Guess who financed the February Trade Deficit?

A hint.

Actually, we know that China won’t we don’t (yet) know China’s January reserve accumulation, let alone its February reserve accumulation. But some Asian central banks have released data for February, and, no shock, they seem to have resumed accumulating reserves. To an extent, higher reserve balances reflect valuation gains from shifts in the euro/ dollar and the February coupon payment on their Treasury holdings. But it also seems to reflect something more.

Look at what happened to the securities the Fed’s holds on behalf of foreign central banks.

The Fed’s custodial holdings increased by $11.6 billion from late December to late January. And then between January 27 and March 2, it increased by $31.3 billion, with a particularly big pick up in interest in Agency securities, which rose $21.6 billion. Incidentally, the story does not change much if you use February 3rd instead of January 27 as the start date.

If the February trade deficit was around $60 billion, and the increase in Federal Reserve custodial accounts reflects a real increase in central banks’ dollar reserves, not just a shift from existing account into the central banks’ account with the Fed, the growth in the Fed’s custodial holdings provided about 1/2 the total financing the US needed.

The Fed’s custodial holdings reflect just that -- the securities held by the Fed. It does not always track the TIC data perfectly. It clearly is not a perfect indicator of broader central bank flows, as it -- like all the US data -- understates central bank financing for the US.

In 2003, the Fed’s custodial holdings increased by $216 billion, and we know from the BIS that global dollar reserves (leaving out the $45 billion transferred to two state banks) increased by $442 billion. That leaves over $200 billion for central banks to have invested in other securities, or to placed in dollar denominated bank accounts.

In 2004, the Fed’s custodial holdings increased by $269 billion (The TIC data, by comparison, showed official net purchases of long-term securities of $236 billion). Nouriel and I estimate that dollar reserves increased by between $465-470 billion. If we are right, that means that there are an additional $200 billion in central bank assets floating around somewhere. Net of valuation gains, reserves grew by more than $600 billion, so, even if dollar reserves increased by something like $470 billion, nearly $150 billion flowed into non-dollars reserve assets.

My point: the Fed data just shows the most visible part of the iceberg, so to speak. The world’s dollar reserves have probably increased by more than $43 billion in January and February.

A $21-22 billion monthly increase in the Fed’s custodial holdings translates into a healthy $260b in annual increase in the Fed’s custodial holdings, about the same about as the US got in 2003 and 2004 EVEN THOUGH the Bank of Japan/ MOF are out of the market. It is a bit too early, though, to know if the growth in custodial accounts reflects central banks putting the dollars they bought in q4 to work (with a lag), or if it represents continued strong reserve accumulation in q1.

Remember, emerging economies added something like $200 billion to their reserves in q4 2004. Only about $45 billion of that showed up in increased custodial holdings at the Fed, and only $53 billion in the TIC data. I doubt that the remaining $150 billion was lent out locally, though apparently that is something Korea is now considering. And, we won’t have a good sense of the current pace of reserve accumulation until China releases its January and February reserve data.

Guo Shunqing -- the head of China’s State Administration of Foreign Exchange -- presumably has access to most of the answers, but he did not give them out over the weekend. He did say that China has no intention of selling dollars, but, as readers of this blog know, that is a second-tier issue. What really counts is how many dollars China is willing to add to its portfolio to defend its exchange rate peg. For the moment, though, it seems safe to assume that the Bretton Woods 2 system of reserve financing for the US remains alive and well.