Figure 1: US dollar (blue, right scale), US dollar plus 60% of unallocated reserves (green, right scale), and log nominal value of US dollar against major currencies (red, left scale). NBER defined recession dates shaded gray. Source: IMF, COFER, June 30, 2009, Federal Reserve via FREDII, NBER and author’s calculations.
Notice that the USD share has not declined, despite a decline in the dollar’s value against major currencies. Following Brad Setser’s observation that the reason the demand for the dollar as a reserve currency rose is because total demand for reserves increased, I also plotted the levels — rather than shares — for the most recent data.
Figure 2: US dollar reserves (blue), US dollar plus 60% of unallocated reserves level (green), and total reserves (black), in millions of US dollars. NBER defined recession dates shaded gray. Source: IMF, COFER, June 30, 2009, NBER, and author’s calculations.
I think it’s an interesting that reserves have been shrinking for the past three quarters — and at a pretty rapid clip. They were declining by an annualized 10.3% in 2009Q1 (q/q in log terms; 9.8% in base terms). This development suggests that, even if the dollar retains its share of total reserves, demand for dollar assets might still decline.
While this might constitute a secular force for dollar weakness, it’s important that there are forces in working in the other direction, including cyclical factors. Deutsche Bank for instance projects 12.7% appreciation in the DB dollar index by end-2009 (16.1% against the euro, both in log terms). Their forecast implies only a slight depreciation in the dollar index by end-2010, and further appreciation against the euro (20.4% relative to June 26).