What to do with Asian central banks’ FX reserves?

Our view is that there is little that the region’s central banks can do with their existing stock of reserves. In May 2011, roughly 33% of on-Japan Asia (NJA) reserves were parked in US treasuries. In level terms, holdings were USD1.65trn while total reserves were USD5trn. The number would be higher if holdings of agency debt and if China’s purchases of treasuries via offshore centres such as London are included.

There is little that Asian central banks can do with this existing stock of investment. Treasuries remain the most liquid asset with traded volumes exceeding NJA holdings of these treasuries. The dollar itself remains the dominant currency for trade settlement and for pricing of commodities. Furthermore, any large scale exit from treasuries will only backfire on regional central banks with respect to both capital and FX losses. It does make sense to maintain an alliance with the USD.

But what about deployment of incremental reserves? Here we believe the region’s central banks have a clear desire and motive to diversify. To an extent, it is also achievable if done gradually. The motive is related not only to the recent downgrade of the US sovereign rating but also to its monetary stance and the existing weak state of public finances. After all, US monetary base has tripled over the last three years while public debt to GDP now stands at almost 100% of GDP.

As a consequence, the USD has lost much of its lustre against gold. Going forward, the Fed has also signalled an unambiguous commitment to an exceptionally low interest rate environment. These concerns over US monetary policy have also been echoed by the region’s central banks. Our interactions with various central banks in the region also suggested a rising urgency to diversify.

Coming to the ability to diversify, historical trends do indicate a gradual change in the composition of reserves. Both the IMF COFER data on reserve allocation by currency as well as that of Asian
holdings of US treasuries speak the story. The COFER data for emerging and developing economies shows a decline in the share of reserves allocated to the USD from 75% in 2001 (average) to around 57% in Q1 2011.

On the other hand, the three currencies to whom allocations have increased are the GBP, EUR and ‘others’. We suspect that the ‘others’ category comprises of currencies like the CAD, AUD and NZD. Trends in allocation to the EUR are interesting – the share of the EUR increased from 18% to around 30% in mid-2009 but has tapered off since. Interestingly, the share of the CHF has remained low at around 0.2%, suggesting that the private sector and not central banks has been the dominant force behind the recent strength of the CHF.

Considering US treasuries as a share of Asian FX reserves, the trend has been more volatile. While holdings have remained between 30%-33% from 2006, there have been periods of sharp fluctuations. For example, the share jumped to more than 40% in the early stages of QE2. On the other hand, the share had dropped to a low of 23% in early 2008. The point we are making is that it is possible to change the allocation of reserves. YTD data also establishes this: holdings of treasuries have remained roughly the same even though reserves have increased by around USD400bn. This holds for most countries including China.

So where will NJA central banks go? Of course, the optimal solution would be to simply appreciate their currencies rather than accumulate reserves. To some extent, the on-going drop in USD/CNY fixing may be signalling a tilt in this direction. It is however, unlikely that NJA will give up on reserve accumulation. The alternatives investments range from increased holdings of gold, alternate currencies and emerging market fixed income products.

‘Other’ currencies and the CHF in which emerging markets and presumably NJA are underinvested are likely to become favourites. Gold is also a possibility and it is pertinent to note that the Bank of Korea has purchased approximately 25 tonnes over the last two months. This was the first purchase in more than a decade. There is also scope to increase investments in emerging market bonds and in Asia in particular for reasons of familiarity. Now it can be argued that the region has only two AAA rated economies, Hong Kong and Singapore, both of whom are very small.

China and Taiwan are AA rated and Korea is A. The point to bear in mind is that the decision to diversify is not contingent on US ratings alone. As stated earlier, the decision also depends on the policy framework in the US and here we believe that the Fed’s commitment to retaining its funds rate at near zero percent through mid-2013 is reason enough to be nervous on the USD.

In the local fixed income space in NJA, the size and liquidity of the markets will define the extent of
reserve diversification. Unfortunately, data on government bonds for purposes of our analysis is not available and we have to consider the aggregate (corporate and government) bond markets. At end-2010, these bonds amounted to USD6trn – this was 9.3% of global bonds outstanding and 23% of the level in the US. In 1996, NJA bonds were only 2.4% of the global outstanding. The share of government bonds is likely to be lower but still, the pace of growth has been solid enough to facilitate reserve diversification.

We are aware that Asian governments have guarded foreign investment in local currency bonds via capital controls etc. Neither are investment rules homogenous and differences in withholding taxes etc. can be large. That said, there has also been progress. For example, yield curves in the region have become more homogenous, linkages between bond and FX markets have become tighter and the regulatory environment has been relaxed to an extent. For example, central banks with bilateral swap lines with PBOC, China can invest in China’s bond market. Similarly, there is a proposal to develop a central securities depository in the ASEAN region.

None of these measures are by themselves adequate and much more will need to be done with regards to depth, traded volumes etc. Still it is simply wrong to assume that treasuries will remain the preferred choice regardless of the policy framework in the US.