The Federal Reserve has backed itself into quite a corner with its three rounds of “quantitative easing.”  In order to make sure that the resulting increase in bank reserves did not cause hyperinflation, the Federal Reserve has had to keep those reserves quarantined in bank vaults, away from bidding on the prices of commodities in the real economy.  To persuade banks to keep their money sitting idle in vaults, the Federal Reserve has had to pay interest on the excess reserves that resulted from those quantitative easings.

The Federal Reserve’s quantitative easings (creating electronic money out of thin air and using that money to buy various financial assets on the open market) brought about an expansion of the monetary base that was unprecedented.


Normally, this would have produced Weimar-like hyperinflation.  I’m not exaggerating. Just look.  We went from $0.8 trillion in the monetary base to $4 trillion in monetary base in about 5 years.  That’s about a 100% annual rate of inflation.  The problem is, when you get that level of inflation, people start to react to it by dramatically increasing the velocity of money (trying to get the increasingly-worthless hot-potato money off their hands before it loses more value), which would have started to increase the rate of inflation exponentially as the same dollar bills got to bid on more and more commodities in a period of time.

So, why hasn’t there been hyperinflation?  Because, as I’ve said, the extra monetary base has been mostly quarantined in bank vaults in the form of excess reserves.  How did the Federal Reserve accomplish that?  By paying interest on excess reserves (IOER) to make holding that money in bank vaults more attractive than making new loans on the basis of that money to the real economy.

At first, the Federal Reserve did not have to pay much interest to coax that money to stay in bank vaults—only 0.25%.  The reason was, banks saw few prospects for making good (credit-worthy) loans to companies at any interest rate higher than 0.25%.  That’s how bad the economy was during the bottom of the Great Recession.

But as the economy (haltingly) improves, banks will become more and more tempted to make loans on those excess reserves.

Technically, as many people have rightly pointed out, banks do not “loan out excess reserves.”  No, it is far worse.  They make loans using excess reserves as their required “fractional reserve,” which currently by law is mandated to be 10% of the loan amount.

What this means is that, although banks have $2.3 trillion in excess reserves, they could actually loan out $23 trillion in new money.  Yeah….

And when those loans are made, the reserves (no longer “excess,” but now “required”) stay in the banking system.  In fact, although banks might shuffle some of those reserves between themselves, there is really no way for those reserves (and thus the fractional base for loans) to be drawn down except by the Federal Reserve REVERSING quantitative easing and selling the financial assets on the Federal Reserve’s balance sheet in exchange for bank money (which the Federal Reserve would then electronically delete from existence once it collects that money).

Commentators often focus on the Federal Reserve’s balance sheet and explain that the balance sheet can be gradually decreased without reversing QE, but instead simply by holding the financial assets to maturity.  While this would indeed decrease the Federal Reserve’s balance sheet, it would do nothing to reverse the expansion of the monetary base.  It is the expansion of the monetary base that is the real problem, of which the Federal Reserve’s ballooned balance sheet is only a symptom.  The only thing that having the Federal Reserve hold those assets to maturity will do is remove those assets as something the Federal Reserve can sell back to the market in exchange for clawing back some of the monetary base it injected.  The more those things are held to maturity, the more permanent the expansion of the monetary base will be.

Thought-Experiment – If the Federal Reserve Ended IOER

As a thought-experiment, what would happen if banks this next year suddenly decided to make new loans off of all of their $2.3 trillion in excess reserves (perhaps in response to an improving business climate promising that most of those loans would be paid back, and/or the Federal Reserve ceasing the payment of interest on those reserves, thus no longer giving that money a reason to stay cooped up in banks without loans on top of them)?  Well, the M2 money supply would increase by not $2.3 trillion, but $23 trillion (due to the 10x multiplier of only needing a 10% fraction as a base for the loans).  Here’s where M2 money is right now:


It’s at about, let’s say, $13 trillion right now.  It would be expanded to $36 trillion in one year, which would be an annual increase of 175%.  That gives us a lower-bound for our rate of inflation that year….

