The ten year treasury yield has a lower bound

And the yield, blue line, seems to be stuck above 1.5%, because we are a deficit only nation so our creditors have a currency risk wedge.  But tax receipts, red line, have leveled off for the cycle, so the deficit is rising at about 5% per year.  For Congress, in an election year, the interest expenses of $580 billion/year will be rising 5% each year.  Congress is not liquid enough to handle the budget constraint.

What happens to Congress?

In the simplest currency model, a dual queuing system, then Congress would have exceeded the published liquidity bounds.  They would have incurred a liability that takes them off the member banker network, and a depository bank would have automatically taken the spot.  There would be no shame, the ex-member bank could very well be making more money outside the member bank business.  Not meeting liquidity constraints, for a member bank, means the short term money supply is likely fine and the member bank is off frying long term fish, nothing wrong there.  The member bank will just let itself fall on and off the member bank list as needed.

In the real world, who knows.  For all of us, this is our first monetary cycle.  The fix is to remove all remits back to Treasury, there is no such thing as stable monopoly ownership of the currency banker.  All of the assets and liabilities of the central banker sheet should be loan and deposits held by the member banks.  The central banker really is a spreadsheet function, it considers actuarial earnings as wasted electricity, and actuarial losses as useful electricity; the electric bills being around $2,000/yr, total.

The Huffman window size and the member bank list

My view, not shared, of the math is that we get a windowed, finite Huffman encoder.  In this math, fall off the member bank list simply means that the typical transaction of the member bank are random, inside the window.  This member bank is off discovering long term innovations, the currency function is widowing the short term fluctuations.  So, the accounts of the member bank do not appear in the member bank finite graph, the encoder/decoder graph built by the Huffman.  Precision is finite, generally bound, so another bank gets noticed as short term innovative, and that bank gets promoted up the graph.  These member banks begin betting on the current natural short term rate, based upon the future discovery of NGDP, of when Huffman makes the next rotation of the graph.