Subtract ? from both sides?? (1-g)? -
  (1-g)?? ? 1 ?? 1-g
• Note that ???ep-p. Also ?ei-i. And irp??
  (rp) (rp) p p r r? ? 1
  (r r) 1-g
• For simplicity, assume g0 and ?0? r r

18
Case 2 continuedUIP holds but not PPP (short run)

• Assuming ? r-r
• rOPT a 1 eD d
  eS ? r ß? ß?
  (ß?)(d2 ?) ß?
• Central bank responds to demand and supply shocks
  as well as to the foreign real interest rate. But
  interest rates respond less vigorously to shocks
  than in closed economy case because of additional
  exchange rate channel
• Optimal response to shocks affecting demand side
  (eD and r) does not depend on central banks
  preferences ?
• Reaction of central bank to supply shocks depends
  on ?

19
Open economy IS curve

• IS curvey a ßr ?? eDy a ßr
  ?(rr) eDy (a?r) (ß?)r eD
• Open economy IS curve is flatter than closed
  economy IS curve.
• A cut in interest rates also leads to a
  depreciation in the exchange rate. Both channels
  stimulate demand
• Can conduct same analysis of shocks as in closed
  economy model

20
Monetary policy under flexible exchange
ratesCase 3 Exchange rates as random walk

• When real exchange rate follows a random walk?
  ? eX
• rOPT a 1 eD d eS ? eX
  ß ß ß(d2?) ß
• Random exchange rate movements constitute an
  additional shock to which the central bank has to
  respond with its interest rate policy
• Real interest rates could become very volatile if
  there are big random fluctuations in exchange
  rates. Typically interest rates are smoother ?
  sub-optimal response

21
Monetary policy under fixed exchange rates

• Central bank loses ability to run domestically
  oriented interest rate policy. Since ?e 0, UIP
  condition simplifies to i i
• Since r i p, we have r i p
• This can be transformed into a simple instrument
  ruler (ipT) (-1)(ppT) 0.y
• Taylor principle is no longer satisfied. Under
  fixed exchange rates, real interest rates have to
  fall when inflation rises. Economic system
  threatens to become unstable

22
Monetary policy response to an inflation shock
under fixed exchange rates
r
rs
A
IS
y
p
VPC
PC(pe2)
pT2
A
0
y
y
23
Arguments for flexible exchange rates

• Exchange rate changes needed to compensate for
  inflation differentials (PPP). Equilibrium value
  of exchange rate is unknown. Fixing at wrong rate
  could be damaging
• Fixed exchange rates can only be adjusted
  sporadically, leaving long periods of
  misalignments. Expectation of realignments may
  lead to speculative attacks
• Exchange rate changes needed to cope with shocks
  that alter external competitiveness eg changing
  energy prices. With a fixed exchange rate regime,
  either all prices have to adjust or exchange rate
  must change. With wage and price rigidity, less
  painful to change exchange rate

24
Arguments for fixed exchange rates

• Flexible exchange rates fluctuate by far more
  than can be explained by inflation differentials
  or real disturbances
• Overshooting implies that exchange rate tends to
  move away from its equilibrium level
• Volatility of exchange rates imparts additional
  uncertainty and instability to the economy

25
Dangers of an overvalued exchange rate

• All attempts by UK government to fix sterling
  exchange rate have ended in failure. Two common
  themes (1) Sterling fixed at uncompetitive
  exchange rate. (2) Transitional costs
  underestimated
• Return to Gold Standard in 1925 Ended in forced
  devaluation after six years of stagnant growth,
  high unemployment and a national strike
• Post-war fix vs US dollar in 1946 and rejection
  of devaluation by incoming Labour government in
  1964 Both episodes ended in forced devaluation 3
  years later
• ERM entry in 1990 Sterling was ejected from ERM
  less than two years later

26
Further reading

• Burda and Wyplosz, Macroeconomics, A European
  Text, 4th edition, Chapters 7, 19, 20
• Carlin and Soskice, Macroeconomics, Chapters 2,
  5, 9
• For a very readable introduction to the
  Dornbusch model, seeRogoff, K (2002),
  Dornbuschs overshooting model after twenty five
  years, Second Annual IMF Research Conference
  The Mundell-Fleming lecturehttp//www.imf.org/ext
  ernal/np/speeches/2001/kr/112901.pdf