But wait, there’s more.  Remember when I said that the velocity of money really picks up when people expect inflation?  That’s true, in which case a “modest” amount of inflation can serve as a seed for hyperinflation, as people increasingly react to inflation by changing their expectations and spending money as quickly as possible.  But it’s also true that the velocity of money typically increases during the boom phase of any business cycle.  We don’t even need to postulate huge changes in psychological expectations to see how changes in the velocity of money would further increase the inflation.

So let’s be conservative in our thought-experiment and imagine only that the velocity of M2 money returned to where it was during the boom of the late 1990s.  (As I’ve said, it could go much worse!)


Going from 1.5 to 2.2 would be another roughly 50% increase.  So add that to the inflation rate.

We’d be looking at a yearly inflation rate of about 225%.  By the end of 2016, the price of gold would be about…$3900 under this scenario.

The Lesser of Two Evils

Of course, the Federal Reserve will not allow this sort of inflation to happen (we assume…).  Instead, the Federal Reserve will continue to raise the interest on excess reserves (IOER) so that keeping those reserves as idle excess reserves will be competitive with using those reserves as the base for 10x the amount of new loans.  Making new loans will become more and more attractive as the economy improves (and borrowers will be willing and able to pay higher and higher rates of interest), which means the Federal Reserve will have to keep raising the IOER to keep pace.

Some commentators have been incredibly glib about this reliance on the IOER as the Federal Reserve’s new primary tool for keeping inflation in check.

(From the link above)—”Basically, the payment of interest on reserves allows the Fed to maintain control of the Fed Funds Rate even when the balance sheet is expanded.  So the Fed can raise interest rates no matter what the size of the balance sheet is because it can simply increase the rate of IOER.”

“Basically”?  “Simply”?  Oh yeah, this IOER thing will be a piece of cake!  Why didn’t the Fed think of this earlier?!  Okay, everybody go home, everything’s fine, nothing to see here….

Needless to say, there are other bourgeois commentators who are less enthusiastic about relying on IOER…

There are four major problems with IOER that I see:

  1. It is basically a taxpayer subsidy to the banks.  Right now it is “only” $6.5 billion per year, but if the IOER rate were to go up to something like 4%, it would be closer to $100 billion per year.
  2. It is baking-in more inflation in the long term because IOER is directly increasing bank reserves without investing to expand production in the real economy.  So if IOER is ever unwound, it will have more and more of a severe aftermath the longer it has been in place.
  3. It could cause “supply-side inflation” if the IOER rate were to go above the traditional target inflation rate of 2%.  Imagine that the IOER is 3%.  That means that banks could then reap a real, inflation-adjusted return on their money of 1% by just letting their cash sit idle.  A riskless 1% return that is perfectly liquid (not tied up in bonds or other long-term instruments)?  Not only would existing reserves stay cooped up in bank vaults to get this great interest rate, but even new money that is currently tied up in the real economy would abandon the real economy and flock into bank vaults.  Hey, instead of risking your money running an enterprise like some old-fashioned dweeb, how about you get in on this riskless 1% action?  Of course, if money exits real production in order to chase this 1% real return in bank vaults, the actual amount of goods produced will fall.  With supply reduced, prices will rise.
  4. IOER is a blunt and leveraged instrument.  For example, right now the IOER is at 0.50%.  If it were at, say, 0.10%, it is possible that ALL of that $2.3 trillion in reserves would be establishing new loan money at higher rates instead (and thus, hyperinflation).  If it were at, say, 1.0%, it is possible that a lot of existing productive capital would exit the real economy and flood into bank vaults to obtain that comparatively high interest rate on idle reserves, thus depressing prices (this may already be happening with the 0.50% rate, to some extent—hence the current fear of deflation).  The Federal Reserve now has $2.3 trillion (or possibly more) that it is implicitly responsible for direction—money that could swing wildly in either direction (real economy vs. idle excess reserves) based on small changes in the real economy (the rates borrowers are willing to pay for loans) or the IOER.  Really, the Federal Reserve should be raising in 0.10% increments at more frequent intervals if it wants to use this blunt, leveraged instrument with more finesse.

These problems will only get worse the longer IOER is in effect.  The only way to unwind IOER, though, will be to unwind QE, or tolerate hyperinflation…and the Federal Reserve has implied that it will not be doing the latter two any time soon….