The currency composition of official foreign exchange reserves are a FX market focus given the impact that trends in reserve management can have on currency performance. There has been speculation that the divergence between the US dollar and relative interest rate developments is partly a result of changes in the currency composition of reserve portfolios against a background of rising trade tensions. The IMF currency composition of official foreign exchange reserves (COFER) data provides a window into recent developments and whether this speculation is justified.
Unfortunately, the data over the past three years is less informative than usual because China has been revealing the composition of its foreign exchange reserves in undisclosed amounts and allocations. The impact of this is observable in the split between allocated and unallocated reserves from the total data set, where allocated reserves are now approaching 90% of total reserves. This development should increase the usefulness of the COFER data going forward, but makes it difficult to draw precise conclusions from the data over recent quarters. We nevertheless view the recent data as supportive for our constructive view on the RMB, but feel it raises some questions for the market consensus view on the Euro.
Euro reserve allocations have risen from 19.1% at the end of 2016 to 20.1% at the end of 2017. However, reserve allocations are measured in US dollars and when we adjust for valuation effects (the Euro appreciation by ~14% against the US dollar in 2017), it becomes clear that rather than buying Euro in 2017, central banks likely used the rise in the Euro to rebalance their holding back to the US dollar. Or put another way, the rise in the Euro in 2017 was as a result of the Eurozone’s current account surplus and private capital flows, rather than reserve managers changing the currency composition of their reserve portfolios in the Euro’s favour. There is speculation that reserve portfolio buying of EURUSD was responsible for its sharp rise in early 2018. The lack of evidence of such shifts in 2017 and no compelling evidence from the January US Treasury International Capital (TIC) data raise uncertainty as whether this was indeed the case. We await the Q1 2018 COFER data (released at the end of Q2 2018) with interest.
Two other trends in the COFER data are worth commenting on: