Mark Thoma's Response

As an educated guess, the interest the Fed is paying on total reserves is probably about 3 billion dollars a year. This is money that would go to the taxpayers. So, what is the purpose of paying interest on reserves? One can conclude that the Fed really doesn't want the money multiplier to increase.

The Fed could pay that 3 billion on required reserves only, and then begin charging interest on excess reserves. Banks that loan out their excess reserves get to keep their payments on required reserves. Banks that don't loan out their excess reserve pay an interest penalty from the money they receive from the Fed for their required reserves.

My view of the Fed's paying interest on bank reserves has a different slant.

The Fed, in the past, paid any excess profits to the U.S. Treasury. Now they pay some or all of the excess profits to the banks. Another form of bailout?

If there is no or low loan demand even at low interest rates, the banks would have lower income than if loan demand were higher. Interest payments on reserves increases bank income.

Ironically, the Fed is earning more because it purchased some of the lousy assets of the banks, that still may be yielding some income but have very high risk. Thus the Fed has taken on the risk from the banks, but pays the banks income from those risky assets.

What a deal!

It would help economists a lot if they had worked in the real world, as Kling has. The interest rate isn't the only thing that businessmen consider. It's a major factor, but not the only factor. The problem businesses face today, and the major problem of the Great D, is uncertainty. Businessmen are still trying to figure out the costs of the healthcare bill. Now they have to deal with financial reform and the possibility of multiple stimuli, tax increases, etc.

Sometimes doing nothing is the better part of valor on the part of the state. Doing nothing allows the waters to clear so that businessmen can make reasonable forecasts. The frenzy and fury of political activity in Washington has created so much smoke and noise that no reasonable businessman would dare try to do anything but hold cash.

How about another Scott Sumner idea? If reducing the interest rate from one quarter of a percent to zero on reserves is not enough to get banks to reduce their excess reserves, then how about a negative interest rate (a penalty) on holding reserves? The Fed could increase the penalty on reserves until the desired level in the reduction of excess reserves is obtained. Certainly, there would be a point where banks would decide that the penalty is too great to hold excess reserves and would be inclined to lend them instead.

For years the Fed has been concerned about the shrinking share of financial intermediation taking place in banks, as opposed to nonbank financial institutions. One of the costs that comes with being a bank was the "reserve tax": banks had to hold a fraction of their deposits in a reserve account that paid no interest. Even before obtaining Congressional permission to pay interest on these reserves, the Fed lessened the tax by allowing things like cash management and sweep accounts that greatly reduced required reserves. Paying interest on required reserves is just a continuation of this effort to reduce the cost of being a bank.

Paying interest on excess reserves, however, has an entirely different rationale. It allows the Fed to change the funds rate without having to engage in open market operations to add or drain reserves. This lets the Fed vary the size and composition of its balance sheet without affecting the funds rate, giving it a second policy tool.

The problem with this, of course, is that banks are only going to lend to the private sector if the return they expect from doing so exceeds the return they expect from holding the funds as excess reserves. Paying interest on required reserves works at cross purposes with quantitative easing.

While paying a mere 25 basis points on excess reserves may not seem like it could have much effect, ending the practice would send a strong signal that the Fed is seriously trying to stimulate the economy. Conversely, not ending it send the opposite signal, and expectations do matter.

Arghh! Of course I meant

"paying interest on excess reserves works at cross purposes..."

I gotta get a better proofreader.

Keynesians are blind to the supply side. They also often seem to argue that supply side changes don't have any effect is demand side is low. This is very strange to me.

But Arnold, what do you think of then going further, if cutting the rate to zero wouldn't do enough. Why not make it negative like Sumner recommends?

I'm not predicting that this is what would happen, but wasn't the original justification for paying interest on reserves, was that the Fed wanted to prop up the fed funds rate?

In other words, back when they were pushing down their target, at some point (I seem to remember) they wanted to apply more "monetary stimulus" but didn't want the fed funds target any lower. So, they had to put a floor under the actual fed funds rate by paying interest on excess reserves. (Since nobody is going to lend to another bank at a lower interest rate than the Fed itself is paying.)

Does anyone else remember this academic discussion back in the fall of 2008? Is it possible that eliminating interest on reserves would just make the fed funds rate fall to basically zero?

Or, we could allow the market to determine interest rates. If banks are leery about making loans, they need to feel like it's worth making the loans. Artificially lowering the interest rates more isn't going to do that. Artificially low interest rates are what got us in this mess in the first place! The banks need to feel comfortable making the loans, meaning they believe they will profit from making them.

Now, I know that what we are talking about is reserve interest rates -- but that (and other government policies) has the effect of artificially driving down loan interest rates.

In fact, if interest rates were allowed to adjust to the market, an increase in the rate would actually probably make people feel much better about the economy as a whole. Cheap money makes people stupid. Higher interest rates would send a signal that people are going to be acting less stupidly with their investments, meaning those investments will be more profitable long-term, and this will create more confidence in the economy.

Actually it looks like dropping to negative wouldn't do any more.

Sumner recommended this, and I thought why not, but after thinking about it, it looks like it shouldn't do more than dropping the rate to zero.

If you drop the rate to zero, then the banks should move all of their excess reserves out at least to t-bills, to get even the tiny bit above zero return that they offer (as long as it's enough above zero to cover transactions costs, and they think the reserve rate will stay at zero at least as long as the t-bill term).

So dropping to zero looks like it should do the max you can do – all excess reserves are moved out. You can't do more than that even with a negative rate.

Plus, with a negative rate banks have to pay a percentage on their required reserves to the Fed, and that sucks some money out of the economy.