VoxEU Column Financial Markets

Federal Reserve Policy Actions in August 2007: Answers to More Questions

The problem: About $1 trillion of commercial paper will mature in coming months. If issuers can’t roll it over, firms will turn to lines of credit that they have arranged as insurance against such events. Banks will be forced to make these loans, and credit conditions elsewhere will tighten. Such a credit crunch will inevitably slow the economy. This column explains the what's, why's and how's of the unfolding crisis and provides a progress report on the Fed's actions.

 There were two announcements from the Federal Reserve on Friday 17 August:


  • The Federal Open Market Committee (FOMC) issued a statement.
  • The Federal Reserve Board announced a reduction in the primary credit discount rate from 6.25% to 5.75%. 

During the last two weeks the Federal Reserve’s Open Market Desk operated at levels that are unexceptional with operations that ranged from nothing on Tuesday 14 August and Friday 24 August to a total of $17 billion on Thursday 16 August and $17.25 billion on Thursday 23 August. 

Finally, on Friday 24 August, the Federal Reserve Bank of New York issued a statement clarifying the rules for the collateral it takes in discount lending operations.

As I write this, things do seem to be calming down, albeit slowly.

What is the Federal Reserve so concerned about?

Initially, the Federal Reserve was concerned about ensuring that financial markets remained liquid. With the dramatic reduction in risk tolerance that began around Thursday 9 August, it became difficult to buy and sell certain types of bonds. By injecting reserves into the banking system, the Fed was providing the means for people to trade.

As the crisis unfolded over the past two weeks, focus has shifted to problems that financial institutions are having in obtaining short-term funding. Banks and other intermediaries obtain substantial quantities of financing at maturities of roughly 90 days. A large amount of this is in the commercial paper market – really short-term bonds. The $2.5 trillion or so worth of commercial paper outstanding in the United States comes in a number of varieties. Most of it has some sort of collateral or insurance behind it in case the issuer defaults. One type is backed by mortgage-backed securities – the stuff that has been at the core of the sub-prime crisis. And therein lies the problem – there is a shortage of collateral that investors are willing to take in exchange for a variety of short- and medium-term loans.

Big problems occur when issuers whose commercial paper is maturing are unable to sell new issues to refinance what’s coming due. Unable to “rollover” the matured paper, people will turn to commercial banks for funds. Financial and nonfinancial firms usually have lines of credit with banks – these are like credit cards that they can call on whenever they want. As a matter of course, banks provide lines of credit to a broad variety of customers. But they do it assuming that everyone will not want to borrow at once.

Here’s the problem: Over the next few months, something like $1 trillion of commercial paper will mature. If issuers can’t roll it over – if investors aren’t willing to buy the commercial paper – firms will turn to banks to use the lines of credit that they have arranged as insurance against such an event. Banks, in turn, will be forced to make these loans, and credit conditions elsewhere will tighten up. Such a credit crunch will inevitably result in a general economic slowdown.


So, the Federal Reserve’s actions have been aimed at encouraging investors to accept high-quality collateral in making loans so that borrowers don’t have to resort to bank loans. Let’s turn to a description of the specific actions of the last 2 weeks.

What is the “primary discount rate”?

The 12 Federal Reserve Banks have a standing offer to make loans to the banks in their Districts that they deem to be sound (as measured by the standardised ratings produced by supervisors). Since a January 2003 change in procedures, the primary discount rate has been set at 1 percentage point above the target federal funds rates. As long as a bank qualifies and is willing to pay the penalty interest rate, it can get the loan. The rules allow a borrowing bank to lend the funds again if it wishes. 

Primary credit is designed to provide additional reserves at times when the open market staff’s forecasts are off, and so the day’s reserve supply falls short of the banking system’s demand. (There is also something called “secondary credit” which provides loans to institutions that do not qualify for primary credit at a rate that is ½ of one percentage point above the primary discount rate.)


