Topic 4: Currency Valuation

Prices and Exchange Rates - PPP and the RER

Dr. Sushanta Mallick

http://webspace.qmul.ac.uk/skmallick/

12 August 2010

S K Mallick (QMUL) International Finance 12 August 2010 1 / 24

Currency Valuation
Learning Objectives

What is the fair value of the euro, or the pound or the yen?
Once we know this, we can compare actual spot rates with these
equilibrium or fair values and …nd the overvalued and undervalued
currencies
We can then buy the undervalued and sell the overvalued as our
investment strategy on the assumption that currencies move back
towards equilibrium over our investment horizon
1 Another arbitrage relationship, PPP - the role of goods markets in
international …nance (as distinct from …nancial asset markets)
2 The exchange rate as the relative price of two monies
3 The asset approach to the exchange rate
4 Performance of alternative models of fair value

S K Mallick (QMUL) International Finance 12 August 2010 2 / 24

Purchasing power parity
Law of one price (LOOP)

Arbitrage = the pro…t is known with certainty ! no risk

In contrast, speculation trading is done with an expectation that a
pro…t will be earned 7! risky
Theoretical macro-economic relationships a¤ecting the exchange rates
prices or price changes, interest rates, forward exchange rates

The LOOP states that identical goods sell for the same price
worldwide.
A laptop bought in Paris should cost the same as one bought in New
York when expressed in the same currency.
No arbitrage condition: PitUS = PitEU St$/e
The domestic price and the foreign price of a good are equal when
converted to the same currency

S K Mallick (QMUL) International Finance 12 August 2010 3 / 24

EXAMPLE
EU basket costs e100; US basket costs $120
Nominal exchange rate, S$/e , is $1.20/e1 (the rate at which we can
trade the currency of one country for the currency of another)
spot exchange rate is determined by the relative prices of similar
baskets of goods

The real exchange rate is: q$/e = ($1.20/e1 e100 per EU basket)
/ $120 per US basket (the rate at which we can trade the goods and
services of one country with the goods and services of another)
q$/e = 1 US basket per EU basket [a unit of home currency has the
same purchasing power in a foreign country]
Example: a car in Finland versus US
According to the LOOP:,Pe = Se/$ P$ ; So, S = price of the car in
Finland / price of the car in US = 56,000/36,469 = 1.5356 e/$
This is absolute PPP using just one good.
S K Mallick (QMUL) International Finance 12 August 2010 4 / 24
Purchasing power parity
Purchasing Power Parity (absolute version): The LOOP aggregated across
all n goods in an economy
n n
$/e $/e
∑ αi PitUS = St ∑ αi PitEU ) PitUS = St PitEU
i =1 i =1
In logs, st = pitUS pitEU
We can de…ne the (log) real exchange rate, q$/e as:
qt = st pitUS + pitEU
US EU
And in terms of changes: q = s p p

where: q = change in the real exchange rate

s = change in the nominal exchange rate expressed as the domestic
price of the foreign currency at time t
US EU
(p p ) = change in relative prices
If PPP holds over time then q = 0 =) percentage changes in
exchange rates equal percentage changes in prices or in‡ation
(Relative PPP) between two periods.
S K Mallick (QMUL) International Finance 12 August 2010 5 / 24
RELATIVE PPP states that any change in the in‡ation di¤erential
between two countries is o¤set by an equal but opposite change in
the spot rate
Currencies with high in‡ation should depreciate relative to currencies
with low in‡ation
Often measured in terms of changes in the value of a basket of goods

Example: suppose Mexico has 10% in‡ation and India has an in‡ation
rate of 5%. After one year, one Mexican peso would buy less than
one unit of Indian rupee; We would expect the value of peso to
decline relative to the rupee.
If US in‡ation is higher than that of Euroland’s, then the dollar’s real
value rises (decline indicates appreciation), loss of US competitiveness.
q $ = s $/e p US p EU

If s $/e declines, then the dollar’s real value increases, loss of US

competitiveness
q $ = s $/e p US p EU
S K Mallick (QMUL) International Finance 12 August 2010 6 / 24
Purchasing Power Parity

Absolute PPP states that exchange rates equal the level of relative
average prices across countries
Relative PPP states that the percentage change in exchange rate
should be approximately equal to in‡ation rate (percentage changes
in prices) di¤erential between two countries

EXAMPLE
US price level rises 10% over a year
EU price level rises 5% over a year
Relative PPP predicts a 5% depreciation of the dollar against the euro
so that
it cancels the 5% by which US in‡ation exceeds EU
purchasing powers of both currencies remain the same

S K Mallick (QMUL) International Finance 12 August 2010 7 / 24

Does PPP hold in the real world?

