Morgan Stanley has launched a new global reserves monitor. Their weekly has a bit more detail than in the online version. I can hardly complain though. I have a bit more detail hidden behind the RGE firewall than I put on the blog.
To be honest, I was quite happy when the big investment banks were -- generally speaking-- ignoring global reserve growth. Imitation can be considered a form of flattery, but competing against a bank with Morgan Stanley's resources isn't easy. Over time, I suspect that they will produce a product that I cannot match.
Stephen Jen and Charles St. Arnaud's work shows that global reserve growth remained quite strong in 2006. And given that Stephen Jen likes to emphasize "diversification across instruments" rather than "diversification across currencies," I suspect they also think dollar reserve growth remained strong.
I agree on both poinths. However, for what it is worth, some of the numbers I am looking at remain a bit different that those that Morgan Stanley is circulating.
Morgan Stanley estimates 2006 global reserve accumulation was around $600b, after adjusting for valuation. That seems a bit low to me. I am looking at a total of more like $740b. Morgan Stanley doesn't count the non-reserve foreign assets of the Saudi central bank and I do. But even without the increase in Saudi foreign assets, I am looking at global reserve growth of about $675b for the year.
As importantly, I believe global reserve growth (counting Saudi foreign assets) picked up slightly over the course of 2006, rising from a $700b annual pace in the first half of 2006 to a $800b annual pace in the second half of the year.
My total assumes about $200b in valuation-adjusted reserve growth in q4. China, the rest of emerging Asia and Japan added $140b or so to their reserves in q4 (after adjusting for valuation), Russia chipped in another $37b or so, Brazil and Argentina chipped in $16b and so on. Sum it up, and the countries Christian Menegatti and I track look set to add about $225 to their reserves in the fourth quarter. We don’t track a bunch of smaller countries that are adding to their reserves. The total headline increase in global reserves -- including SAMA foreign assets -- will probably be around $260b in q4. Roughly $50b of that likely reflects valuation gains – hence my overall estimate.
Morgan Stanley estimates that valuation gains – the rising dollar value of the world’s existing euros and pound reserves -- added $65b to world’s measured reserves, at least in dollar terms. That seems on the low side to me.
In the first three quarters, valuation gains added about $65b to the reserves of those countries that report the currency composition of their reserves to the IMF. Based on their reported holdings and currency moves in the q4, that should rise to $100b for the year. And I think valuation gains added $50b to the reserves of countries that do not report data to the IMF over the course of 2006.
$740 in reserve growth and $150b in valuation gains adds up to a headline increase of $890b in reserves (the headline increase, using my measure, was $630b in the first three quarters of 2006), or $820b without the increase in Saudi foreign assets.
What about the country specific detail?
Jen and St. Arnaud quite correctly note that the increase in China’s valuation-adjusted reserves has lagged the increase in Chinas current account surplus. The explanation – at least in the first half of the year – is a rather mysterious $45b in “private” purchases of foreign debt by Chinese residents. Similar outflows must have continued in the second half of the year.
Morgan Stanley’s valuation adjustment for China seemed a bit on the low side, given that Stephen Jen has previously indicated that he thinks only about 60% of China’s reserves are in dollars. If 35% of China’s end 2005 reserves were in euros and pounds and 5% in yen, the valuation gain on those reserves in 2006 should be about $35b – v. only $20b in Morgan Stanley’s table.
Morgan Stanley's valuation adjustment for Russia also is a bit different than what I would have expected. Russia actually reports valuation gains on its reserves – so we know that currency moves added about $6b to the value of Russia’s reserves during the first three quarters of the year. A 50% dollar/40% euro/10% pound currency split in q4 implies another $4b of gains – or $10b for the year. Russia’s valuation adjusted reserve increase for 2006 then works out to $110b (it would have been above $130b if Russia hadn’t paid the Paris Club).
Morgan Stanley seems to be getting a smaller number because they aren’t counting Russia’s “other reserve assets” (reverse repos, according to the central bank; see the footnote) as part of Russia’s foreign currency reserves. I am not 100% sure on this, but I suspect these reverse repos generate fx exposure. The data on total fx reserves that Russia reports with a lag in the same data set that provides a breakdown of currency gains seems to include the reverse repos.
I shouldn't just be critical though. I quite how like Morgan Stanley estimates of the interest income on existing reserves – Christian Menegatti and I need to do something similar. Interest income certainly explains most of Japan’s 2006 reserve growth.
But I don’t agree with Jen’s argument that – setting China aside – emerging Asia has gotten out of the business of adding to their reserves. At various points in 2006, some Asian central banks not named the PBoC actually were quite active – at least according to the data I watch. In early 2006, emerging asia attracted a big wave of capital flows, and its reserves went up. In April, emerging Asia resisted a bout of dollar pressure: many Asian countries were adding to its reserves rapidly before the emerging market sell-off in May/ June. Q3 was a light period of intervention – but there was a big bout of intervention at the end of q4. The following chart suggests that valuation-adjusted reserve growth in emerging asia (ex-China) is correlated with moves in the euro-dollar (the blue line). That makes sense – move sin the euro-dollar can be considered a proxy for private demand for dollar assets.
Some of this reserve growth in 2006 is simply interest income. But some isn’t.
If the dollar comes under further pressure in 2007, emerging Asia is likely to continue to add to its reserves. If the dollar continues its current rebounds, emerging Asia (ex China) could well be able to stay out of the market.
The new frontier here is tracking additions to oil investment funds. In 2006, I would bet that oil investment funds (Norway, Abu Dhabi, Kuwait, Qatar and Kazakhstan) not managed by central banks added about $120b to their assets, though that estimate is subject to a large margin of error.
$120b in oil funds and $740b from central banks would imply total official asset growth of $860b ... That is a big number. It presumably is a big reason why the US hasn’t had trouble financing a huge deficit despite being a less-than-stellar destination for investment.