Historically, banks have been extremely reluctant to borrow. Over the past year, borrowings have averaged less than $200 million per day (that’s with an “m”). Recall from my earlier piece that on a normal day, bank’s hold a total of around $12 billion in their reserve accounts at the Fed. So, $200 million is a very small number.2


Why are banks so hesitant to borrow from the Fed?

Once-upon-a-time (over a decade ago) discount borrowing had a stigma attached to it. If a bank borrowed too frequently, the logic went, the people running the bank must not be very good at managing their balance sheet.  Frequent borrowers would be admonished by the Federal Reserve.

When the discount rate was one-half of one percentage point below the target federal funds rate, discouraging borrowing may have made some amount of sense. But that’s no longer the case. So why don’t banks do it? Specifically, why didn’t banks do it last week? And, are they likely to do it this week?

There are two reasons that banks have not utilised discount borrowing with any frequency. First and foremost, it is clearly more expensive than borrowing from other banks. Over the past two weeks, the federal funds rate in the market has been regularly below not just the primary discount rate (which was 6.25% until 10 August) but below the target of 5.25%.   The following table reports the effective (quantity-weighted) federal funds rate along with the reported high and low levels for the past 15 business days.

Federal Funds Rate
Overnight Trading

Effective Rate


Standard Deviation


Primary Discount Rate


*Indicates last day of a maintenance period.

The effective federal funds rate is the quantity-weighted average of transactions reported by federal funds dealers to the Federal Reserve Bank of New York. The standard deviation is also quantity weighted.

Source: Federal Reserve Bank of New York

But there is another very good reason that banks are unlikely to borrow. Here’s the problem. When I attended meetings of the Board of Directors of the Federal Reserve Bank of New York I would receive a copy of the notebook the Directors had. Inside this book was a listing of all of the discount loans the Bank had made since the last meeting – the size of the loan and to whom it went. Since the Reserve Banks are private nonprofit corporations and chartered banks, this made sense. The Directors had a fiduciary responsibility for the management of the bank.

That’s all well and good, but by statute three, each of the Reserve Bank’s directors is a banker. That meant that they were able to see who was borrowing. Since the directors rotate every few years, and bankers surely talk to each other, I strongly suspect that everyone knows some of their competitors will find out they have borrowed from the Fed. Would you borrow under those circumstances?

Nevertheless, several banks borrowed. Why did they do it?

On Tuesday 22 August there were reports of four large banks obtaining primary credit. Citibank, JPMorgan Chase, Wachovia and the Bank of America were reported to each borrow $500 million for total borrowings of $2 billion. And on Thursday 23 August, in their weekly data release, the Federal Reserve reported Wednesday end-of-day primary credit of $2.001 billion and average weekly borrowings of $1.193 billion. This suggests that the four large banks had borrowed for an average of between 3 and 4 days.

There are several reasons the banks did this. First, they clearly did this very publicly in order to show their cooperation. Second, it was inexpensive. To understand the cost of the borrowing, we need to take a short detour into a few details of how reserve accounting works. Banks are legally required to hold reserves to back certain types of checking account deposits. The requirement is computed as an average over a two-week period ending every other Wednesday. As it turns out, the most recent reserve maintenance period began on Thursday 16 August and ends on Wednesday 29 August. Because the requirement is met on average of the 14-day period, a bank that holds a high level of reserves at the beginning of the period can then hold a low level at the end.3

This brings us back to the four large banks that borrowed. My guess is that they now have held a high level of reserves for a few days early in the maintenance period, so they can economise over the next week.

Turning to the cost, in the United States, bank reserves receive zero interest. In the absence of holding reserves in their own account, a bank can lend them out to another bank. The rate of interbank lending is the federal funds rate, so that’s what economists call the “opportunity cost” of holding reserves. Assuming that a lending bank obtains roughly the effective rate on its lending, the opportunity cost is something close to 5%. Banks paid 5¾ for the privilege of obtaining a total of $2 billion of primary credit. So the daily cost of this is the difference between these two – that turns out to be a total of about $29,250 per day, or $7300 per bank per day. To put that into perspective, a bank CEO with annual compensation of $10 million per year is paid nearly 4 times that per day, 365 days per year.