If PPP is expected to hold, then the best prediction for the

US
1 +p
one-period spot rate (st ) should be: st = s0 EU
1 +p

PPP is the application of the LOOP across countries for all goods
and services. If PPP is a valid equilibrium concept, then it can be
used as an indicator of fair value

In empirical studies, PPP has not received much support

The only empirical merit for PPP is in long term forecasting
Its usefulness is limited because of the extremely slow speed of
reversion to equilibrium
The equilibrium real exchange rate may be shifting, producing
systematic divergence from PPP (in part due to productivity
di¤erentials)

S K Mallick (QMUL) International Finance 12 August 2010 8 / 24

The real exchange rate

The real exchange rate measures the cost of one country’s goods in
terms of another.

Real dollar depreciation: A dramatic example of the real depreciation

of the dollar in terms of the euro has been the increase in the dollar
price of the euro from $0.90 to $1.55 from 2005 to 2008, while there
has been little di¤erence in the in‡ation rates. The euro has thus
experienced a real appreciation, while the dollar has su¤ered a real
depreciation.

Naturally, there are many currencies, so the best RER measurement is

one that takes into account the weight of each currency in terms of
trade. The IMF provides such a calculation.

S K Mallick (QMUL) International Finance 12 August 2010 9 / 24

The real e¤ective exchange rate
The IMF’s real e¤ective exchange rate (REER) measures the price of
a home country’s goods in terms of foreign goods
It is calculated as a weighted average (geometric mean) of bilateral
real exchange rates using the annual value of a country’s trade with
other countries as its weights. It is then converted into an index using
a base year.
n wi
P
REER = 100 ∏ Si Pi
i =1
w1 w2 w3 wn
P P P
= 100 S1 P1 S2 P2 S3 P3 ... SnPPn
where ∏ : the product of the i terms
P : the consumer price index (CPI) of the home country (USA); Pi :
the CPI in country i
Si : the home currency (USD) price of currency i; wi : the currency
weights of the di¤erent countries.
n wi
1
NEER = 100 ∏ Si
i =1
S K Mallick (QMUL) International Finance 12 August 2010 10 / 24
The real e¤ective exchange rate I

When the home country’s price rises relative to other countries, a real
appreciation takes place. It costs fewer home goods to buy foreign
goods.

When foreign in‡ation is higher than domestic in‡ation, a real

depreciation occurs. It costs more home goods to buy foreign goods.

If changes in exchange rates just o¤set di¤erential in‡ation rates – if

purchasing power parity holds – all the real e¤ective exchange rate
indices would stay at 100
If a rate strengthened (overvalued) or weakened (undervalued) then the
index value would be ?100

S K Mallick (QMUL) International Finance 12 August 2010 11 / 24

Stylized facts on the Indian NEER and REER
20/1/10
105

100

95

90

85

80
93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09
IN INDIA RUPEE - NOMINAL EFFECTIVE EXCHANGE RATE INDEX NADJ
IN INDIA RUPEE - REAL EFFECTIVE EXCHANGE RATE INDEX NADJ
Source: Thomson Datastream

Trend nominal depreciation

Indian In‡ation is higher than in trading partners
Broadly stable REER (mean-reverting but at a slow speed).
S K Mallick (QMUL) International Finance 12 August 2010 12 / 24
Empirical Evidence
What do we know about the empirical validity of PPP?

