Contents of the course

1
International Finance
Part 1 Fundamentals of International Finance
Lecture n2 Exchange rate determination
2
Exchange rate determination

• Three approaches for exchange rate determination
• 1. Balance of payments approaches
• Current account balances
• Portfolio investments
• 2. Parity relations
• Purchasing Power Parity (PPP)
• Interest Rate Parity (CIP UIP)
• 3. Asset approaches
• Relative interest rates
• Supply and demand for assets

3
BOP approaches

• 1. Balance of payments approaches
• FX rates are adjusted so that the BOP is in
  equilibrium
• In case of fixed exchange rates
• Government in charge of the BOP equilibrium
• FX rates maintained via the change in currency
  reserves
• In case of deficit the central bank ease
  devaluation pressure by buying the domestic
  currency and selling foreign currencies (out of
  its reserves) and gold. If the disequilibrium is
  too large and the central banks run out of
  reserves, the domestic currency will devalue.
• BOP disequilibrium are then used to forecast the
  evolution of FX rates. Example Thailand and the
  Thai baht.

4
BOP approaches

• 1. Balance of payments approaches
• FX rates are adjusted so that the BOP is in
  equilibrium
• In case of floating exchange rates
• Deficit countries FX rates are expected to
  depreciate , due to the excess supply of money.
• Surplus countries FX rates are expected to
  appreciate, due to the relative shortage of
  domestic currency.
• In case of managed floats
• Changes on relative interest rates to influence
  the capital inflows or outflows impacting the BOP
  and the valuation of a currency. Ex rise in
  interest rates to increase money demand (capital
  inflows) and support the value of the currency.
  Bop trends trends help to forecast such moves.

5
Purchasing Power Parity

• 2. Parity Relations - PPP
• Absolute Purchasing Power Parity - Definition
• P S.P law of one price
• Domestic Prices Exchange Rate . Foreign prices
• P and P general price indices
• Rearranging S P/P
• The spot exchange rate between 2 currencies is
  equal to the ratio of general price levels
  between the 2 countries.

6
Purchasing Power Parity

• Absolute Purchasing Power Parity - Hypotheses
• Hypotheses
• No transports costs
• Perfect information (on prices in both countries)
• Homogeneous goods
• No trade barriers
• -gt Equality brought by arbitrage
• Definition of arbitrage buy and resell without
  risk but with a profit
• Example of arbitrage buy 20 kg of gold in
  Belgium at 50.000 euros, and resell is
  immediately in France at 53.000 euros. Question
  is this imaginable for any goods?

7
Purchasing Power Parity

• Relative Purchasing Power Parity - Definition
• S b.P/P
• Prices across countries might differ by a
  constant factor b, accounting for transport
  costs and information costs.
• Focuses on the movements in the exchange rate and
  the extent to which they reflect differential
  inflation
• dS/S dP/P - dP/P

8
Purchasing Power Parity

• PPP absolute relative - Interpretation ?
• No precision of causality does the prices
  determine the FX rate, or the reverse?
• Goods included traded non traded? If yes,
  hypothesis of perfect substitutability and
  similar productivity levels.
• If only traded goods included PPP close to a
  tautology
• Short-run or long-run anchor?
• -gt alternative cost parity theory
• (more seducing, but same nature of problems)

9
Purchasing Power Parity

• PPP absolute relative - Theoretical criticisms
  (6)
• Information costs, transport costs, trade
  barriers exist, and could change over time.
• The direction of causality is unclear -gt exchange
  rates could determine prices.
• All disturbances are monetary, or more important
  than real ones -gtno account of productivity
  changes in one country.
• No account of productivity differential between
  traded non-traded goods sectors.
• Ignores the role of income in determining
  exchange rates, and its consequences on a change
  in Demand.
• No role of capital flows. Only focus on exchange
  of goods.

10
Purchasing Power Parity

• PPP absolute relative - Empirical testing
• Tests of commodity arbitrage, of the absolute
  version and on the relative version.
• Methods used
• Regressions
• Plots of data
• Cointegration tests (for the long-term
  relationship)
• -gt Commodity arbitrage appear not to maintain the
  law of one price. Little empirical support.
• Recent multivariate cointegration tests some
  support (relative version) for some large
  currencies.

11
Interest Rate Parity

• 2. Parity Relations - Fisher - CIP - UIP
• The Fisher Effect - Definition
• The Fisher Effect states that nominal interest
  rates in each country are equal to the required
  real rate of return plus compensation for
  expected inflation.
• That is i r ? i r ?
• Empirical tests show that the Fisher effect
  generally exists for short-maturity government
  securities, but are inconclusive for longer
  maturities.
• The International Fisher Effect states that
  nominal interest rates in each country are equal
  to the required real

12
Interest Rate Parity

• The Fisher Effect - Definition
• The International Fisher Effect (or Fisher-open)
  states that the spot exchange rate should change
  in an amount equal to but in the opposite
  direction of the difference in interest rates
  between countries.
• That is
• Or, in a simplified form
• Justification for the international Fisher effect
  is that investors must be rewarded or penalized
  to offset the expected change in exchange rates

13
Interest Rate Parity

• The Fisher Effect - Empirical evidence
• Empirical tests lend some support to the
  international Fisher effect, despite large
  short-run deviations.
• A more serious criticism comes from studies
  suggesting the existence of a foreign exchange
  risk premium for most major currencies.
• Also, speculation in uncovered interest arbitrage
  creates distortions in currency markets.

14
Interest Rate Parity

• 2. Parity Relations - reminder
• The Forward Rate
• A forward rate is an exchange rate quoted today
  for settlement at some future date
• The forward exchange agreement between currencies
  states the rate of exchange at which a foreign
  currency will be bought or sold forward at a
  specific date in the future (typically 30, 60,
  90, 180, 270 or 360 days)
• The forward rate is calculated by adjusting the
  current spot rate by the ratio of euro currency
  interest rates of the same maturity for the two
  subject currencies
• The forward premium or discount is the percentage
  difference between the spot and forward rates
  stated in annual percentage terms

15
Interest Rate Parity

• 2. Parity Relations - CIP
• Covered Interest Rate Parity - Definition
• CIP states that returns between assets in
  different countries should be equalised.
• If they are not, equalisation is brought by
  arbitrage.
• It gives
• (1it) . Ft/St (1it)
• Return of foreign invt return of domestic invt
• where
• it foreign interest rate it domestic
  interest rate
• Ft forward exchange rate
• St spot exchange rate

16
Interest Rate Parity

• Covered Interest Rate Parity - Numerical example
• Let it 5 it 4 St 0.9/
• Then
• Ft St (1it)/ (1it) 0.9 x 1.04/1.05 0.8914
  forward rate (one year) of per . The
  difference with spot rate reflects the difference
  of interest rates in each currency.
• If the parity is not applied, we have
• (1it) . Ft/St gt (1it) or Ftgt St (1it)/
  (1it)
• arbitrageurs will change large amounts of
  against at spot rate (0.9), pay i to borrow and
  lend at i, and sell them forward at F rate
  (gt0.89), making profit with no risk.

17
Interest Rate Parity

• Covered Interest Rate Parity - Hypotheses
• Assets same risk, same maturity
• No transaction costs
• No information costs
• No control on capital flows
• Plus, the transaction of in the forward market
  implies that there is no foreign exchange risk
  (risk that S changes while investing abroad).
• Arbitrage Ex. return greater abroad, we have
• (1it) . Ft/St gt (1it)
• -gt Investors will buy spot rate and sell forward
  (to invest abroad), causing S to rise and F to
  fall, getting back to equality.

18
Interest Rate Parity

• Covered Interest Rate Parity - Empirical tests
• Methods
• Plot the different interest rates separately and
  compare
• Test if deviations are significantly different
  from zero
• Regressing the forward premium - the difference
  between spot and forward rate - on (i-i) by
• fp a b(i-i)
• if b1, a 0, CIP holds (i- fpi)
• Deviations not uncommon, supposedly due to
• transaction costs (to what extent?)
• the existence of capital controls
• the existence of political risks risks of
  capital controls or taxes before funds are
  repatriated.

