Purchasing Power Parity And Floating Exchange Rate Experience PDF

Purchasing Power Parity and Floating Exchange Rate Experience Pilbeam, chapter 6 Prof. F. Klaassen  What you will learn Goods prices matter for the exchange rate in the long run: higher inflation in H vs. F comes with depreciation of H currency vs. F. How g...

Purchasing Power Parity and Floating Exchange Rate Experience Pilbeam, chapter 6 Prof. F. Klaassen  What you will learn Goods prices matter for the exchange rate in the long run: higher inflation in H vs. F comes with depreciation of H currency vs. F. How good is P/P* at forecasting S? Short run: terribly bad Long run: some evidence that P/P* is useful (PPP makes sense). High labor productivity in producing traded goods causes high aggregate price levels (Balassa-Samuelson):  Useful to explain: rich countries have higher price levels: deviation from PPP. 2/28  Motivation Earlier chapters: S matters for BoP and policy  Important to know how S comes about.  See chapters 6-8: theories on exchange rate determination. (Note: chapters 6-8 concern a floating exchange rate, obviously.) 3/28 Theories of exch. rate determination Chapter 6 Exch. rate determined by goods market purchasing power parity (PPP) Empirical testing of PPP incomplete explanation of exchange rate. Chapter 7 Add money market and bond market monetary models Two main theories: flexible and sticky prices Assumption: domestic (H) and foreign (F) bonds are perfect substitutes Empirical testing substitutability assumption is sometimes rejected. Chapter 8 Allow for imperfect substitutability of H&F bonds (risk premium). 4 Idea behind PPP: law of one price  Law of one price Arbitrage equalizes goods prices internationally  Need to compare prices internationally: use exchange rate to convert foreign price into domestic currency  For good g: Pg = S Pg*: “Law of one price” (LOP). Equilibrium restoration: if Pg > S Pg*, then only demand F good more demand F currency S. 6  Example: Big Mac parity The Economist has collected Big Mac prices around the world  Compute Big Mac exchange rate, the rate that equalizes Big Mac prices: SBM=PBM/P*BM: a reference/equilibrium exchange rate. Useful for measuring undervaluation of domestic currency: (S-SBM)/SBM Positive if S>SBM, as should be for undervaluation of H currency. Examples (more in tutorial): July 2023 – S = 0.91 €/$ – SBM = €5.28/$5.58 = 0.95 €/$  Undervaluation of euro is -4% (i.e., overvaluation of 4%). 7/28  Big Mac parity 1999-2011 Euro per dollar exchange rate and Big Mac rate (red) 1.3 1.2 1.1 1 0.9 14% 0.8 0.7 0.6 0.5 01-1999 01-2001 01-2003 01-2005 01-2007 01-2009 01-2011 In 2nd half: actual price of a dollar is lower than Big Mac reference price Dollar is undervalued (euro is overvalued). 8/28 Data sources: Datastream (blue) and The Economist (red)  Using Big Macs for exchange rate? Critique on Big Mac parity: Prices are distorted by taxes and price discrimination There exist reasons for price differences Trade in Big Macs is hampered by transportation costs Arbitrage mechanism does not work well  Prices are not equalized perfectly. True, but red Big Mac line isn’t that bad; if Big Mac rate were unrelated to S, the line could have been around 10, say  Big Mac rate is truly related to S, but we cannot say how strong the relation is. 9/28  Comparing Big Mac to... iPod: arbitrage works better  iPod exchange rate is SiPod=PiPod/P*iPod. Euro/dollar exch. rate, Big Mac (red) and iPod rates iPod touch, 29-1-2011: 1.3 €229 in Germany 1.2 $229 in US SiPod=1 €/$. 1.1 1 0.9 27% 0.8 0.7 0.6 0.5 01-1999 01-2001 01-2003 01-2005 01-2007 01-2009 01-2011 Euro overvaluation of 27%: in line with Big Mac message. 10/28 Data sources: Datastream (blue) and The Economist (red)  PPP and its variants Timers  From LOP to PPP If LOP holds for all g in the economy [i.e., Pg = S Pg*], then for the bundle of all g, with aggregate prices P and P*: P = S P*: this is PPP  Purchasing power of €1 is the same in H and F, 1/P resp. 1/(SP*). Other interpretation: S=P/P*: exchange rate equals the ratio of H and F price levels. Note: S=P/P* is not how the exchange rate is defined; it is just the value according to the PPP theory [definition of S is the number of home currency units per unit of foreign currency, so has nothing to do with goods prices P&P*] PPP does not imply LOP, as deviations from LOP for some goods may offset each other when aggregating, so that PPP can still hold. 12 PPP is old idea: Cassell (1920s) or even Ricardo (1810s). According to PPP  Absolute and relative PPP (1/2) PPP concept of previous slide is actually absolute PPP: P = S P*. Problem: P and P* are not published in levels, only in index form (P/P0); think of CPI  Absolute PPP cannot be verified  Need additional PPP concept. 