Is Federal Reserve Policy In Violation Of The Law?
No, I don’t mean sections 8 and 10 of the Constitution’s first article — though goodness knows a case can be made (and has been made recently, and most eloquently, by CMFA Adjunct Scholar Dick Timberlake), that it hasn’t adhered to the letter of that law, either. I’m referring to the law authorizing the Fed to pay interest on depository institutions’ reserve balances, or IOR, for short.
I bet the Board of Governors makes some mean lemonade!You see, according to Title II of the 2006 “Financial Services Regulatory Relief Act” — that law that originally granted the Fed authority, commencing October 1, 2011, to begin paying IOR — the Fed is allowed to pay interest, not at any old rate it chooses, but “at a rate or ratesnot to exceed the general level of short-term interest rates.”
As the name of the 2006 Act suggests, its purpose was to relieve financial institutions of unnecessary regulatory burdens. The fact that depository institutions’ reserve balances at the Fed, including minimum balances they were required to hold, bore no interest, had long been regarded as one such unnecessary burden.
So long as reserve balances paid no interest, reserve requirements amounted to a distortionary tax on bank deposits subject to them. In the words of then Fed Governor Donald Kohn, who testified in favor of IOR back in 2004, the payment of interest on reserves, and on required reserves especially, would result in improvements in efficiency that “should eventually be passed through to bank borrowers and depositors.”
Since the original intent of IOR was to remove an implicit tax on deposits, and not to have the Fed subsidize those deposits, it’s easy to understand the law’s insistence that the Fed pay IOR only at “a rate or rates not to exceed the general level of short-term interest rates.”
It also easy to see why most economists, including the Fed’s own experts, treat the federal funds rate as an appropriate proxy for the opportunity cost of reserve holding, and hence as one of the short-term rates that the rate of interest on bank reserves ought “not to exceed.” Indeed, because overnight lending involves some risk and transactions costs, while banks would earn IOR effortlessly and without bearing any risk, the IOR rate should logically be strictly below, rather than below or equal to, the federal funds rate.
The Law Tweaked
I wish someone deeply committed to making sure that financial institutions don’t get away with any hanky-panky would go ahead and sue the Fed.Fast forward to 2008. Among its other provisions, the “Emergency Economic Stabilization Act” passed on October 3rd of that year “accelerated” the Fed’s authority to pay interest on bank reserves, making that authority effective as of the new law’s passage, instead of as of October 1 of 2011. Significantly, the 2008 measure did not otherwise alter the language of the original legislation. The rush to implement IOR was, nevertheless, based on motives quite different from those that informed the 2006 Act.
As then-Chairman Ben Bernanke explained, in an October 7th, 2008 speech he gave at the annual meeting of the National Association for Business Economics, in the wake of Lehman’s failure, the extent of the Fed’s emergency lending had begun to run ahead of our ability to absorb excess reserves* held by the banking system, leading the effective funds rate, on many days, to fall below the target set by the Federal Open Market Committee.
This problem has largely been addressed by a provision of the legislation the Congress passed last week, which gives the Federal Reserve the authority to pay interest on balances that depository institutions hold in their accounts at the Federal Reserve Banks.
The Federal Reserve announced yesterday that it will pay interest on required reserve balances at 10 basis points below the target federal funds rate, and pay interest on excess reserves, initially at 75 basis points below the target. Paying interest on reserves should allow us to better control the federal funds rate, as banks are unlikely to lend overnight balances at a rate lower than they can receive from the Fed; thus, the payment of interest on reserves should set a floor for the funds rate over the day (my emphasis).
Thus, although the Fed was now chiefly concerned, not with relieving banks of an implicit tax, but with reinforcing its ability to hit its federal funds target, its plan was to do so in a manner that nevertheless complied with the letter, if not the spirit, of the 2006 law, by having IOR serve as a new, non-zero lower bound to the federal funds rate.
Alas, things didn’t work out quite as Bernanke and other Fed officials intended. Instead, the “floor” they’d laid out so carefully turned out to be rotten, chiefly because, although they keep balances with the Fed, Fannie and Freddie and other GSEs aren’t eligible for IOR.
Consequently, their involvement created an arbitrage opportunity that Fed officials hadn’t anticipated, with banks borrowing funds from GSEs overnight at rates sufficiently below the IOR rate to turn the banks a tidy (if modest) profit.
The crux of the matter is that the effective federal funds rate — that is, the rate that was actually being paid for overnight funds — quickly ended up falling below the IOR “floor” and, therefore, below the Fed’s target rate. In fact, as the red plot in the figure below shows, since December 2008, when the Fed set the IOR rate for both required and excess reserves at 25 basis points, the rate has always exceeded the effective federal funds rate, besting it on occasion by as much as 19 basis points.
Faced by this reality, the Fed made the best of a bad job by declaring (1) that instead of setting a fed funds rate target it would henceforth set a target “range;” and (2) that the rate of IOR was to define, not the lower bound (or “floor”) of the new target range, but the upperbound.
Man, I bet the Board of Governors makes some mean lemonade!
But while the Fed may have succeeded in saving face, it doesn’t follow that it managed to do so while still obeying the law. For converting the IOR rate from a floor to a ceiling meant setting it above rather than at or below “the general level of short-term interest rates,” taking that “general level” to be appropriately represented by the effective federal funds rate.
Nor does letting the three-month T-bill rate proxy the “general level” of (risk-free) short term rates — a reasonable alternative —