19
Interest Rate Parity

• Uncovered Interest Rate Parity - Definition
• UIP CIP foreign exchange risk
• States that returns between assets in different
  countries should be equalised, plus a deviation
  accounting for exchange risk. It gives
• i ? Se i
• where
• i foreign interest rate i domestic
  interest rate
• ? Se expected change in the spot exchange rate
• This relation holds if the path of the exchange
  rate is known with certainty, or if arbitrageurs
  are risk neutral.

20
Interest Rate Parity

• i ? Se i
• States that if the foreign interest rate is
  higher than the domestic rate, then the domestic
  currency must be expected to appreciate, to
  maintain this relationship (otherwise
  arbitrage).
• Uncovered Interest Rate Parity - Hypotheses
• Rational expectations, i.e., forward market is
  efficient
• Risk neutrality of arbitrageurs
• If risk aversion of arbitrageurs introduction
  of a risk premium i ? Se - i ?

21
Interest Rate Parity

• Uncovered Interest Rate Parity - Empirical
  evidence
• Difficulties
• Assess expectations on S
• Joint test of rational behavior of investors and
  of market efficiency (ex. no bubble phenomenon)
• General result UIP does not hold
• very little empirical support
• (possibly due to the existence of a risk premium)

22
Forward Market

• The forward market for foreign exchange
• Next to arbitrageurs, another important group on
  the forex markets speculators. They
  deliberately expose themselves to exchange rate
  risk.
• Speculators will trade on the basis of the
  difference between f (forward) and se (spot
  expected) at a given time horizon.
• Trade until f se
• Hypotheses underlying this relation
• Speculators are risk neutral
• Not prevented from operating on the forward
  market
• No transaction costs

23
Forward Market

• We rational behavior hypothesis, we have
• st ste ut , ut being a random walk (mean
  zero)
• with the arbitrage relation se f
• we have st ft-1 ut (1)
• meaning ft-1 non biaised estimator of St
• (1) is the efficient market condition relating
  the actual spot exchange rate to the forward
  rate.
• Tests over market efficiency of forward rates
• Difficulty joint test, both on market
  efficiency and on fundamentals of the model
  supposed to derive se
• Methods using regressions and serial
  autocorrelation tests

24
Forward Market

• Some results of the econometrical tests
• Empirical support of existence of a risk premium
  (time-varying), but no clear model of formation
• The lagged spot rate (st-1) outperforms the
  forward rate at predicting the spot rate -gt
  abnormal profits could have been made, trading on
  the basis of the difference between the current
  spot rate and the forward rate at a given time.
• Survey data about expectations formation of
  agents
• Expected change in spot rates is not an unbiased
  predictor of actual change in the spot rate
• Agents bias their estimation of spot rates, based
  on extrapolation of recent trend -gt destabilising
  expectations on exchange rates.

25
Summary

• Main results of the section
• Serious theoretical questions on PPP theory and
  few empirical support.
• CIP theory includes the role of capital mobility
  and arbitrage. More empirical support of CIP
  while UIP does not hold empirically.
• Relationship between spot and forward rates
  suggest the existence of a time-varying risk
  premium and some irrationality of market
  participants while forming expectations of
  exchange rates.

26
Summary

• Important policy implications
• If agents form their expectations extrapolatively
  then a policy of leaning against the wind may
  be beneficial.
• That is, a forex market intervention attempting
  to break a trend in the market.
• The existence of a risk premium that assets
  domestic and abroad are not perfect substitutes,
  and that interest rates in any country may not be
  identical to those abroad, even with no
  particular expectations of spot rate changes.
• It also implies that sterilised intervention may
  work.

27
Asset approach

• 3. Asset approaches - Portfolio models
• Consider 3 assets that investors hold and
  diversify (imperfect substitutability - UIP does
  not hold)
• M Money
• B domestic bonds
• F foreign bonds
• Well-defined asset-demand function
• Function of expected rate of return (on both the
  asset and its various substitutes)
• Expected rate of return of foreign assets
  defined as i expected rate of depreciation of
  domestic currency.
• Function of wealth implies that changes in
  price of the assets, and changes in S, will
  affect assets demand.

