As of December 4, 2013, US banks had $2.493 trillion on deposit at the Fed. (Source: FRB H.4.1 Report). This amount includes required and excess reserves. The amount has increased by 63% over the past 12 months and approximately 300% since the Fed started paying interest on the balances. Bernanke started paying IOER during the financial crisis, but banks had wanted this for years. Some fraction of this reserve growth is due to QE and some fraction is due to the above-market rates that the Fed is paying. (This is the so-called IOER "Interest on excess reserves.") Right now, the Fed is paying about 0.25% on IOER and the t-bill rate is 0.02%. So the Fed is paying more than 0.23% above the market. On a balance of $2.5 trillion, this is a direct subsidy to FRB member banks of roughly $5.75 Billion per year and with each QE day, the amount grows.

This subsidy is theoretically being financed by the Fed's holdings of longer-dated securities so it's positive carry for the Fed. However, from the perspective of a risk-averse banker, and ignoring capital haircuts and the risks/spreads etc., a banker would need to buy treasury securities with a maturity of greater than 2 years to get the same yield as parking overnight money at the fed. So banks are behaving quite rationally.

The elephant in the room is the rate that the Fed pays on IOER. Talk is brewing that along with the announcement of a taper, the Fed will reduce the IOER rate. I submit that this is a highly unstable equilibrium and a change in IOER will have unintended (and unpredictable) consequences. Let's imagine that the Fed cuts IOER to zero. You will suddenly have $2.5 trillion looking for a new home. Where will it go? T-bills are already at 0 yield. So if banks just buy T-bills (even outside the fed) then that is a classic liquidity trap. Or, it's possible (but improbable ) that it will suddenly go into the real loan market. If that happens, the economy would go gangbusters with possibly little upward pressure on rates since $2.5 trillion in supply is a lot of money. Or, this gusher of ?dumb? money will listen carefully to the fed's forward guidance and collapse all rates towards zero out to the 2-year etc. I think this helps explains why Bill Gross is bullish on the front end of the curve because the curve is highly arbitraged between 2 years and 5 years. So it's possible that a taper announcement combined with a drop in IOER could turn out to be very bullish for the bond market. And this would persist until the Fed actually raises the funds rate.

Additionally, dropping the IOER might appease some critics about the size of the fed's balance sheet (ignoring the sheer quantity of bonds that remain). The IOER has been a subsidy to re-capitalize the banks. And now that this process is largely complete, the subsidy of $5.75 Billion/year should end and watching the gusher of $2.5 trillion leave the reserve account will be interesting, to say the least.

Bottom line: The IOER is a bigger deal than the taper announcement. The pundits will figure this out in due course.

Alston Mabry writes: 

"Remember that money we gave you, so you could give it back to us, and then we'd pay you for keeping it with us?"


"You can't have it back."

Bud Conrad writes: 

Rocky, Thanks.

The Fed has to buy up the new debt issuance from the government to keep rates low. It is also buying the MBS to keep mortgage rates low and to allow the banks to keep on their books holdings that might otherwise be declared toxic waste from being written off. So they can't stop QE purchases.

They have to fund the purchases some how. At present the Fed has been paying over market rate to keep the deposits of Excess Reserves to obtain the money to buy the Treasuries and MBS/Agencies. I don't see how the Fed balances its books if the banks withdraw $2.5 trillion. Then the Fed would look like a commercial bank that has a run from depositors and is quickly iliquid. The equity account is only $65 billion. The Fed is like a very leveraged hedge fund. If the depositors want to withdraw their money, the Fed would have to sell off assets or EXIT, which would cause panic in the markets.That seems even less likely. So Al is right: "You can't have your money" has to be the response.

So the Fed is trapped into continuing the payments on the deposits (IOER) as long as they have income from the Treasuries and MBS to pay for it. The idea that the Fed prints up currency is a little misleading because the actual physical demand for paper is decided by the public's conventions, and there is less use for the dollar bills with more transactions being done with credit cards. So as rates rise they will be raising the IOER rate, and at some point that gets so big that it uses all the asset income, and then the Fed has to go to the government for a bailout, which means the tax payer supports the banks getting their huge interest payments.

As an aside, does anyone know if the big banks can go to the Fed and add money to their deposits to earn the above market rate? Banks are supposedly free to with draw the accounts created out of thin air to pay for QE purchases, but can they add to those deposits? It would seem not because the amounts would rise even more dramatically.