14/28  Absolute and relative PPP (2/2) Alternative: consider changes over time, i.e., relative PPP: dP/P = dS/S + dP*/P*. Note: dP/P is observed, since dP/P=d(P/P0)/(P/P0): no need to know price levels absolute-PPP problem has been avoided. Other interpretation: dS/S = dP/P – dP*/P*: depreciation is excess inflation in H over F  Higher inflation comes together with depreciation. 15 Relations between abs. and rel. PPP Pilbeam uses several other PPP formulas; here they are collected. Switch to logs for convenience: Absolute PPP: p=s+p*, where p=log(P), s=log(S) Relative PPP: dp=ds+dp*. Encompassing form: p = s + p* + c (c is a constant) Absolute PPP: c=0 Relative PPP: c is unrestricted. Real exchange rate (in logs) is sr = s+p*-p: Absolute PPP: sr=0 Relative PPP: sr is unrestricted constant.  If abs. PPP holds, then rel. holds too: abs. is more restrictive than rel. 16/28 Is difference absolute vs. relative PPP clear?  PPP in reality Timers  Evidence on long-run relative PPP 82 countries vs. US Relative PPP is useful in long run. Source: Feenstra and Taylor: International Economics, first edition  Loooong run: 200 years of CPIUS vs. CPIUK$  PtUS  CPI US t  log  US   P1900   PtUK St  CPI UK $ t  log  UK  P S  1900 1900  St is $/£ - exchange rate.  CPIs show similar developments: supports long-run relative PPP. Still, CPIUS seems to grow faster than CPIUK$: why? See later slide. Goods arbitrage increased in period, but no clearer co-movement at end than at begin.  Evidence of PPP is not conclusive yet. Source: Taylor and Taylor (2004): The Purchasing Power Parity Debate, Journal of Economic Perspectives Explaining deviations from PPP  Explaining deviations from PPP Economic reasons 1. Trade costs and impediments goods arbitrage ↓ 2. Tradables vs. non-tradables (see next slides) 3. Productivity differentials: Balassa-Samuelson (see next slides). 22/28  2: Tradables vs. non-tradables Basic mechanism behind PPP: goods arbitrage  Unrealistic for nontraded goods (non-tradables): PN ≠ SPN* even if PT = SPT* (N for no-tradables, T for tradables). Consequence for aggregate price levels: P ≠ SP*.  Finding P≠SP* may cause false conclusion of incomplete goods arbitrage  Also test PT = SPT* (so using just tradables) when testing for PPP. 23/28 Redo 200-years analysis: PPI instead of CPI  PPI series show stronger evidence of long-run relative PPP than CPI series before  consistent with fact that PPI has fewer nontradables than CPI: PPI is more like PT. Source: Taylor and Taylor (2004): The Purchasing Power Parity Debate, Journal of Economic Perspectives  3: Productivity differentials Take prices of baskets of a limited number of similar goods across countries  Scatter plot, where each dot is one country in 2010 US  Observation: Prich > SP*poor. China Real per capita income (in 2005 U.S. dollars) 25/28 Source: Krugman, Obstfeld, and Melitz: International Economics, eleventh edition Balassa-Samuelson explanation of Prich > SP*poor Idea: Take a rich country and compare it to a poor country. Prices of tradables are equal (due to international competition), but productivity for tradables is higher (in rich)  Wages in tradables sector are higher  Wages in non-tradables sector are higher (as labor mobility within country reduces labor supply in non-tradables)  Prices in non-tradables sector are higher (as productivity is equal)  Aggregate price level is higher: Prich > SP*poor. 26 Conclusion  Evaluation of PPP PPP is based on goods market arbitrage. Empirical validity: In short-run:(relative) PPP is invalid. In long-run: (relative) PPP is valuable. Conclusions: PPP is useful for long-run exchange rate determination PPP is incomplete explanation of exchange rate in short-run  add money/capital flows: next chapters. Do we now understand the “What you will learn”-slide? 28