Regular readers will know that I greatly dislike Zero Hedge. Today, via The Browser, I came across an argument on the site that QE2 had been a secret support programme for European banks. This doesn’t sound right to me, and I have had a go at a response. This is mind-bending stuff so any criticism will be appreciated. Here is my comment:

The argument here is not easy to follow, but I think I can summarise it as follows:

1. The effect of QE2 has simply been to hand reserve balances to foreign banks, with no increase in the reserves of US banks. That is why US banks have not been lending.

2. The mechanism for this transfer has been the purchase of Treasuries from the foreign banks who are registered as primary dealers.

The argument is based on incomplete data — hence the use of cash holdings as a proxy for reserve balances in the piece. That is problematic because cash is created by various means other than an increase in the liabilities of the Federal Reserve. For clarity and simplicity I will consider the argument using the Q1 flow of funds data, from which the Zero Hedge argument can be made equally well.

It is true that the increase in foreign banks’ reserves mirrors the Fed’s asset purchases in the available data. The flow of funds tables ( show that foreign banks added $325.8bn to their reserve balances at the Fed between Q310 and Q111, which pretty much matches the expansion of the Fed’s liabilities in that time ($333.1bn). This is the essence of the Zero Hedge argument. However, US banks added to their reserve balances to the tune of $119.7bn in the same period, which tells us that the argument that US banks have not been lending because they were unable to lend does not stand up: both foreign and US banks added to their reserves during the period. This disposes of point 1.

I have also looked into the implication in the Zero Hedge piece that QE2 has simply been an exercise in buying Treasuries from foreign banks. In the course of the six months, the Treasury holdings of foreign banks have hardly moved at all (they have actually increased from $73.4bn to $74.6bn). Foreign banks have reduced their bank credit and holdings of corporate and foreign bonds and moved the proceeds to the Fed; they have not sold out of Treasuries. This disposes of point 2.

Now, you might ask how the banking system was able to increase its reserves at the Fed by more than the increase in the Fed’s liabilities. The answer is that there are other moving parts in this system. I have had a look at the composition of the Fed’s liabilities in the flow of funds. What happened between Q310 and Q111 was that deposits of were withdrawn from the Fed, mainly by the Federal government whose deposits fell by $191.7bn. If you were feeling mischevious, you could argue that foreign banks have helpfully allowed the Fed to have a stimulative balance-sheet expansion at a time when the American government was withdrawing balances from the system.

I suspect, however, that the reality is that one cannot consider parts of the system in isolation as Zero Hedge has done. I need some more time to think through how to expand on that; but the main point here is that the Zero Hedge argument does not stand up.