Wednesday, January 17, 2018

Loan to deposit and multipliers

The basket model is stripped down, we cheat. We remove money transaction costs, then we auto trade to a neutral, or known profit function. As a result we aren't carry around lots of tiny baskets per person.

So we get this simple equivalence between Walmart and money. Everytime the panel truck delivers in the back, we get a card swipe, and each panel truck generates N customers out, each doing a card swipe. We assume stable queues, apply sphere packing assumption (all baskets optimally full). Then we see a coherence between the queue structure of the deposits matched to loans; and the queue structure of ins and outs at Walmart.

Loan to deposits is a Walmart panel truck to customer ratio. It is the really stripped down version of the money multiplier equation, one loan triggers N deposits.

The multiplier comes from a known congestion level, which we can call price uncertainty.  Walmart knows hoy many customers mill about in the store, its window size. These customers have uncertain baskets, not yet in the structured queue leaving.. In store congestion is a different uncertainty than the S&L currency risk or bit error. Thew in store congestion uncertainty is price uncertainty, it is what the  S&L attempt to discover.

Bit error is the tradebook uncertainty in reading the S&L book. Quantization separates uncetainties, so price uncertainty will be comed of tradbook uncertainy qiants.  Our assumption minimizes that bit error, we still always have the best quants to count prices.

No comments: