Leverage and Theory

David Wessel presents a discussion among Mankiw and Greenspan about how much leverage is desirable for banks.

According to me:

1) Firms maintain inventory variance, but my theory says inventory variances are set to a constant by biology, for bankers the same should hold. 2) Finite precision fixes the number of terms, so space on the bankers yield is pre-determined. 3) The Bankers yield always tends toward Normal for utility reasons.

The key to determining the leverage ratio is to determine the number effective terms on the yield, do an entropy check (how many self correlated terms are on the curve). The inventory variance is proportional to the occupied portion of the curve relative to the whole curve, and by construction should be constant for all terms.

Which is to say, a proper banker's yield curve represents the biologically fixed SNR function, it is the finite diagonal of the error matrix in our Kalman filters. Within the limits of the environment, production systems try to produce the same SNR function in the aggregate.