# Pilbeam Chapter 6 João Ejarque University of Lund November 2009

Document Sample

```					Pilbeam: Chapter 6

João Ejarque
University of Lund
November 2009
PPP: Purchasing Power Parity
• So far we looked at IS/LM/BP models where AS is horizontal. In
these models the exchange rate is simply a variable that
somehow adjusts to equilibrium (or is fixed at some value).
There is no clear theory of how the exact value of the exchange
rate is found.
• The monetary approach to the BP thinks long run and assumes a
vertical AS. It also has an explicit theory of exchange rate
determination. It is the PPP, where P = SP*. With their aggregate
demand (money market) assumption this shows us that
– S = *M/M*+ / *(ky)/(k*y’)+
– Excess supply of money at home makes S increase (depreciation).
• However, is the PPP a good theory of exchange rate
determination? This is what chapter 6 is concerned with.
– It turns out the PPP is at best a reasonable long run theory, but a rather
useless explanation for short run behaviour of S. For better short run
theories we look to chapters 7 and 8.
PPP: Purchasing Power Parity

• The PPP is also an old theory. When countries first floated their
exchange rates it was believed that goods arbitrage would be a
dominant force.
• The Bretton-woods system collapsed in 1971, after the United
States unilaterally terminated convertibility of the dollar to gold.
• At that time capital movements were of comparatively smaller
importance and the current account was still a major source of
international capital flows.
PPP: Purchasing Power Parity
• A restatement of the law of one price
–   Assumes perfect competition and no barriers to trade
–   The law of one price is a single good theory.
–   In exchange rates it is applied to baskets of goods.
–   Exchange rates will adjust to eliminate arbitrage between baskets of
goods.
• Absolute PPP (strong)
– P = SP*
– Here P and P* are the price indices for the same basket of goods. Note
that goods belonging to the same basket of goods may carry different
weights in different countries.
• Relative PPP (weak)
– (ΔP/P) = (ΔS/S) + (ΔP*/P*)
– The exchange rate will adjust to eliminate differences in inflation rates
(depreciation is the result of excess inflation at home relative to abroad).
– Relative PPP is compatible with P = λSP*, where λ is a constant.
PPP: Purchasing Power Parity
• Traded versus non-traded goods
–   Assume PPP holds for traded goods, PT = S P*T, and define:
–   P = α PN + (1-α) PT
–   P* = β P*N + (1-β) P*T
–   Divide P/P*
–   And then divide the numerator by PT and the denominator by SP*T.
–   Get our λ factor!

 PN 
    1   
P 
P
S    T
P *
 PN 
*
  *   1   
P 
 T 
Note: the λ factor may not be constant. Preferences (α,β) may well vary over time.
PPP: Purchasing Power Parity
• Traded versus non-traded goods (issues)
– Non traded goods are also subject to competitive pressures from abroad: a
barbershop pays rent, and a competitive real estate market will incorporate the
marginal value of office space for a tradable goods company.
– Again, goods belonging to the same basket may carry different weights in different
countries. The relative importance of the non traded sector changes from country
to country.
• Suggestion:
– Measure the impact of house prices on the empirical performance of the PPP. Use
UK and German data? Note also that within the Eurozone house prices evolved
differently for example in Germany and Spain. (How differently?)
• Empirically:
– PPP performs well as a long run indicator but fails completely in short run fit.
– PPP performs better for countries that are geographically close to each other, and
countries with lower tariff and trade barriers.
– Exchange rates are far more volatile than national price levels.
Regressions (using tradables only)
Null Hypothesis: a0 =0, a1 = 1

T
log( S )  a0  a1 log P   log P *  
T             

T
 log( S )  a0  a1  log P    log P *  
T                
S is the nominal exchange rate. P is often the CPI but the PPI works better.
Graphical evidence on PPP
The Dollar/DM Rate and Relative U.S./German Price Levels, 1964-2000

Note: PUS and PG are indices  graph is about relative PPP.
Conclusion: PPP does not hold in short-run, but maybe in long-run.
8
PPP: Purchasing Power Parity
• Empirical failure:
– Trade costs
– Imperfect competition
– Exchange rate overshooting (goods prices are slow to adjust and the
exchange rate reacts to many other factors)
– Productivity differences and non traded goods (NEXT: BS model)

• Balassa-Samuelson Model
• Facts
– PPP prices are higher in rich countries
– Non traded goods prices are more different across countries
than prices of traded goods
– Non traded goods prices are higher in rich countries
PPP: Purchasing Power Parity
• Balassa-Samuelson Model
• Facts
– PPP prices are higher in rich countries
– Non traded goods prices are more different across countries than prices
of traded goods
• Model
–   Production technologies (fixed coefficients, only labor)
–   PN = WN/QN and PT = WT/QT ,
–   PN* = WN*/QN* and PT* = WT*/QT* ,
–   Market clearing in each country
–   WN = WT and WN* = WT*,
–   Technology differences (foreigners are more efficient producing tradables)
–   QT* > QT but QN* = QN ,
–   Assume PPP holds for traded goods:
–   PT = S PT*
PPP: Purchasing Power Parity
• Balassa-Samuelson Model Implications

– The country with the more efficient tradable goods technology (the rich
country) has a higher relative price of non tradables:
• PN*/PT* > PN/PT ,
– Cross country (cross section) variation of non tradables prices is higher
than the variation of tradables prices.
• PN*/PN > PT* /PT
– Non tradables are more expensive in poor countries (PPP fails for non
tradable goods).
• SPN* > PN ,
• The model of course does not explain why in the data PPP fails for tradables since it
assumes PPP holds for tradables.
– Wages are higher in rich countries
• SW* > W ,
Wikipedia Example: English Pub Beer
• A pint of Beer in a London Pub is more expensive than in a
Newcastle Pub.
• However, supermarket beer in London costs the same as
supermarket beer in Newcastle.
• Why?

– Supermarket beer is tradable. Pub beer is non tradable.

– Pub waiter productivity is the same in London and newcastle.

– However, productivity in the ”export” sector is different.
• Financial services in London are more productive than Industry in newcastle.

– All this with the same currency and a nominal exchange rate = 1.

```
DOCUMENT INFO
Shared By:
Categories:
Stats:
 views: 9 posted: 3/10/2010 language: English pages: 12