At the end of every quarter, the IMF reports official reserve holdings. It is the most comprehensive and authoritative source of such information. This is especially true of the currency composition of the reserves, which some countries publish freely and others regard as a national secret.
The IMF reports the figures with a quarter lag. Data as of the end of the second quarter have now been made available. It collected data from 146 participating countries. There were 145 countries in Q1 that participated, but of special note, starting with Q2 figures, China has formally accepted the international best practices and will report its reserve figures to the IMF on a quarterly basis.
The caveat is that China only reported part of its reserve holdings, claiming that they were broadly representative of its allocation. It indicated that over the next two years, it will include more of its reserves. This may set the conspiracy-oriented among us to wonder if China has the reserves it claims, as some do about various central bank gold holdings.
Occam's Razor would suggest a simpler explanation. By gradually reporting its reserves, China can minimize the loss of secrecy. It is Chinese diplomacy. Begrudgingly (in terms of how long it has taken China to adopt this rather straight forward reporting that the vast majority of countries practice) accept the convention, but do so in China's way.
It is not exactly clear how much of China's reserves were reported. It appears to be around $550 bln. Here is how we arrive at that estimate. Some countries report the currency allocation of their reserves and others simply report their overall reserves. The unallocated reserves rose by $580.2 bln. The allocated reserves rose by $604.8 bln.
A point we keep hammering is that the reserves are not simply quantities of money, but there is a valuation component that needs to be taken into account. The IMF explains in a technical note that countries report their reserves in terms of US dollars, and the preferred exchange rate is the quarter-end rate. The fluctuation of exchange rates impacts the dollar value of reserves. Broadly speaking, a rising dollar puts downward pressure on the value of reserves, while a falling dollar lifts the valuation of reserves.
However, the dollar holdings themselves are not subject to valuation shifts. This is a pure quantity. The amount of dollars held by those central banks that report the allocation of their reserves rose by $363 bln. If we assume that China's holdings of dollars were in line with the global average (of allocated reserves), which is 63.75%, it means China reported about $567 bln of the $3.69 trillion of reserves (as of the end of June). This is broadly consistent with our earlier estimate.
The US dollar fell against all the reserves currencies in the second quarter, except the Japanese yen, which slipped 1.9%. This means that the dollar value of reserve holdings has been bolstered by the dollar's decline. For example, the value of the reserves denominated in euros rose by $105 bln or 8.3% to $1.367 trillion. The euro appreciated in Q2 by 3.9%. That means that almost half or $49 bln of the increase in the (dollar) value of euro reserves can be explained simply by change in the euro-dollar exchange rate.
In addition to including Chinese figures in the currency allocation, the IMF made another critical change. Previously it broke down the reserve holdings by what it calls advanced economies and emerging and developing economies. Apparently without explanation, the IMF says this level of detail will no longer provide such data. While the motivation may not be clear there is a clear and practical consequence. It makes it harder to interpolate the allocation of China's reserves, which is consistent with the PBOC bleeding the currency allocation of its reserves into the IMF data gradually.
Nevertheless, some creative accounting forensics can still let us draw some additional though admittedly preliminary conclusions. If China's dollar holdings are in line with the global average, it appears that it may be underweight euros, but overweight sterling and the dollar-bloc.
The dollar value of sterling's reserves rose $312.8 bln from $237.3 bln of the allocated reserves. This represents almost a 32% increase. Sterling appreciated 6% against the dollar in Q2. Adjusting for the foreign exchange movements still gives sterling a large relatively large boost. It share of allocated reserves rose to almost 4.7% form 3.9% in Q1.
The dollar value of Canadian and Australian dollars rose 13.3% and 20.9% respectively. The currencies rose a 1.5% and 1.3% against the US dollar over the period, which means that the increased dollar value largely reflects new additions to the tune of $15 bln of Canadian dollar and almost $22 bln of Australian dollars.
The key takeaway from the IMF's Q2 COFER data is that China has begun reporting the allocation of its reserves. It has done so in a way that limits our ability to detect China's precise allocation, and the IMF appears to have cooperated by no long providing a breakdown between the advanced and developing economies. Nevertheless, we can make some preliminary assessments that suggest China may underweight euros relative to the global averages and may be overweight sterling and the dollar-bloc.
Overall global reserves rose by about $23.5 bln or 5.7%, a good part of which can be explained by the dollar's decline against the other reserve currencies except the yen. The dollar holdings, which are not impacted by foreign exchange fluctuations, rose by $363 bln though the dollar's share of allocated reserves slipped to 63.75% from 64.12%. The euro's share slipped lower to 20.5%, since the initial days of the euro's launch.
Where does this leave us in terms of China's quest to be included in the SDR? First, adopting the IMF's best practices is helpful. Although it may not be a formal requirement, it seems it is operationally necessary. Second, the IMF recently reported that the yuan's share of global reserves were about 1.1% (dollar value of about $120 bln). This lower than Australian and Canadian dollars (1.9%), but above the Swiss franc's 0.3% of global (allocated) reserves.
Most discussions see the SDR decision as binary. It is either included or it is not. However, this appears to have led to under-discussion of an important aspect of the issue. If the yuan is included, what will be its weighting? The currency's weight in the SDR basic is a function of the importance of its exports and its role as a reserve currency. China ranks high on the former but low on the latter.
The current configuration, set in 2010, gives the dollar a 41.9% weight, the euro 37.4%, sterling 11.3% and the yen 9.4%. If the yuan is included, it would seem consistent with the gradualism that characterizes such decision-making to be the least disruptive as possible. That would mean giving it a modest weight to start, may be less the yen's current weight, with an eye toward increasing it in the future as the yuan's ease of access and reserve status increases in the coming years.