Consider price discovery as a dual queuing problem

In finance, for example, is there an optimum number of short term depositors to match the short term borrowers? How would we know?  How many bids are in the hopper? As a borrower, show me two to three deposit bids in the hopper.  As a depositor, show me one to two borrowers in the hopper.

Why those numbers? The most amount of inside information with the smallest number of transactions.  If the deposit queue is too big, the potential depositor holds on to cash, or jumps into the borrower queue.  And visa versa.

When the currency banker sees one or the other queue become too short or long, it will change the borrower or depositor rate, ex post facto, and re-balance.

Well, what is a transaction?

Sort of a big catch here.  No one knows shit until the monthly bills are paid at least, so start with monthly changes in deposit to loan.  So the Fed is dealing with queues, at equilibrium, in which member banks adjust their balances on  a monthly basis.

The Fed can change rates every three to five months, but this is extremely efficient for humans.  Add in the institutional delays, the sparse spectrum, and our frequency scale spans a generation.

So, we would expect that significant member bank account changes would be spread out over the week or so after the end of month.  Each account owner would sit at his slot machine and watch the queue.  At opportune moments, he.she would push some or all of their monthly account balances out to the hopper.  In the interaction the collective is performing a finite precision Huffman encoder, adjusting its window size to maintain pre-announced precision.

But its true.  Shannon Theory implied information transfer efficiency, which is spwctral efficiency. But the  theory can be boiled down to optimum queuing over the optimum network; where transaction and network both adapt, decompression with boson stats, the compression with fermion. Compression and decompression is the loan to deposit and its invers.

This idea is the essence of banker bot.

Adjust risk by telling the bot to be less precise, or have the bot lean the collective network left, you have inside info on a significant change in deposit to loans.  Your bot, sitting in your protected smart card, will sniff the transaction windows on your typical sequences. Then match your profile to the collective of saving and loan sites,using protected smart card transaction that place bets.

All doable, hardware through software security, hardware protected and masked keys, self organizing key blocks to verify key security, guaranteed honest account, when required and user consents, all of thi, as protected as gold.  Each web site having its own uniform smart card.

The hardware key chain is infinite precision, block chain.  Thus we can support the billions and billions of card holders.  Coin bankers are finite precision, because that is what currency bankers sell, a fix precision accounting system.

The finite precision encoeder is like a neural net, it locks the coefficients when  queues are optimum.  This is a quantization effect,forced by the fixed precision bankers.  The member banks are finding the best pricing 'quants', the nearest minimal encode/decode tree with fixed precision.

Precision refers to the information loss when transactions go from uncompressed to compressed. Lost information is quantization energy, dunno where it goes.  The compression ratio partitions the lenders and depositors, keeps the equipartion conditions sustained, we can do Ito's grammar.