Friday, October 24, 2008

On The Huge Spike In The Fed's Balance Sheet

Bob Murphy emailed me to ask me my thoughts on the huge spike in the Federal Reserve balance sheet. This is an important enough topic that I will outline my thoughts in this post.

There has been huge coverage in the econ/blog world with regard to this spike--a lot of the commentary being confused or just plain wrong. Anyone, for example, calling the spike a major inflationary injection is way off.

The best coverage I have seen has come from James Hamilton at Econbrowser.

Here's one of the charts that everyone is getting worked up about.

Here's Hamilton's take on what is going on:

...the real action began last month...the Fed expanded its total asset holdings by $600 billion over the last 30 days, with less than a third of this going directly into reserve balances...

Reserves ballooned [immediately after 9-11] to $67 billion, as excess reserves simply piled up in some banks while others remained in need. Last week's spike of $171 billion was 2-1/2 times as big-- the breakdown of interbank lending last week proved more profound than that caused by the physical disruptions in New York in 2001.

Anyone who suggests that last week's ballooning reserve deposits represent inflationary pressure or the Fed monetizing the deficit simply doesn't know what they're talking about. Banks are sitting on the reserves, not withdrawing them as cash. When markets settle down, the Fed can and will absorb those reserves back in with sterilizing sales of Treasury securities, just as it did in 2001 or after the more modest spike in August 2007....

...we see that creating new reserves, as dramatic as it was, was dwarfed in magnitude by some of the other actions the Fed took over the last month...

I gather that the Treasury auctioned off some extra T-bills to the public, in addition to their usual weekly auction, and simply kept the receipts as deposits in an account with the Fed. If that were the end of the story and the Fed kept its total liabilities constant, it would result in a huge (completely infeasible technically) drain on reserve balances and currency in circulation, as banks sought to deliver reserves to the Treasury's account to honor their customers' purchases of the T-bills. So the Fed offset the supplemental Treasury auction with a matching purchase of private assets, such as the PDCF and AMLF, thereby temporarily delivering reserves to banks which the banks in turn could hand over to the Treasury supplementary account. The net result of such dual Treasury/Fed operations is that the newly created "reserves" would just sit there in the Treasury supplementary account doing nothing other than standing as an accounting entry. In other words, the device allowed for a huge expansion of the Fed's balance sheet without causing any change in currency in circulation or reserve deposits.

So there you have it. The flight to quality is resulting in the Treasury being able to raise huge amounts in the T-Bill market. By treasury depositing the proceeds at the Fed, it is the same as the Fed drawing these huge reserves out of the economy. Thus the Fed re-injects the funds by its its purchases through currency swaps, primary dealer credit facility operations and the like. Resulting in a non-event, from a money supply perspective, in spite of the huge increase in assets on the Fed's balance sheet.

The theory then is that when the flight to quality reverses itself, the entire operation will reverse itself. The T-bill owners will demand payment for their T-bills, the Treasury will draw down on its reserves at the Fed to pay off the T-bills, and the Fed will drain money from the system to offset the new money coming into the system from payments to T-bill holders by the Treasury. But, here's the rub, to drain reserves the Fed will have to sell off the securities it has purchased. It may be able to sell off its commercial paper, but who is going to buy the junk mortgage backed securities it has purchased? Thus, this could all turn out very inflationary once the Treasury needs to pull its deposit with the Fed.

Bottom line: Changes in M2 nsa money supply remains the best indicator of how much net-new money the Fed is adding to the system.

8 comments:

  1. Question: why is the adjusted monetary base (http://research.stlouisfed.org/publications/usfd/page3.pdf) going through the roof? If it's really a non-event, how can new bank reserves be getting created?

    ReplyDelete
  2. notgreat, the answer is really in this post. You have to read it slowly.

    The Treasury drained reserves by dumping the proceeds of their Treasury sales with the Fed. The Fed added money by their purchases to balance things out.

    ReplyDelete
  3. Yes, I understand the drain part: people take money from their money market account or bank account and purchase Ts. Treasury deposits its newly raised cash with the Fed. The Fed/Treasury has somehow gotten that money back to the banks (through what mechanism is another interesting question). But they're not merely balancing things out -- bank reserves have gone through the roof. Where is the net add coming from?

    ReplyDelete
  4. Think of it this way. You have $1,000 cash that you deposit at the Fed. Fed deposits have SKYROCKETED! But this drains $1,000 from the system.

    And you need the cash to live on, so the Fed let's you borrow this money. Now you have your $1,000. The Fed shows your IOU on its books against the cash it lent and your deposit is still at the Fed.

    The amount of money in the system hasn't changed but their are new accounting entries everywhere.

    ReplyDelete
  5. But where did the $1000 cash that I have come from? Presumably I didn't have it under my mattress before, it was in the banking system to start with -- in a bank's vault or on deposit w/the fed, no?

    ReplyDelete
  6. Yes. The $1,000 is money that is already in the system, but when it is deposited at the Fed it leaves the system..

    ReplyDelete
  7. I think we're talking past each other. If the $1000 cash was in the system before, it would've been counted in the monetary base graph that I linked to. So there would've been no net change if things worked the way you describe, but clearly there was a huge net add, so those new reserves must have come from *outside* the banking system somehow. The question is how?

    ReplyDelete
  8. This gets tricky, if you have $1,000 in currency, it is available to bid up goods. Once you buy T-bills and the Treasury deposits it with the Fed, there is no change n the Fed balance sheet your currency s off the books the deposit by the Fed increase their reserves at the Fed. Thus net neutral BUT the Treasury deposit is not out there bidding up goods, thus it is a drain on currency in circulation. The Fed then adds reserve to increase the currency by $1,000, so that it is back where it was, but it will show a doubling of the reserves on the books.It's really hocus pocus stuff.

    The Treasury is selling T-bills because the markets won't buy anything else, they launder it through the Fed and buy what the markets won't.

    ReplyDelete