A bank that takes short term deposits and lends long term. How much insurance does the bank need to protect against deposit run?
Consider a flow of long term loans and a flow of short term deposits. We know those rates, they are the 1,Y,Z chain, and that ratio goes toward three as you count up. So, if we serialize transactions, and keep within bound error, we would find the two flows would be equally panicked and neither the one or the other over run. Deal flow is the key, maintain serialization. Bank runs are about sudden parallelism.
But banks are not quite serialized. But, I can create the equivalent contract in which the match is serialized using Eth2.0 and even Swift codes. The chain manages dep[osits ins and loans outs to keep the ratio until the banking contract collapses or times out. Each set of matches exits the contract and spins out into a separate payments contracts between parties.
I have the equivalent problem, actually, I have to measure miner fees to enter and exit my banking contract. That banker/miner fee is the insurance, it puts depositor and borrower in a guaranteed path, fee paid ex ante. And estimating that fee is an inverse estimate of insurance, it is the cost of smoothing the volatility.The better the bank serializes, the safer the matching, the lower the insurance cost.
This is a theoretical model on counting the risk of time in a closed bounded system. But there is no absolute mechanism to extract time, it only assume a consistent demand for thing marked as a time unit. You will still be stuck, going one layer up, looking at the sandbox in aggregate, and isolating consistent time units.
No comments:
Post a Comment