So, the Federal Reserve just raised the federal funds target rate again to 0.75%. I wouldn’t consider this particularly noteworthy, except for the way that the Fed is doing it.

It used to be that commercial banks tended to expand their loans on their balance sheet up to (or even slightly past) the point where they had just enough reserve funds (whether through opening checking and savings accounts for depositors and holding their money, or by soliciting short-term loans from other banks) to cover the fractional backing requirement for those loans. After all, “excess” reserves that were not backing loans was just money that was sitting there doing nothing for the bank, earning 0% interest, as opposed to a much better risk-adjusted rate of return that they could get by pyramiding those reserves into a 20x greater loan (assuming that the reserve requirement stays fixed at 5%).

It used to be that, when the Fed wanted interest rates to increase, they would sell Fed assets (usually govt. treasuries) in exchange for bank reserves (money), thus vacuuming up some reserves from the banking system. That made it harder overall for banks to get emergency loans to shore up their reserve requirements after a busy day of making new loans, and those desperate banks would bid up the interest rates on those loans until such a point that it became pointless to extend new loans at a certain going interest rate because the banks would have to pay just as much interest for the reserves to back it. After an evening of musical chairs, some banks would sometimes fall short of securing short-term loans to meet their reserve requirements, and they would have to come crawling to the Federal Reserve for a loan from the “discount window,” which was another way that the Federal Reserve could influence what the competitive going interest rates were.

All of that is out the window now thanks to the several rounds of Quantitative Easing (QE), which were basically just open market operations on steroids.  In isolation, the QEs would have allowed for an explosion in the amount of commercial bank loans thanks to the extent to which they increased the monetary base of reserves held by those commercial banks.

However, at the same time that the Federal Reserve was doing its QEs, it also started a new policy meant to counter-act the effect of its own QEs, interest on reserves.  (Why the Fed would want to do a huge QE, and then shoot its previous QE in the foot with IOR, instead of just doing a smaller QE in the first place, has to do with the Fed’s idea of “Operation Twist,” in which they wanted to specifically bring down long-term interest rates without unleashing inflation).

Paying interest on reserves has disincentivized commercial banks from using those reserves to reach for riskier gains when they could just sit back and reap risk-free returns from the Fed, especially if borrowers were not willing to take on debt at interest rates much higher than the interest on reserves.  The slight added margin wouldn’t be worth the risk.  So, commercial banks have sat on these reserves.

However, as the business cycle improves, borrowers are becoming more eager for loans and feel more confident about being able to pay them back, even at slightly higher interest rates.  This would ordinarily tempt banks to starting writing more loans (at those slightly higher interest rates), threatening to explode the broad money supply.  Therefore, as the business cycle improves, unless the Federal Reserve wants to undo its QEs, or unless it wants to tolerate an explosion of inflation above its declared 2% target, the Fed must increase the interest on reserves that it pays to commercial banks to make that money stay put rather than chase greener pastures.

Some economists seem to see nothing wrong with interest on reserves and assume that it will become the new normal, but I see reasons for why interest on reserves is unsustainable, and it has to do with the Fed’s balance sheet.

Why the Fed Balance Sheet Matters

For my understanding of why the Fed balance sheet matters, I must thank Mike Sproul for explaining it in his many comments on blogs such as those of David Glasner (uneasymoney), Nick Rowe (, Scott Sumner (themoneyillusion), and JPKoning (moneyness).

Here’s the thing:  people assume that the Fed plays by entirely different rules because it has access to the printing presses.  It cannot go out of business if it makes bad deals.  It does not need to make a profit to continue to exist.  All of that is true.  However, if the Fed cares about the integrity of the dollar as a currency, it would behoove the Fed to act as though it was a normal bank.  In other words, if it cares about the dollar, the Fed ought to not make a habit of giving things away for free.  For every dollar it issues, it ought to try to have a dollar’s worth of assets in its possession.  Why?

It is true that the Fed’s dollar “liabilities” are unlike typical liabilities in that the Fed can never be put under the legal obligation to redeem outstanding dollars for Fed assets.  That’s true.  But the Fed has a practical obligation to offer such redemption if it wants to protect the value of the dollar and hit its inflation targets.

Dollar velocity and the demand for dollars are contentious topics.  For the moment, I won’t delve into what determines these.  However, I think it should suffice to say for our present purposes that these two things are not stable.  As much as we would like to assume that velocity stays constant, we have empirical evidence that it doesn’t.  Likewise, demand for dollars can suddenly increase or decrease for reasons that are difficult to understand.  Regardless, there may come times when the velocity of money surprisingly increases or the demand for dollars suddenly decreases.

When these things happen, if the Fed wants to hit its inflation target, it must be ready to sell its assets on its balance sheet in exchange for vacuuming up as much base money as it needs to in order to bring the supply of dollars in line with the new demand for, and velocity of, dollars.

To be perfectly prudential, the Fed ought to have enough assets so that it can vacuum up ALL of its outstanding dollars, if need be.  That way, no matter how low the demand for dollars temporarily dips, or how high the velocity gets, the Fed can always decrease the supply of dollars to make the dollar hit its inflation target.  That would give the dollar supreme credibility, and would dissuade people from even testing the Fed with a temporary run on (away from) the dollar.

