The Subprime Solution: How Today's Global Financial Crisis Happened, and What to Do about It

The Subprime Solution

How Today's Global Financial Crisis Happened, and What to Do about It

by Robert J. Shiller
(based on 21 customer reviews)

The Subprime Solution: How Today's Global Financial Crisis Happened, and What to Do about It (Hardcover)
Author: Robert J. Shiller
Publisher: Princeton University Press


List Price: $16.95
Amazon's Price: $11.53
You save: $5.42 (32%)

click for more info




Most useful review as voted by customers:
5 out of 6 people found the following review helpful.
Review Date: 8/27/08


3.5 stars-Shiller can't deal with uncertainty versus risk problem due to his allegiance to the SEU Rational actor model of

This could have been a major contribution to economic theory and history.Unfortunately,Shiller is unable to think outside the box of the basic neoclassical rational actor model of SEU(Subjective Expected Utility)theory ,and its extension in the form of the Tversky-Kahneman Prospect (Cumulative Prospect Theory )Theory that underlies the behavioral economics(finance)school of thought that has arisen since the late 1970's.Shiller makes it clear that he is an avid supporter of this school(Shiller,pp.117-120).SEU theory is actually a more advanced mathematical form of Jeremy Bentham's Benthamite Utilitarianism as expressed in his 1787 book, " Introduction to the Principles of Morals and Legislation".Bentham,the founder of neoclassical economics, asserted that all rational,and even some irrational, human decision makers are able to accurately calculate the outcomes of their actions .However,he failed to present a method describing how such decisions were made.Modern neoclassical economics filled this gap by combining the Ramsey-de Finetti-Savage subjective theory of probability,based on the premise that all probability calculations are precise,exact,accurate,unique,linear,additive,single number calculations, based on the laws of addition and multiplication of the probability calculus,combined with the expected utility theory of Morgenstern and Von Neumann,which claimed that all utilities can be shown to also be exact,precise,linear,additive calculations.Neoclassical economist Herbert Gintis gives a good summary of the neoclassical SEU theory :"...the model can be shown to apply over any domain in which the agent has transitive preferences " so that " there is a probability pi subscript,0<=pi subscript<=1 such that the agent is indifferent between Ai subscript and a lottery that pays A1 subscript with probability pi subscript and pays An subscript with probability 1-pi subscript.Clearly,these assumptions are extremely plausible "(Gintis,2004,Politics,Philosophy,and Economics,3,p.40).Unfortunarely,this is not the case.Gintis has presented an abbreviated summary of Savage's sure thing postulate that both Keynes(A Treatise on Probability,1921,p.315,ft.2)and Ellsberg showed required that the decision maker would have to be able to specify a complete information set.Keynes expressed this by the condition that the weight of the evidence,w,equalled 1.Elleberg expressed it by the condition that there was no ambiguity in the information base so that his variable ,rho,equalled 1.This problem ,of a complete information base ,showed
up in the 20 year exchange between L Jonathan Cohen and Tversky and Kahneman ,carried out primarily in
the pages of Brain and Behavioral Science journal between 1974 and 1994,over the blue -green taxi cab problem.Tversky had to come up with a Keynes/Ellsberg like w or rho variable,that he called s for support,in order to counter Cohen's point that Tversky -Kahneman were claiming that the experimental subjects
were irrationally using heuristics and rules of thumb,instead of the mathematical laws that a rational decision maker would use,rather than recognizing that the experimental subjects realized that their information
base was incomplete so that w and rho were less than 1.In cases of uncertainty /ambiguity ,where w or rho are less than 1,it is irrational to attempt to use the mathematical laws of probability which only work if w or rho are equal to 1.This was exactly the same point made in 1931 by Keynes in his review of Ramsey's subjective theory of probability.SEU theory can only deal with situations of
risk(w=1,rho=1).
Shiller constantly refers to the need to better manage risk through the risk models used by modern financial mathematical models .These risk models all result in the use of some sort of normal probability distribution(joint normal,cumulative normal,bivariate normal,multivariate normal,log normal).Benoit Mandelbrot
has demonstrated continuously for 50 years that none of the time series data supports the use of any type of normal distribution.The data supports the use of the Cauchy,Frechet,or power law distributions like the Pareto.Mandelbrot has correctly demonstrated that decision makers face the wild risk of the Cauchy and not the mild risk of the Normal.All 6 of the solutions proposed by Shiller on p.122 and discussed in depth on pp.123-169 can't deal with Keynesian uncertainty or Ellsbergian ambiguity or Mandelbrotian wild risk.The only way to deal with the uncertainty and lack of confidence created by the speculative and securitization behavior of the large Wall Street investment banks and the commercial banking system is a preventitive one-Prevent the speculators from getting their hands on the bank loans that they need to leverage their debt position in the first place.Thisis the solution arrived at by both Keynes and Smith(See Smith,WN,1776,pp.339-340;Keynes,GT,1936,pp.321-327,338-353,and pp,374-377).There is only one reference to uncertainty in this book.Shiller puts uncertainty in italics on p.103:"Right after the 1929 crash,the forecasters,although they did not predict the depression that was to follow,expressed unusual uncertainty(uncertainty is in italics for emphasis)about the economic outlook.Romer believes that it was this uncertainty that led to the sharp contraction in consumer spending that ultimately caused the Depression ".(Shiller,p.103,2008).Unfortunately,none of his solutions,based on the standard neoclassical SEU
risk models,that are taught universally in all economics and finance classes where Shiller teaches,can deal with the collapse in investor and consumer confidence because confidence
is a function of Keynes's w,which is assumed to always equal 1 in the SEU theory.Keynes gave the correct solution on p.158 of the General Theory-"A collapse in the price of equities,which has had disasterous reactions on the marginal efficiency of capital may have been due to the weakening either of speculative confidence or of the stste of credit.But wheras the weakening of either is enough to cause a collapse,recovery requires the revival of both(Keynes placed " both"in italics for emphasis).For whilst the weakening of credit is sufficient to bring about collapse,its strenthening,though a necessary condition of recovery,is not a sufficient condition."(Keynes,p.158,1936).None of Bernanke's current policies or of Shiller's recommendations on risk management will have any impact on confidence.

Shiller's position,in this book and the others he has written,is that the problem is one of irrational exuberance combined with information cascades.
"An information cascade occurs when those in a group disregard their own independently,individually collected information because they feel thateveryone else simply couldn't be wrong.(Shiller,p.47).Keynes had already shown that the reason this occurs is that each individual regards his w to be very low.This means that you are now dealing with uncertainty and not risk.Risk management techniques,no matter how mathematically advanced,will not be able to deal with this problem.

Shiller has correctly identified the problems of financial speculation and securitization.Unfortunately,his new risk management techniques would have no more of a chance of dealing with the wild risk of the Cauchy Distribution than an ice cube would have of not melting in the Sahara Desert.An ounce of Keynesian/Smithian prevention is worth more than a pound of risk management techniques build on the standard deviation of a normal probability distribution." Excessive Volatility " automatically means you have to deal with uncertainty as opposed to risk.










Books

(display all categories)

Adjust text size.