Interest rates and the reverend's parade of school girls

In my model I have two queues,depositors and borrowers.  I use maximum entropy conditions to prove a relationship between interest rates and queue size.

First, my assumption is that loan and deposit sizes are free to adjust, I use only transaction rates.  Later I cvan go back and how accurate my assumption is.

OK, the reverend has gone into the banking business, he has a queue of depoitors.  The reverend does not  want two depositors to get in the queue twice, he gets a redundancy.  Thiose two will form a wholesale lending bank and move up the banker distribution tree.

So, you see, the reverend wants all sets of unique depositors to be in his retail queue, his transaction costs are minimized.

What process eliminates redundancy in a transaction series?  The Huffman encoder, and the bankers tree is the Huffman encoding graph.  Liquidity bounds are set by setting the Huffman moving window size.

There, you have it.  The pros can add in the borrowers.   The Huffman encoder is really tracking the difference between the reverend input and output queues, using the hyperbolic model.  So the Huffman encoder, decode graph generates typical liquidity values when the banker's graph is  minimized..  Along those lines, I might think about is some.