So, the four large banks that borrowed were able to purchase some goodwill at what looks like a pretty low price.

What was the purpose of reducing the primary discount rate?

The reduction in the primary discount rate was prompted by the impression that financial firms were having trouble obtaining term financing – that is, loans of 30 to 90 days. The signal for this was that the 90-day uncollateralised lending rate for banks, what is called the “London Interbank Offer Rate” or LIBOR, was nearly 2 full percentage points above the 3-month Treasury bill rate. That, together with reports that it had become extremely difficult to obtain funding at anything but the shortest maturities, prompted the action.

In order to address this problem, the Federal Reserve not only lowered the discount rate, but also increased the term for which it is willing to lend. Normally, discount lending is overnight. On Friday 10 August, the Fed announced that it would lend for up to 30 days.

While discount loans require banks to post collateral, what qualifies is quite broad – much broader than is accepted in open market operations. The Federal Reserve Bank of New York’s statement on 24 August was a public statement of this fact. As far as I can tell, the primary point of the policy change on 10 August and the New York Fed’s public information statement two weeks later is to encourage people to accept broad collateral.

What impact has the Fed’s actions had on markets?

The hope of the Fed is that it can bring confidence back into financial markets. One way to understand what has happened is that financial market participants suddenly shifted from trusting each other to not trusting each other. Without trust, people look for the safest possible places to put their funds – U.S. Treasury securities, for example. 

The lack of trust arose from two complementary sources. First, the sub-prime crisis caused financial professionals to question the ratings provided by people at Standard and Poors’ and Moody’s. These ratings normally provide the basis for pricing the risk in various types of bonds. The sub-prime crisis made it clear that the rating agencies were doing a poor job of evaluating risks in securities that were backed by sub-prime mortgages. Once investors started to question one set of ratings, it wasn’t a big step to question all of them.

This first problem generated a classic flight-to-quality that drained resources from certain borrowers. As credit dried up, investors and lenders started to question the medium-term macroeconomic outlook.  This increased the perceived riskiness of lending to virtually everyone (outside of the U.S. Treasury) and so financial institutions started to have a difficult time financing themselves.

The objective of the Fed’s actions was to bring confidence back into the financial markets, getting people to trust each other again. But there is a clear limit to what central bankers can do. First and foremost, they provide short-term liquidity to ensure there are funds for trading. Beyond that, what they can do is hope that people trust them to stabilise growth and inflation in the medium term. That is, Chairman Bernanke and his colleagues can try to convince financial market participants that they will continue to succeed in meeting their stabilisation objectives so that this source of risk is eliminated. What they cannot do is increase the quality of information that helps investors and lenders distinguish more from less risky borrowers.4

Has it worked? There are several ways to measure the impact of the Fed’s actions on the financial markets. First, we see that the intra-day range of the federal funds rate is much lower this week than it was last week. Second, after hitting a low near 2.5%, the three-month Treasury bill rate has risen back to 4%. As I write this, it appears that high-quality borrowers are able to borrow again.

Why didn’t the Fed put the funds into banks using open market operations?

As described in my previous VOX essay, the Open Market Desk of the Federal Reserve Bank of New York regularly injects reserves into the banking system through repurchase agreements. These open market operations are done through a group of 21 primary dealers and require very specific, high-quality collateral.

While it would have been possible to engage in repurchase agreements that are up to 65 days in length (that’s the current maximum term in the authorisation), the counterparties and collateral are severely limited.

Lending is very different. First, any of the more than 7000 banks in the country can borrow. And second, the collateral requirements are much less stringent. So, the hope is that by relaxing discount lending, the Fed will encourage banks to borrow and use the funds to lend to their customers.

It is important to note that in principle discount lending has virtually no restrictions whatsoever. The Fed can (and has) taken virtually anything as collateral, and they can lend to anyone. The relevant portion of the legislation reads as follows:

“Emergency credit for others. In unusual and exigent circumstances and after consultation with the Board of Governors, a Federal Reserve Bank may extend credit to an individual, partnership, or corporation that is not a depository institution if, in the judgment of the Federal Reserve Bank, credit is not available from other sources and failure to obtain such credit would adversely affect the economy. If the collateral used to secure emergency credit consists of assets other than obligations of, or fully guaranteed as to principal and interest by, the United States or an agency thereof, credit must be in the form of a discount and five or more members of the Board of Governors must affirmatively vote to authorize the discount prior to the extension of credit. Emergency credit will be extended at a rate above the highest rate in effect for advances to depository institutions.”

Reading this you can see that it means that the Fed could, if they felt the need, make me a loan and take my car as collateral. The only real restriction currently in force is that since there are only five sitting Federal Reserve Board Governors, the decision to do it would have to be unanimous. (I have a pretty nice car, so they might go for it.)

More seriously, this is the portion of the legislation that would have allowed the Fed to make loans directly to specialist firms on the New York Stock Exchange in the aftermath of the October 1987 collapse.

What is the FOMC statement?

A little background will be helpful in understanding what this is all about. The FOMC is the 12-member committee that makes monetary policy decisions in the United States – that means they set the federal funds rate target. Members include the seven members of the Board of Governors, the President of the Federal Reserve Bank of New York, and the presidents of four Federal Reserve Banks. While only five Federal Reserve Bank Presidents vote, all 12 participate in the deliberations.

The committee has eight scheduled meetings per year, and since May 1999 has issued a statement following each of them. The purpose of the statement, which is usually only a few hundred words long, is to both explain the current decision (raise, lower, or maintain the federal funds rate target) and to provide some guidance about future policy actions. 

Why did the FOMC issue the statement?

The issuance of a statement on 17 August was unprecedented. Never before has the committee issued a statement between regularly scheduled meetings that did not announce a policy change. Its purpose is clearly to signal that the increased likelihood of a cut in the federal funds rate target at the next regularly scheduled meeting on 18 September.

Up until now, the committee’s communication strategy had been working reasonably well. Chairman Bernanke and his colleagues want to ensure that they do not shock markets. That means telegraphing their intentions. Between the official FOMC statements and the public comments by Federal Reserve officials this has all gone very well. And on 7 August the FOMC met and issued a statement that said they were concerned about prospect of inflation rising; implying that the next move in interest rates might be up.

Last week things changed extremely quickly. Conditions in credit markets deteriorated to the point where some fixed-income markets (those are markets for bonds and loans) were failing to function adequately. If this continued, then the FOMC would have to cut its official target. The purpose of the unscheduled statement was to signal this concern.

The FOMC’s procedure of what might be called “prospective transparency” has backed it into something of a corner. 

What comes next for policymakers?

The next step would be to cut the federal funds rate target. Many informed observers clearly believe that they will do this. And investors who bet on this in a futures market think that the Federal Open Market Committee will cut the target at or before the next scheduled meeting on 18 September. 

My own reading of this is that policymakers are hoping that they will not have to do this. Chairman Bernanke and his colleagues have been very careful to discuss the federal funds rate target in the context of their medium-term stabilisation objectives. That is, whenever they mention possible changes in the rate, they do it with reference to the need to keep inflation low and stable, while encouraging high and stable real growth. The implication of this is that if the recent sub-prime crisis jeopardises these objectives, then the FOMC will act. If, on the other hand, the crisis is short-lived and contained, then there is little need to act by changing the target rate.


For details, see my Vox essay “Federal Reserve Policy Actions in August 2007: Frequently Asked Questions,” 14 August 2007.

The aftermath of the 9/11 terrorist attacks is a glaring exception to this. Then bank borrowing peaked at roughly $40 billion (with a “b”).

For details of how this all works you can see page 437 of the second edition of my book Money, Banking, and Financial Markets, New York, N.Y.: McGraw-Hill Irwin, 2009.

It is interesting to note that the 60-day commercial paper rate for high-quality (AA) nonfinancial borrowers remained close to 5.25 percent, where it has been for some time.


105 Reads