The Atlanta Fed's macroblog provides commentary and analysis on economic topics including monetary policy, macroeconomic developments, inflation, labor economics, and financial issues.

Authors for macroblog are Dave Altig, John Robertson, and other Atlanta Fed economists and researchers.

« When is rapid growth in a central bank's balance sheet not cause for concern? | Main | More on interest on reserves »

May 14, 2009

Tools for managing the Fed's balance sheet

In his question and answer session following a speech to the Chicago Fed's Conference on Bank Structure and Competition on May 7, one of the topics Federal Reserve Chairman Ben Bernanke addressed was various tools for managing the Fed's balance sheet and the implementation of monetary policy in the future.

As I see it, the issue is that the Fed has created a very large amount of reserves and its conventional monetary policy tool, the federal funds rate target (the desired market rate for bank reserves), is very close to zero. If the funds rate target is going to be raised at some point in the future then either the Fed needs to be able to simultaneously keep demand for reserves at a high level or it has to reduce the amount of reserves available in order to drive the market rate higher. In other words, a positive interest rate would require eliminating a large portion of the excess reserves or elevating the demand for reserves in a way that is consistent with the market rate near the desired (target) level.

The options mentioned by the chairman last week fall along these two lines. One possibility is to have the demand for excess reserves at a high level, with the interest rate paid on reserves being the primary lever for the implementation of monetary policy. Last week I wrote about interest on reserves in the context of New Zealand's floor system and the fact that a large level of bank reserves replaced central bank daylight overdraft facilities for the purposes of meeting payment system needs. I also noted that U.S. banks have been putting their large holdings of reserves to similar use—significantly replacing daylight overdrafts. Nonetheless, it remains an open question as to whether market rates could be controlled satisfactorily by adjusting the interest rate paid on excess reserves alone given the size and complexity of the U.S. banking system.

What about reducing reserves? Beyond the decline in reserves that would automatically accompany reduced demand for the various Fed liquidity facilities, one tool for reducing reserves is outright sales of assets—Treasury securities, agency debt, and agency mortgage-backed securities—held in the Fed's portfolio. Doing so would reduce the level of bank reserves and also reduce the overall size of the Fed's balance sheet by the same amount.

Another option mentioned by the chairman is to use term reverse repos against these same Fed assets. A repo transaction resembles a collateralized loan, but it is technically a purchase and subsequent sale agreement with the price differential effectively reflecting the interest on the transaction. A Fed repo operation temporarily buys acceptable securities in exchange for reserves, while a reverse repo temporarily sells Fed-owned securities in exchange for reserves. To the Fed, a repo is an asset on its balance sheet that is matched by an increase in reserves on the liabilities side. In contrast, a reverse repo is a Fed liability that is matched by a decrease in reserves. Hence, a reverse repo does not reduce the size of the balance sheet, but it does reduce the amount of reserves. (Central bank trivia: Unlike most other central banks, the Fed convention is to talk about repos and reverse repos from the perspective of the counterparty.)

One option not mentioned by the chairman last week but talked about by others (see, for example, Janet Yellen's speech on May 6) is for the Fed to seek authority to issue its own unsecured debt. These Fed bills would be similar in effect to a reverse repo—replacing reserves with interest-bearing Fed obligations, but these obligations would not be backed by Fed-owned collateral. Many other central banks have the authority to issue their own unsecured short-term debt, and a number of central banks utilized this authority to manage their balance sheet during the heat of the financial crisis in 2008. For example, sterling bills and reserve bank bills have been used in the United Kingdom and New Zealand, respectively, as the primary means of draining reserves when reserves were created as a result of short-term liquidity facilities.

However, it is important to note that the central banks that introduced bills were generally facing large increases in reserves and declining amounts of government debt on their balance sheets. This situation significantly limited their ability to conduct large reverse repo operations or outright sales of securities to drain excess liquidity. For example, the Bank of England's holdings of government debt at the beginning of 2009 were less than 2 percent of its balance sheet, down from about 11 percent in early 2008.

In principle, then, it seems the Fed has plenty of tools available for managing reserves and the federal funds rate in the future. As my former countrymen in New Zealand are apt to say—no worries, mate! Of course, that leaves plenty of other things to worry about.

By John Robertson, vice president in Research at the Atlanta Fed

May 14, 2009 in Monetary Policy | Permalink


TrackBack URL for this entry:

Listed below are links to blogs that reference Tools for managing the Fed's balance sheet :


I have a question that reflects my ignorance of how these things work more than being really substantial.

Wouldn't the reserves we are talking about fall as well as a result of banks actually lending money out to their customers?

If so, does that mean that the problem is how to balance the increased lending (and monetary expansion which that implies) against the fear of future inflation?

And again, if I've understood correctly, isn't there a sort of "buffer zone" created by the risk of deflation,the lack of economic activity, etc. which gives the Fed time to manage the balance sheet contraction?

When the economy eventually recovers, reserves will start to fall, the velocity of money will start to rise, the monetary base will grow and in this context the Fed needs to start shrinking its balance sheet. Is that right?

Posted by: Alexander Singer | May 15, 2009 at 07:02 AM

Re: "If the funds rate target is going to be raised at some point in the future then either the Fed needs to be able to simultaneously keep demand for reserves at a high level or it has to reduce the amount of reserves available in order to drive the market rate higher."

With interest on reserves (IOR), why would a bank not want to keep its reserves at the existing level after the policy rate is increased? They are operating on a horizontal segment of their demand for reserves. I just don't see why raising the policy rate will be a problem with IOR. What am I missing?

Posted by: tom | May 15, 2009 at 02:12 PM


If you are still out there:

"Wouldn't the reserves we are talking about fall as well as a result of banks actually lending money out to their customers? "

Yes, but this is exactly the problem. Increased lending to soak up all of those reserves would result in massive inflation. Remember banks can loan out ten dollars for each dollar in reserves.

Posted by: Karl Smith | May 20, 2009 at 11:48 PM

Hello everybody,

my question goes into the direction of the possibility to drain reserves by reverse repos. In my opinion this is only a temproary solution, at the end of the day increasing reserve balances cause the fed buys the assets back at slightly higher prices. Or is that wrong? what maturity structure would these repos typically have?

Posted by: Philipp | September 22, 2009 at 07:42 AM

Post a comment

Comments are moderated and will not appear until the moderator has approved them.

If you have a TypeKey or TypePad account, please Sign in