Fed Facing a 'Money Trap'

John Maynard Keynes suggested that when interest rates are very low, monetary policy does not work, and there is a so-called “liquidity trap”. Speculators anticipate that interest rates will go up and bond and asset prices will fall. Therefore, speculators transform their wealth holdings into the most liquid asset, money (cash), rendering monetary policy ineffective. It is often argued that a liquidity trap occurred during the Great Depression. Between August 1929 and March 1933 the Federal Reserve Bank (Fed) increased the monetary base by 41% and yet the money supply (M1) decreased by 29% as people started hoarding cash because of bank failures.

So, in a financial crisis such as today, where large banks fail seemingly overnight, and the Fed is responding by lowering the rates, one would think that a liquidity trap could occur. Instead something different is happening, call it a money trap.

As insightfully pointed out by Michael Shedlock, Treasury Bills (T-bills) yields are falling because investors (company treasurers, mutual funds, local government treasurers, etc) prefer them to money. Money amounts above the FDIC $100 000 limit are not safe, and are quite costly too keep in cash and to dispense with (who uses cash anyway). A similar effect was also observed during the Great Depression, by the way.

In a liquidity trap, speculators want to get rid of bonds and obtain cash. Thus, the demand for money shoots up. In the current crisis, individuals and firms with large amounts of short-term savings want to get rid of cash and obtain safe liquid assets. This happens because market after market has been failing and corporations, local governments, investment funds, and wealthy individuals need security on their short-term asset holdings. The only safe liquid assets available are short-term T-bills, as they protect against the risk of bank defaults. So demand for T-bills is very high, the price of T-bills is rising and their yield falling to levels similar to those observed during the Great Depression.

The Fed has been trying to address the problem, to no avail thus far, by means of outright sales of T-bills. The Fed’s aim is to drive the short-term T-bill yields closer to the federal funds target rate. Its holdings of T-bills have fallen by 37% ($90bn) since the end of 2007 (see the Fed’s H.4.1 Release). Yet, the yield on the 3-month T-bill has fallen to a low of 0.26% last week, nearly two percentage points lower than the discount rate, although it has since recovered to 1.2%.

Why is this a problem? Because it means that the Federal Reserve Bank is unable to conduct monetary policy, its main function in the current macroeconomic policy making framework, since it currently does not control short term rates other than the federal funds rate.

The Fed targets the short term federal funds rate, currently at 2.25%, through open-market operations. These operations are used to exchange money (cash) by securities and vice-versa between the Fed and the banks, and occur regularly through temporary repo operations but in relatively low volumes.

For example, one year ago, Fed repos plus other loans to commercial banks amounted to just $32bn. As of March 19th, these Fed assets had risen to $160bn including the TAF (Term Auction Facility) and PDCF (Primary Dealer Credit Facilty), still a small amount in the grand scheme of things. In normal circumstances, the yields on the 3- and 6-month T-bills, as well as the yields on the interbank rates (Dollar Libor rates), where trillions of dollars of money and securities are exchanged between private agents, follow the federal funds target rate fairly closely. Therefore, the federal funds rate “guides” the short-term market rates to “an equilibrium”. Currently, however, the yields on the 3-month T-bills are one percentage point lower and the yield on the 1- and 3-month Libor rates about 35 basis points higher than the federal funds target rate.

The Federal Reserve has taken unprecedented measures (lowering the discount window rate, the Term Auction Facility, the Primary Dealer Credit Facility, and starting March 27th the new Term Securities Lending Facility - TSLF) in order to lower the Libor rates to the current levels, which are closer to but still above the federal funds target rate, and to try to reduce the squeeze on T-bills and bonds that has occurred as their prices have risen in the flight to safety. These measures in effect subsidize commercial banks and primary dealers with tax-payers funded liquidity and Treasury securities.

However, despite the Fed’s intervention, private savings holders don’t want to hold cash and prefer the safety of T-bills with much lower yields. In fact, demand for T-bills is so large that at last week’s rate of intervention the Fed holdings of T-bills would be depleted in 5 weeks. Alas, given that T-bills are zero cupon bonds, it is quite likely that we will observe negative short-term yields on the T-bills in the near future, for the first time in history (negative rates on T-bills repos occurred already on March 20th). This means investors will be paying more to the U.S. Treasury than what they will get after 3- or possibly 6-months, just in order to ensure that their money is safe. It is like paying taxes they don’t actually owe to Uncle Sam, and in addition let Uncle Sam hold on to their cash, just because the investors don’t trust the private financial system.

What is happening is that the demand for money (cash) by savings holders is falling. If the Fed had maybe more than a trillion dollars of T-bills on its balance sheet it might be able to keep the yield on the T-bills close to the federal funds target rate, by means of outright sales of T-bills. But as of March 19th it just had $150 billions of T-bills left, so it is unable to push the T-bills yield higher.

