As the euro has plummeted against the USD, there’s been concern that efforts to rebalance the global economy will face increasing headwinds. [Bergsten] [Duy]. This worry is only added to by the already widening US trade deficit . In this post, I don’t want to dispute the difficulty of effecting global rebalancing.
It was already a difficult task, even before the euro area’s recent debt-related travails. What I do want to do is to put the recent exchange rate movements in perspective. My three observations are as follows:
- The euro is a relatively small component of the US trade weighted exchange rate.
- The persistence of exchange rate movements is important in determining the impact on trade flows.
- Rebalancing was never going to be effected by exchange rate re-alignments alone.
The euro and the dollar. First, here is a graph of the dollar/euro exchange rate expressed in units of USD per EUR.
Figure 1: USD/EUR exchange rate, monthly averages (blue line); synthetic euro before 1999M01. USD/EUR exchange rate on 6/4/2010 (blue square), Deutsche Bank forecasts as of 6/4/2010 (red squares) and forward rates (green triangles). Source: Fed via FREDII, Deutsche Bank, Exchange Rate Perspectives, June 8, 2010 [not online]. Clearly, the euro has declined precipitously since late last year. As of June 4, the USD/EUR rate was 1.2. Forward rates (which have little predictive power, and usually point in the wrong direction  ) imply no change. The Deutsche Bank forecast is for strengthening against the dollar, rising to 1.35 USD/EUR in a year’s time.
Clearly, that’s just one forecast, and given the uncertainties regarding the resolution of the euro area’s fiscal problems, there is little reason to put too much credence on this particular forecast. But this brings me to the second point. Despite the euro’s depreciation, the dollar has exhibited much less movement on a real, trade-weighted, basis.
Figure 2: Log broad trade weighted real USD (blue bold), CNY (red) and EUR (green). Source: BIS and author’s calculations.The data on this graph only extend up to 2010M04. The USD in nominal terms, on a broad basis, has appreciated by 2.8% (log terms, not annualized) in May. Still, that puts into perspective the fact that the USD is roughly where it was on the eve of the Lehman collapse.
Persistence. Deutsche Bank forecasts a 10% depreciation in the DB US trade weighted dollar (10.7% in log terms) in one year’s time. (I know that simple univariate tests cannot typically reject the null that the trade weighted dollar follows a unit root process, but there seems to be some evidence of threshold cointegration of the nominal exchange rate with price levels, which is consistent with stationarity in the real rate — see this post.)
Further note that the real, trade-weighted euro is not particularly weak by historical standards. And the Chinese yuan, interestingly enough, is pretty close to a trend line drawn from 2005m07 to 2010m04. (Which is not to say it is at equilibrium, or that more appreciation wouldn’t have a salutory effect on the Chinese and global economies — see this post).
Exchange rates and rebalancing. I’m currently working to update my research on trade elasticities (which explains the brevity of this post). One point that comes out is that econometrically, the exchange rate impact on trade flows is pretty small; price elasticities seldom exceed 0.5 even in the long run (short run can be much different). This highlights the point that in rebalancing, growth effects, and composition of GDP effects, might be of comparable importance to relative price changes effected via exchange rates (almost heresy for an open economy macro guy, but there it is). I mentioned this point in a series of posts:     . We need consumer retrenchment in the US, increased consumption/GDP ratios in China, as well as in other perennially surplus countries (e.g., Germany, Japan). To the extent that fiscal retrenchment occurs in the euro area, that will make rebalancing more difficult than otherwise, at least in the short run.
But as I say, nobody ever said rebalancing would be easy .
Originally published at Econbrowser and reproduced here with the author’s permission.