More on Transatlantic Cash Flows

Last week I noted the recent buildup of cash assets by foreign banks in the US. I suggested that, when paired with ECB data showing that European monetary financial institutions (MFIs) have been drawing down their deposits in Europe, this provides circumstantial evidence that European institutions are reducing their exposure to Europe in part by moving cash to the US. (Note that most of that cash is being deposited with the Fed in the form of reserves.) But as noted by some commenters, it's worth considering a couple of other possible explanations.

For example, Kate Mackenzie at Alphaville draws our attention to commentary by Lars Pedersen at Alliance Bernstein. In June Pedersen noted the substantial buildup of cash assets by foreign banks in the US, and suggested that "building up dollar balances at the Federal Reserve might be viewed as insurance by these foreign banks." But he also noted that there may be something else going on (pdf):

There may be other motivations beyond safety. Deposits at the Fed are unusually attractive for foreign banks because of the regulatory landscape. Borrowing money via deposits that don’t exact an FDIC surcharge, and depositing them at the Federal Reserve, earning 25 basis points, is much more attractive to a foreign bank with a US branch than it is to a US bank. That regulatory gap also provides insight into why foreign-related banks have maintained sizable deposits at the Fed.

Could recent rule changes at the FDIC explain the recent buildup in cash reserves by non-US banks? Unless I'm missing something, this seems an insufficient explanation. Because if I understand the FDIC assessment scheme correctly, the FDIC charges don't seem to depend on whether bank assets take the form of deposits with the Fed or some other sort of asset. So it's not clear how this would specifically affect Fed reserves and not other asset classes.

But what about a second possibility raised by Pedersen and others: perhaps the rise in Fed reserves of foreign-owned banks is simply an artifact of QE2. Though this was not the program's primary goal, QE2 had the unavoidable side-effect of increasing the total quantity of reserves owned by the US banking system. And after all, someone has to hold all of the reserves created by QE2.

But while this may be a part of what's going on, this explanation again seems insufficient by itself. First, there's no obvious reason why the reserves created by QE2 should have gone so overwhelmingly to foreign-owned banks. QE2 created new bank reserves, but the Fed has no say in who actually holds those reserves. There must be some force at work that has made foreign-owned banks in the US more eager to increase their holdings of Fed reserves than domestic US banks. A greater fear of unexpected losses by non-US banks would do it.

Second, if the funds that foreign-owned banks in the US are stashing with the Fed were simply coming from QE2, then we would expect the rise in reserves held with the Fed to be matched by a decrease in those banks' other domestic US assets; the story would be one of foreign-owned banks in the US effectively trading one type of US asset (such as US Treasuries) for another (deposits with the Fed).

But foreign-owned banks in the US have been getting the funds to deposit with the Fed from their foreign parents, not from their base of existing US assets. The following chart shows FRB data detailing the amount owed by the US branches of banks in the US to their overseas affiliates, broken down by whether the banks are domestic or foreign-owned. (A negative number means that the foreign office owes the US office.)


One year ago the US branches of foreign-owned banks were effectively "owed" about $400 billion by their non-US parents. Then, beginning around the end of last year, foreign banks began a large net transfer of funds to their US branches. At first this transfer of cash had the effect of essentially "repaying" the net obligations that existed in 2010, and then from June 2011 onward those continued cash transfers from overseas offices to affiliated US branches meant that the US branches of foreign banks became net borrowers from their parents. The total cash transfer to US branches of foreign-owned banks over the past 8 months, according to these figures, was around $600 billion. And what happened to the cash received by US branches of foreign-owned banks in the US? They deposited almost all of it with the Fed.

Meanwhile, it's interesting to note that over the past 2 months, domestic banks in the US have been paying down the debts of their overseas affiliates to their US parents. While this has been a much smaller and more recent phenomenon than the cash movements of foreign-owned banks, I would be eager to hear some theories as to what's behind this. (I have a couple of thoughts about it, but no story that I find convincing yet.)

Note the one other time in the recent past when non-US banks shifted large amounts of cash to the US: during October and November of 2008, during the height of the (first?) worldwide financial crisis. Then as now, foreign banks were eager to accumulate the safest form of cash there is: dollar deposits with the Fed. This may be yet another sign that we are currently in the midst of a time of financial stress similar to that at the end of 2008. And that probably explains this queasy feeling I have...


UPDATE:For another alternative interpretation of this data, see this post by Rebecca Wilder.