Changes in national price levels do not tell us much about changes in

the nominal exchange rate
The choice of the price index: Based on indices and the weights
attached to similar goods in aggregate price indices, they might di¤er
across countries

LOOP also does not do very well

LOOP implies that PPP holds for traded goods
Equilibrating the purchasing power of incomes (nominal exchange rate
adjusted for in‡ation di¤erential) implies PPP holds for a broad price
index

Relative PPP does better but still not a good indicator of exchange
rate changes
Departures from PPP can be higher in the short-run than in the
long-run due to price stickiness
S K Mallick (QMUL) International Finance 12 August 2010 13 / 24
Shortcomings of PPP
1 A¤ected by transportation costs - The greater the transport costs, the
greater the range over which the exchange rate can deviate from its
PPP value

2 The border e¤ect [Engel and Rogers (AER, 1996)] - the distance
between cities can explain a considerable amount of the price
di¤erential of similar goods in di¤erent cities of the same country.
This is even larger for two cities in di¤erent countries.

3 Goods arbitrage can be a¤ected by trade barriers (tari¤s and duties) -

governmental trade restrictions make trade expensive and in some
cases create non-tradable goods or services

4 Importance of non-traded goods (Parsley and Wei, 2004): Prices of

nontradable goods and services are determined by domestic demand
and supply curves. So they are also not linked internationally allowing
systematic deviations from PPP
S K Mallick (QMUL) International Finance 12 August 2010 14 / 24
Does PPP hold?
Early empirical literature on PPP

PPP states: the currency with the higher in‡ation rate is expected to
depreciate relative to the currency with the lower rate of in‡ation
Until the late 1970s, the PPP relationship was tested with a
regression model of the following form:

s t = α + β 1 pt + β 2 pt + e t
The hypothesis is: β1 = 1; β2 = 1, which would be interpreted as a
test of absolute PPP, while a test of the same restrictions with the
variables in …rst di¤erences would be interpreted as a test of relative
PPP.

Evidence
PPP holds for hyperin‡ation countries (H0 is not rejected)
PPP is rejected for low in‡ation countries (H0 is rejected)
Problems
Spurious regressions and Endogeneity problems
S K Mallick (QMUL) International Finance 12 August 2010 15 / 24
Does PPP hold?
Long-run relationships

Does PPP hold as a cointegrating relationship?

st = α + β 1 pt + β 2 pt + e t
The hypothesis is: β1 = 1; β2 = 1
Is q (the real exchange rate) stationary?
qt = α + β 1 qt 1 + e t
If β1 = 1, q does not revert to an average level and PPP does not hold
Most experts agree that PPP holds in the long-run even if the
empirical evidence does not support that
I Shapiro, A. (2009), Multinational Financial Management, 9th Edition,
J.Wiley & Sons, Chapter 3
I Robin, J.A. (2010), International Corporate Finance, 1st Edition,
Chapter 5.
I Lucio Sarno and Mark Taylor (2002) The Economics of Exchange
Rates, Chapter 3.
S K Mallick (QMUL) International Finance 12 August 2010 16 / 24
Why might PPP not work?
Arbitrage only works for tradeable goods
How about services (non-tradeables)? Haircuts cost 10 times as much
in the developed world as in the developing world

Barriers to trade: tari¤s and quotas can lead to deviations from PPP
Transportation costs
Many goods are not perfect substitutes
Cross-country di¤erences in tastes and preferences
Imperfect exchange rate pass through
I Mallick, Sushanta, and Helena Marques (2008a), Exchange rate
transmission into industry level export prices: A tale of two policy
regimes in India, IMF Sta¤ Papers, 55 (1): 83-108.
I Mallick, Sushanta, and Helena Marques (2008b), Pass-through of
exchange rate and tari¤s into import prices of India: Currency
depreciation versus import liberalization, Review of International
Economics, 16 (4): 765-782.
S K Mallick (QMUL) International Finance 12 August 2010 17 / 24
Exchange Rate Pass-Through

Incomplete exchange rate pass-through (ERPT) is one reason why a

country’s RER can deviate from it’s PPP equilibrium point

The degree to which the prices of imported & exported goods change
as a result of exchange rate changes is termed as pass-through

Example: assume BMW produces a car in Germany and all costs are
incurred in euros. When the car is exported to the US, the price of the
BMW should be the euro value converted to dollars at the spot rate

$
PBMW e
= PBMW S $/e
Where P$ is the BMW price in dollars, Pe is the BMW price in euros
and S is the spot rate

