Let’s Talk About Seigniorage

Scott Sumner has a post up this morning quoting me in a bunch of out of context quotes from one of his readers, Mark Sadowski.  Scott finished the post asking:

“I just have one question for Mr. Roche.  If the Fed buys currency at par, how did the Fed earn its seignorage [sic] prior to 2008?”

I had been having a friendly talk with Nick Rowe over the past few days and I pointed out a contradiction within Market Monetarism that is strange – the fact that Market Monetarists often like to talk about how bank deposits are convertible on demand into base money.  I pointed out that coins are purchased from the US Mint at face value (I also once accidentally said notes were as well, but later pointed out that that was a typo, something Mark Sadowski was not gracious enough to point out).  This point is not important, however.

The point that is important is that the Fed buys coin from the US Treasury (the Mint is an arm of the Treasury) at face value and obtains credit in its account at the Fed.  The US Treasury earns seigniorage when it sells coins to the Fed.  And the Fed sells coins to the public at face value.  In other words, you can always redeem bank deposits for coins at face value, which come directly from the US Treasury and are merely distributed by the Fed.  So there’s an internal inconsistency in the way Market Monetarists tell their story about conversion as they conveniently ignore the fact that the US Treasury determines par value for all coins in the system and makes them redeemable at par for bank deposits. In this regard, the Fed is a mere middleman in this transaction (as it is in most operations) and not this overarching “central” price setter as they so often like to claim.

Immediately after claiming I was “wrong” about this Mark Sadowski went on to contradict himself when he showed I was right about the coin situation and then tried to trivialize it stating that coins aren’t a large part of the monetary base, which is totally irrelevant because conversion at par is not necessarily a function of the SIZE of the monetary base.  It reminded me of how cash is a mere 0.2% of the value of transactions in the USA yet Sumner loves to place some special importance on the use of cash in the system – he calls it “paper gold”….

But none of that answers Sumner’s bizarre question.   So let’s explore that a bit further.  Seigniorage is just the difference between the cost to produce money and the actual value of the money.  So, if it costs $0.10 to produce a $0.25 coin then the Treasury earns the difference in seigniorage.  In a fiat system seigniorage is pretty different from the metallic monetary system days.   Of course, the Fed doesn’t mine up the monetary base today and hold it as gold or something.  It just credits accounts when it purchases assets from the private sector.  The interest on these securities is substantial and has always been the Fed’s primary source of revenue.  Of course, the Fed also remits over 90% of its earnings to the US Treasury so once again, the beneficiary of seigniorage is Treasury.  I don’t know where Scott Sumner thought the Fed earned all of its revenues, but he seems to be very confused on this topic (or he’s confused because one of his readers quoted me out of context, I don’t know).  So let’s keep exploring.

The Fed also buys notes from the Bureau of Engraving at cost and makes them available to the public on demand.   Banks can buy the notes at face value from the Fed.  Notes, however, do not count as part of the monetary base until they are sold to the banks.  And when this occurs the Fed simply swaps its liabilities with the banking sector resulting in no gain.  So, to answer Scott’s question – the Fed earns seigniorage primarily on securities it holds.  And it passes the vast majority of this seigniorage on to the US Treasury.  It only holds a marginal amount to cover future operating expenses.  And of course, the payments from those securities come from…the revenue US Treasury earns.

To summarize:  in a modern fiat monetary system notes are provided to the banking system as a mere convenience tool for bank deposit account holders and little else.  They bear none of the special importance that Scott Sumner claims when he compares them to “paper gold”.  And the vast majority of seigniorage earned on securities held by the Fed gets remitted to the US Treasury.  And lastly, coins are always purchased by the Fed at face value from the US Treasury who earns the seigniorage on them.   So who benefits from seigniorage in a modern fiat monetary system?  The US Treasury mostly.  And if you wanted to claim that there was one defining form of asset that everything was convertible into at face value it would undeniably be coins which are issued by the US Mint and sold to the Fed at face value.  So, in a weird way, the Treasury is the banker to the banker who it banks with.

Shorter summary: Market Monetarism is wrong, but this is only one of many internal inconsistencies about what has clearly become a confused version of Milton Friedman’s dying theories….