Rocky Humbert replies: 

Bud: If your head is spinning, I suggest you sit down. If you look at the situation as I articulated it, then don't you agree with my analysis….? (This is a macro-economics conversation. No conspiracy theories allowed. ; ) Namely, the Fed could theoretically exist with only $1 of equity. Their equity is irrelevant because of their ability to print currency. And so long as the currency is accepted and relatively stable, everything works. For the Fed, currency is the same thing as a paper check. So if Citibank and the other big banks say "we want to withdraw $X trillion in excess reserves" the fed can hand them a check for $X trillion. And Citibank can take that check and spend it however they want. Whether the check has a picture of Ben Franklin or looks yellow or purple or is electronic is not material. It's credit creation… (This is when the S-Man chimes in.) I believe that before the Fed existed, this was how all banks operated — namely, there was essentially no difference between XYZ Bank's check/draft, their self-issued currency, etc etc.

Rudolf Hauser writes:

There is a bit of misunderstanding here. A reserve balance at the Fed is a bank's checking account at which it holds bankers money. That is the only money, other than currency, that another bank will accept in payment unless it is willing to keep a deposit in the bank that is in the negative position of the transaction. When a bank wants to reduce its balance at the Fed, it does so by buying other assets, such a T bills, or making loans. The seller or borrower now either deposits that money in their own bank or makes loans. This process continues if no other bank receiving deposits or proceeds of sales of assets to these spenders decides to hold excess deposits. Eventually enough ends up in checking accounts so that all the excess reserves reduced by the first bank have either become required reserves or held by other banks that have increased their excess reserve balances. The Fed does not have to sell any assets or pay out anything. The reserve balances just get moved around and converted from excess to required reserves. This of course increases M1 and M2 balances and is inflationary. If the Fed wants to avoid this it either has to make holding excess reserves more attractive by raising the rate it pays, selling assets it holds, borrowing cash via reverse repos or by converting excess reserves into required reserves by raising required reserves that have to be held against any checking or other accounts.

The risks are that eventually the banks might want to reduce excess reserves, resulting in a expansion in M1 and M2 that will be inflationary. Real growth is being held back by factors other than lack of liquidity. While faster M1 and M2 growth might push some demand forward in time resulting in some temporary faster real growth, the type of growth that would clearly have to lead to higher prices for either assets and/or goods and services. Alternatively, the Fed could take the measures noted above. It's ability to pay more on excess reserves is at some point limited by what the Fed earns on its assets and the amount of equity it has. But do not forget the first hit is on the U.S. Treasury which is currently getting large contributions from the Fed, which pays most of its profits to the Treasury. This is currently a large cushion. Selling assets will cause interest rates on those assets to rise, potentially considerably depending on how much the Fed sells among other factors. Even if the Fed does not try to upset the situation, rates might rise because of actual and expected inflation. This might create problems for some holders of long term debt and securities. The least destructive way might be to raise reserve requirements, but this might create problem to the extent that excess reserves are not evenly distributed among the banks. All these moves would be politically unpopular. This is why I am somewhat skeptical of the Fed to get us out of this situation. They could do it, but it will require a FOMC with a lot of wisdom, determination and courage to do so and a Congress that does not take away the Fed's nominal independence to pull off.

anonymous writes: 

Zerohedge quotes Bridgewater on the process of QE noting that not just the amount spent, but what it is buying dictates what the economic effects are. If the assets are more risky and less like cash, the effect is supposed to be more. Seems to me the creation of new money is the big cause of the effect. and then how that money is used is the other half of the equation. It's my view that the new money sits on the Fed balance sheet and impairs its inflationary effect. The reason it sits as excess reserves is that the Fed pays above market rate on the deposits. The $ 2.5 trillion times a reasonable interest rate in normal times of 4% would cost the Fed $100 B, and that is close to it current earnings for its assets of Treasuries and MBS Rising rates is not good for the Fed either.

Bridegwater and commentary:

In the past we have explained how QE continues to "fail upward" because instead of injecting credit that makes its way into the economy, what Bernanke is doing, is sequestering money-equivalent, high-quality collateral (not to mention market liquidity)- at last check the Fed owned 33% of all 10 Year equivalents - and by injecting reserves that end up on bank balance sheets, allows banks to chase risk higher in lieu of expanding loan creation. Alas it took a few thousands words, and tens of charts, to show this. Since we always enjoy simplification of complex concepts, we were happy to read the following 104-word blurb from Bridgewater's Co-CEO and Co-CIO Greg Jensen, on how QE should work… and why it doesn't.

The effectiveness of quantitative easing is a function of the dollars spent and what those people do with that money. If the dollars get spent on an asset that is very interchangeable with cash, then you don't get much of an impact. You don't get a multiplier from that.

If the dollar is spent on an asset that's risky and very different from cash, then that money goes into other assets and into the real economy. That's really how you see the impact of quantitative easing. What do they buy? Who do they buy it from? What do those people do with that money?

Of course, this is why sooner or later the Fed will proceed to "monetize" increasingly more risky, and more non-cash equivalents assets, until "this time becomes different." Which it never is, but the Fed will still try, and try and try.





Speak your mind


Resources & Links