S K Mallick (QMUL) International Finance 12 August 2010 18 / 24

Exchange Rate Pass-Through I

Due to incomplete ERPT, a country’s RER can deviate for lengthy

periods from its PPP-equilibrium level

If the euro appreciated 20% against the dollar, but the price of the
BMW in the US market rose to only $40,000 from $35,000, and not
$42,000 as is the case under complete pass-through, the pass-through
is partial

The degree of pass-through is measured by the proportion of the

exchange rate change re‡ected in dollar prices
$
P BMW ,2 40,000
$ = 35,000 = 1.1429, or 14.29%
P BMW ,1

The degree of pass-through in this case is partial, 14.29% 20.00%

or approximately 0.71. Only 71.0% of the change has been passed
through to the US dollar price
S K Mallick (QMUL) International Finance 12 August 2010 19 / 24
The remaining PPP Puzzle
Balassa-Samuelson hypothesis

Real exchange rate appreciation for countries which are experiencing

increases in productivity levels relative to other countries

Reason: Higher productivity increases wages but no e¤ect on traded

goods prices (PT ), because the domestic price level is tied to the
world price level and thus the exchange rate. In non-traded goods
sector, the higher wages will be passed on to higher prices, i.e., PNT "
! P ".

Rich countries with higher labor productivity in the tradables sector

will tend to have higher nontradables prices (PNT ) and higher price
levels (P).

If US productivity rises relative to Eurozone, this will in turn cause a

real exchange rate appreciation of the dollar (a rise). That is, an
increase in PUS will lead to a real exchange rate appreciation.
S K Mallick (QMUL) International Finance 12 August 2010 20 / 24
General Price levels in the US (P us ) and Europe (P eu ):
us 1
P us = (PTus )θ (PNT θ
)
eu 1
P eu = (PTeu )θ (PNT θ
)
In logs: p us = θpTus + (1 us
θ ) pNT
p eu = θpTeu + (1 eu
θ ) pNT

The real exchange rate (RER):

q$/e = s$/e (p us p eu )
= s$/e (θpTus + (1 θ ) pNT
us θpTeu (1 θ ) pNT
eu
)
us eu
= (s$/e pT + pT ) + (1 θ ) (pT us eu us
pT ) (1 θ ) (pNT eu
pNT )
LOOP for T makes the …rst term zero
RER dynamics driven by PNT in terms of PT

S K Mallick (QMUL) International Finance 12 August 2010 21 / 24

BS e¤ect holds: Productivity statistics give some empirical support to
the Balassa-Samuelson di¤erential productivity proposition.
Sustained appreciation in Japan’s real exchange rate against the $ in
1970’s

Developed versus Less Developed countries

In traded goods, DC are more productive
In non-traded goods, both are equally productive
Productivity di¤erences are sharper in traded than in nontraded goods
PT is same across countries but lower productivity in traded goods in
LDCs leads to lower wages in non-traded sector and thereby lower PNT
and lower P. So PLDC < PDC .

Countries with high Government expenditure can also experience real

exchange rate appreciation

S K Mallick (QMUL) International Finance 12 August 2010 22 / 24

PPP and Investment Strategy

PPP-determined exchange rates still provide a valuable benchmark

If you have long horizon and substantial risk capital

Purchase undervalued currencies; undervaluation will be corrected via

an appreciation sometime in the future

Sell overvalued currencies (take short position)

Take more aggressive (unhedged) positions as the deviations from PPP

grow and persist

S K Mallick (QMUL) International Finance 12 August 2010 23 / 24

Conclusion

When economic changes are in‡uenced only by monetary factors, and

when the assumptions of PPP hold, nominal exchange rates are
determined by relative PPP in the long run

When economic changes are caused by factors that a¤ect real output,
exchange rates are not determined by PPP only, but are also
in‡uenced by the real exchange rate

This long-run exchange rate model will be used even when we discuss
short-run macroeconomic events
long-run factors are important for the short-run because of the central
role expectations (about the future) play in day-to-day determination
of exchange rates

S K Mallick (QMUL) International Finance 12 August 2010 24 / 24