The prospect of increased supply has some saying further tweaks to the central bank’s policy-setting tool will be necessary.

“While no one knows exactly where the funds rate will set, there are reasons to think that it may continue to rise, in which case another adjustment to IOER would be warranted,” said Brian Sack, director of global economics for the D.E. Shaw Group and former head of the New York Fed’s markets desk.

TD’s Goldberg expects the Fed to adjust the IOER rate again as early as its September meeting, and sees fed funds converging with IOER by the end of the year.

Fed Futility

Others argue such policy action will prove futile.

Because the cause of the drift is the result of a glut in short-end supply, only coordination with the Treasury can solve the problem, according to Credit Suisse Group AG analyst Zoltan Pozsar.

“When the Fed’s floor is getting higher, the solution is not to lower the ceiling too, because then markets are going to suffocate and start punching holes in the ceiling, Pozsar said. “I find it odd that the Fed’s response to rising rates is to lower IOER when the issue is about Treasury issuing more bills.”

It’s been less than three weeks since policy makers cut by five basis points the rate they pay on excess reserves relative to the top of their official target range, and the effective funds rate has already crept higher once again. On a number of days last month it settled just three basis points from the IOER rate and eight basis points from the central bank’s upper band. The gap was four basis points on Monday.

The culprits, according to Wrightson ICAP, are likely domestic banks that have ramped up their bidding for fed funds. U.S.-based firms took 23 percent of the market share in March, up from an average of 17 percent in 2017, according to economist Lou Crandall, based on the Fed’s expanded summary of selected money-market rates. The surge in bidding likely contributed to the repricing of fed funds in mid-March, he said, and signs point to similar demand last month.

Liquidity Problem