28
Portfolio approach

• Portfolio models - Specification
• Static expectations (? Se 0)
• Macroeconomic model in an open economy price
  flexibility, full employment (results may vary
  the way the real economy is modelled)
• The model distinguishes between short-run and
  long-run exchange rate determination.
• Short-run depends of investors preferences
  between foreign and domestic assets S changes
  to insure that assets markets are in equilibrium.
• Long-run role for the real sector, and in
  particular, the current account. The short-run S
  determines the current account, which, in turn,
  represents net foreign asset accumulation, that,
  again, causes S to change.
• Stability is reached when the current account is
  in equilibrium.

29
Portfolio approach

• Portfolio models - Specification
• Differentiates between stocks and flows (1)
• It is the variation of the differential of
  interest rates that determines capital flows, not
  the absolute difference (Mundell-Fleming model).
• Differentiates between stocks and flows (2)
• Implications of current account imbalances for
  asset accumulation
• Consider a country with a current account surplus
  (XgtM)
• If S are floating a current account surplus
  will be accompanied by a financial account
  deficit (FOgtFI, financial outflows), so that the
  BOP sums up to zero.
• This capital account deficit will increase the
  investors (residents) holdings of foreign assets.
• Then, a current account surplus implies an
  increase in residents holdings of foreign assets.

30
Portfolio approach

• Portfolio models - Graphical Equilibrium

S
B
M
F
Seq
F
M
B
i
i0
31
Portfolio approach

• Portfolio models - Graphical Equilibrium
• MM equilibrium on the Money Market
• Positive slope, since a rise in S increases
  wealth (by increase of SF - the value of holdings
  of foreign bonds) and thus the demand for Money,
  and i should rise to restore equilibrium.
• BB domestic bond market equilibrium
• Negative slope, because a rise in S generates
  excess demand on bonds, raise their prices, so i
  falls.
• FF foreign bond market equilibrium
• Negative slope, for the same reason as BB, but
  less steeper, since the impact of a rise in
  demand is less strong on foreign bonds than on
  domestic bonds (imperfect substitutes, preference
  for national investments).

32
Portfolio approach

• Portfolio models - Empirical evidence
• Difficult to test due to important data problems
• Good supporting evidence for the tests run
• But bad performance at forecasting (in
  particular, it does not outperform the random
  walk)
• Econometrical problems could explain this
  failure, like poor data and poorly specified
  dynamics.

33
News approach

• Testing models - News approach
• Try to distinguish between expected / unexpected
  components of exchange rates determinants
• Models sensitive to the way news are constructed,
  and to the choice of the type of news
• Poor empirical performance
• -gt research question what type of news is
  important to influence expectations on exchange
  rates?

34
Misalignments

• Recent attempts to explain misalignments
• Misalignment departure of exchange rate from
  its long-run equilibrium
• Two types of explanations
• Rational bubble
• Still assuming rational behaviour of markets
  participants.
• Pt discounted cash-flows Bt
• Bt E(Bt1) / (1r) bubble component
• Bt1 (1r) Bt Zt
• Bubble has a probability of bursting at each
  period, but grows at an expected rate of r if
  investors are risk neutral.
• Testing for evidence joint hypothesis of bubble
  existence and of the model of exchange rate
  determination.

35
Misalignments

• Recent attempts to explain misalignments
• Rational bubble
• Mixed empirical evidence for speculative bubbles
  on the exchange rate markets
• Theoretical questions
• Since prices are bounded to zero, a negative
  bubble could never starts
• With rational expectations hypothesis, a bubble
  should begin at the start of the asset life
• An asset with a finite life and a fixed
  redemption value cannot become the object of a
  rational bubble
• Need for the hypothesis of limited rational
  expectations and near rational bubble there
  is always a greater fool out there...

36
Misalignments

• Recent attempts to explain misalignments
• Heterogeneous expectations
• Wide dispersion of opinions observed, in
  particular for longer maturities
• Model of two groups of forecasters (Frankel
  Froot, 1987)
• Chartists extrapolate past experience
• Fundamentalists using Dornbushs overshooting
  model
• Portfolio managers use a weighted average of
  these two forecasts, and update the weights
  according to who is doing better.
• Broad empirical support explained the rise and
  fall of the dollar in early 1980s.
  Questionnaires among forecasters supported the
  approach.
• -gt still at an infant stage.