It works in practise - BUT - does it work in theory?

This was the famous statement - allegedly by Chrusjtjov viewing a combined harvester demo at a sovhos.

It came to my mind when reading the often so eminent Economist - now the "Modern Economy Theory" July 18th article. Without going into any interesting detail  on this or on the rebutting views by Robert Lucas (Aug 6th) I think that we have ample evidence from past runs that most markets - especially financial ones - from time to time lose their sense.  Run hysterically high - only to fall into total depression - hardly any price is low enough and hardly any partner strong enough. There are many - almost too obvious - levers driving things exponentially in either direction - in the free fall phase destroying value for shareholders and taxpayers on an astronomical scale. Should it not now be clear that:

1. Mark to market is an absolute must for trading. It should not be necessary to say that trading is a very important activitity - but that position sizes should be restricted to capital deployed. The value of widespread liquidity created by trading can hardly be overstated - even if it disappears in crisis situations.

2. No mandatory mark to market for loans - irrespective of instrument used. We have seen examples of how investment-banking-encouraged and regulator-adopted mark to market practises have destroyed banks on a large scale - tax payers lose big time in the fire sales - and smart actors pick up spoils.

3. Disappearance of liquidity phenomenas should be an assumption - and investment portfolio carriers should have a much larger part of their funding assured than has been the case before.

Building theories is fascinating and useful - but we should not blame theorists for not foreseeing also-before-by-practioners-seen consequences of overheated markets. Incentive systems are sometimes leeding to even faster value destruction and begs the question if analysts are up to speed - or who they serve by - too frequently -  not looking beyond the next quarter.