Outright sales of T-bills by the Fed decrease the monetary base (money in circulation plus bank reserves), clearly something it does not want to happen in the middle of a financial crisis as it would be strongly contractionary. Therefore, the Fed has had to sterilize these sales of T-bills to maintain the monetary base. It accomplished this by making temporary purchases of other securities (e.g., mortgage based securities), namely through the new facilities (TAF, PDCF), while maintaining roughly constant its permanent holdings of Treasury notes (2- and 5-years) and bonds (10- and 30- years). The Fed has also maintained the monetary base quite stable, which is very surprising given the stresses in the system. At this point, it seems very likely that the Fed will have to significantly increase the monetary base, i.e., print more U.S. Dollars, like it did between 1929 and 1933, to ameliorate the effects of the upcoming steep contraction in money supply.

The Fed has been trying to solve a crisis for which it lacks the instruments, and in the process, is damaging its credibility. Holders of short-term savings and banks correctly perceive that the risk of default by banks is very high. Thus, there is a large spread between 3-month T-bills and the 3 month Libor rate (the TED spread), and rightly so. The market is functioning. There is no inherent problem in having a large TED spread, although it may result in further failures of weaker banks. Still, if policy makers decide intervention is required to reduce the TED spread, the market is also telling who should act: the US Treasury, either through blanket coverage of deposits, bank bailouts, or nationalization of insolvent banks.

Comments

[Just Say Whoa!:]

Interesting analysis on the limits of FED policy in fixing this problem.

The root of the problem was lack of regulation of the "shadow banking" system that issued "opaque" securities, which turned out to be a giant Ponzi scheme.

The prospect of negative T-Bill rates brings memory of a quote from the movie "Animal House" where Kevin Bacon's character screams at the top of his lungs:

"All is well! Remain calm!"

2008 Mar 26 12:08 PM

[vaduz:]

very disturbing thoughts

2008 Mar 26 01:04 PM

[flow5:]

The "monetary base" is not a base for the expansion of money.
The "demand for money's" meaning is exactly the opposite of your usage (see money paradox).
The growth in currency held by the non-bank public has grown by 4% since 1/2/06 and the growth in legal reserves have grown by 2% since 1/4/06. Pretty small numbers. However reserves are no longer "binding". Therefore you can't measure the "money multiplier" (it has doubled in the last 8 years) and you cannot discern whether the Fed is following a "restrictive" monetary policy or an "easy" one (with your data).
One more thing. It's impossible to conduct monetary policy using interest rate targets.

2008 Mar 26 01:18 PM

[timjowers:]

The best article I've read on the FED in a while. Ponzi scheme. $30B to a bank but not one dime to mortgage holders. Hmph!

Also, if the US government defaults then who will be first in line to take the government lands? Treasury holders? I think 12% of the land in the USA is claimed by the US government so each dollar does have a real hard asset behind it I think.

2008 Mar 27 04:49 PM

[Uncle Billy...:]

"For example, one year ago, Fed repos plus other loans to commercial banks amounted to just $32bn"

Ricardo, where did you get this number from? Do you know if there's a way to still put together m3 numbers? Is the info still reported by the fed, just not in a consolidated format?

2008 Mar 27 04:50 PM

[Ricardo Cabral:]

Flow5, thanks for the comments
I believe comparisons with 2007 are more informative than comparisons with 2006. Growth in currency held by the non-bank public has been just 1% between Feb 2007 and Feb 2008. M1 growth has been 0% and Demand and other checkable deposits have fallen by 1.3% although the retail sweeps program (RSP), which has transferred around $30-$35bn per year in the last few years, explains this fall. Growth in legal reserves was 0.3% in the period.

Nonetheless, you are correct in arguing that the changes in the reserve requirements make it difficult interpret whether the Fed is following a restrictive or easy monetary policy. If we define M1’=M1+RSP (this is a somewhat questionable comparison to do since people might have used their transaction deposits differently than did, had the RSP not existed), then M1’ has been growing at a rate below 2% since 2006, and with a monthly average slightly under 1% in the last twelve months, well under the inflation rate. The Fed has also reduced the rate of nominal monetary base growth to just 1.1% year-on-year as of Feb 2008 (from 2% in Feb 2007 and 4.3% in Feb 2006), meaning negative real monetary base growth. So, according to these measures, the Fed’s monetary policy has been restrictive. M2 growth, on the other hand, has accelerated to a 6.9% annual rate in February. Thus, one could also argue, as you seem to suggest, that such growth rate represents “easy” monetary policy and continuing trust in the banking system, contrary to what I argue in my article. But I may expand on this issue in a forthcoming post.

Different schools of thought (e.g., monetarists) may argue that it’s impossible to conduct monetary policy using interest rates targets. However, the use of interest rate targets to conduct monetary policy is explicitly stated by the Fed (see www.frbsf.org/publicat...) and the Bank of Canada (see www.bankofcanada.ca/en... ), for example.



Uncle Billy:

The numbers are available from H.4.1. Release (www.federalreserve.gov.../). I don’t think the M3 numbers can be so easily collected, otherwise somebody would already have posted them on the internet.

2008 Mar 30 05:49 PM

[Armagedon:]

The St. Louis Fed publishes the number for MZM, which is an acceptable proxy for M3. It is freely available on their website

2008 Oct 15 09:39 PM