Another thing that would be prudent for the Fed to do is keep some assets on its balance sheet whose values are not denominated in dollars.  The reason is, if the value of the dollar ever starts to slip, that is precisely when the Fed would need to rely on selling its assets to vacuum up enough dollars to re-establish the desired target.  Yet, as the dollar loses value, so will, to the same degree, any assets that the Fed has that are denominated in those same dollars.  A government bond or an MBS promising payment in dollars will lose value precisely proportional to the loss in the value of the dollar.  The Fed could find itself chasing its own tail.  However, a large gold hoard that the Fed would be willing to partially sell off would be a stabilizing influence because, as the dollar lost value, a given ounce of gold would buy (vacuum up) more and more dollars, thus making the problem actually easier to correct as it became worse.

How the Fed has harmed its balance sheet since 2008

The Fed has done two things to impair its balance sheet since 2008:  buying toxic MBS assets during its rounds of QE, and paying interest on reserves.

Banks have only been too happy to get a lot of their MBS assets off their balance sheets and onto the Fed’s balance sheet since 2008.  I imagine that the Fed was betting that those assets only seemed toxic at the time due to an abnormally distressed situation on the financial markets.  However, what if those assets really are toxic in a fundamental way?  What if too many of those MBS tranches have home loans that will never be fully paid back, not during a recession, not during a boom, never?  If so, then the Fed overpaid for those assets.  If the markets correctly perceive this, then when the Fed eventually goes to unwind its QEs and sell those assets back to the market, the market will only accept those MBS assets at a discount compared to what the Fed originally paid for them (let us disregard for the moment changes in price due to partial maturity of the assets in the intervening time).  What that means is that, even after all of the MBS assets are once again off the Fed’s books, some dollars that were originally injected into the monetary base from the original QE will remain permanently stranded out there, permanently swelling the monetary base (unless the Fed wants to deplete some of its other assets to compensate).  In any case, the Fed’s ammo for vacuuming up dollars in the future, should it need it, will have been permanently depleted.

As for interest on reserves, there the Fed is giving commercial banks value without getting anything for the Fed’s own balance sheet in return.  Some commentators talk as if this is to compensate for the fact that the Fed is getting a constant revenue stream from its financial assets bought in the QEs, and the banks are not.  This is a red herring, though.  Those revenue streams were already factored into the open-market prices of those financial assets when the Fed originally bought them.  The banks have already been paid for them.  The banks freely chose to trade a lump-sum of reserves for an ongoing (and uncertain) revenue stream, and the banks were able to freely hold out as long as they wanted until the Fed was offering enough of a price to allow the Fed to buy all of these revenue streams to hit its QE targets.  The Fed never “forced” banks to part with these revenue streams.  That’s what “open-market operations” means.  The Fed merely had to increase the offer price until the banks were willing to sell them however much the Fed wanted to buy.

The Fed’s paying commercial banks 0.75% interest on its reserves each year is a bit like the Fed having overpaid originally for those financial assets by 0.75%, multiplied by however many years this is allowed to continue and adjusted according to whatever the average interest on reserves payment happens to have been when integrated over the whole time from the start of the QEs until the unwinding of the QEs.

Interest on reserves is a permanent increase in the monetary base that the Fed has no way of vacuuming up if the Fed should ever need to do so.  The Fed has nothing to offer for that money.

There is a way out, however:  the fiscal authority could compensate by doing either of two things that amount to the same thing:  bestow non-monetary assets to the Fed’s balance sheet (taking them from somewhere else on the fiscal authority’s balance sheet, of course, such as social security) so that the Fed gets more ammo that it can use to vacuum up dollars, or run a tax surplus that is then destroyed (rather than recycled back into the economy via debt repayment and/or program spending).  These would alternative methods of vacuuming up dollars if the Fed’s balance sheet should ever become exhausted.

However, I find these alternatives politically unrealistic.  The fiscal authority is well-known to have an inflationary bias.  Would the public really tolerate their incomes being taxed, only to hear that that tax money would then be shredded (rather than going towards paying down the debt or some social program)?  This is why we have an independent monetary authority in the first place—to do the tough jobs that the popularly-elected fiscal authority won’t do.   If the “independent” monetary authority can’t do the job, why would we presume that the fiscal authority could?

In short:  the Fed’s balance sheet matters.  Let’s imagine that interest on reserves goes on for long enough and reaches a high enough level that fully one half of the monetary base was dispensed via interest on reserves, with the Fed having no assets on its balance sheet to show for it.  At that point, if the demand for dollars should ever suddenly drop in half, or the velocity of dollars should ever double (both plainly imaginable scenarios), the Fed would be out of ammo on its balance sheet for vacuuming up unwanted dollars, and inflation would be off and running.

And if the Fed should ever try to get away from Interest on Reserves, it would also need to unwind all of its QE or else risk an explosion of loan creation (and even if QE is unwound fully, I fear that there will be some permanently stranded dollars that the Fed will not be able to buy back with the toxic assets on its balance sheet).  Unwinding either QE or IOR will be painful.

Interest on reserves is a poisoned chalice.  Everything is treating it like a cure-all, but it will slowly poison the Fed’s balance sheet and tarnish the dollar’s long-term value prospects.