Kling's latest post on Recalculation and the debate over EMH and Bubble prediction requires a response.
Kling:
"However, I would suggest that explaining the depth of the recession is a challenge for just about any macroeconomist. There is no well-established theory that can explain how we got to 10 percent unemployment."
I go with James Hamilton on the global oil readjustment. Energy is highly correlated, the actual core, of transaction costs. Which Kling mentions later:
"Tightness in the oil market means that we have to convert to less oil-intensive patterns of consumption growth and productions. Just as in the 1970's, this creates big adjustment problems."
Another issue in the Blogs are the interviews between the Economist and the Chicago School about bubbles and EMH. The Coordination Blog has a post on that. Can we predict bubbles? Absolutely, if one is on the inside of the fraud as was Goldman Sachs. A related issue is whether price signals can predict bubbles. No, not necessarily because price is simply the approximate estimate of inventories by finance, a variance estimate. The estimate is only valid during relative smooth economic times, the real market norm being entropy.
Can entropy measures predict bubbles? Yes, better than price signals. Look at divergence measures, however the timing is difficult to predict.
Another point, was innovation stopped by a barrier? Kling:
"The recent innovation slump was disguised by the housing boom. That is, if you take away the housing boom, you would have seen a steady increase in unemployment, due to the lack of new business formation."
Yes, true in fact. If one followed the innovation bubble, it moved through the economy, in each turn attacking a constraint on transaction costs. The innovation moved from producer to household, then was stopped at transportation. It stopped at transportation for one simple reason, Robots are not allowed drivers' licenses, a government problem.
In other words, government prohibits the application of innovation to public goods.
No comments:
Post a Comment