mike777, on 2013-July-19, 00:38, said:
Richard Roll discusses the issue of markets.
Let's examine that Roll paper that you keep citing, because I don't find it particularly convincing:
First and foremost, lets start with his explanation why a bubble isn't the correct explanation for the collapse in the housing market.
Ultimately, his argument collapses down to the following:
Quote
Of course, one can never completely rule out the possibility of an irrational bubble in hindsight.
A simple explanation for the 2007–08 debacle is that global real estate values were simply too high and
that a crash was inevitable. Why there was a bubble in the first place and why it burst when it did might
forever remain unanswerable questions, which is the problem with any bubble story; it can always be concocted to “explain” everything.
This is a fancy way of saying "I don't like bubbles, so let me come up with another explanation"
Here's what Rolls offers instead:
Quote
Consider an alternative rational explanation: a 2007 reduction in the value of human capital, the overwhelmingly dominant component of collective total real wealth. Human capital is also the
most important determinant of real estate values. People will pay what they can afford for housing. There are plausible reasons to conjecture that a negative human capital valuation shock actually
occurred in 2007, although hard evidence is lacking because human capital prices are not directly observable.
Machinery, equipment, and intangible values can be observed daily, at least in part, through equity market values. Real estate values are partially observable, although less frequently—at
best monthly. Human capital value is not observable until long afterward, when labor income is ultimately reported.
Given the impossibility of observing human capital values directly, the diagnostician must resort to supposition. The following is a
plausible chronology of recent events, based on rationality and well-functioning markets. First, human capital values declined precipitously
from mid-2007 through 2008 because anticipated growth rates in labor income declined. This value reduction was not observed
(and could not have been). If the anticipated growth rate in labor income is relatively close to the discount rate, even a small decrease
in anticipated growth can have a large impact on the present value of human capital.
Note that his explanation is subject to the precise same critique as the one for the "real estate bubble:.
Roll directly states that human capital is unobservable and offers no explanation why expectations about growth rates for labor income suddenly changed in 2007.
Unlike the bubble explanation, Rolls alternative requires that there was a sudden and significant change in growth rate expectations that suddenly impacted the world economy in a very short period of time.
What was the driver for this event? Why did it happen all at once, rather than a slow, gradual change in beliefs?
Next, lets look at some of the assumptions that Roll requires for his preferred explanation. In particular, consider the following quote:
Quote
Suppose Warren Buffett gave Bill Gates $20 billion. What is the net change in aggregated (total)
wealth? Would markets crash as a result? Would total human capital be worth less? Would real
estate values fall or rise (except in Omaha, Nebraska, relative to Seattle)? Would there be any
direct impact on existing stocks of productive machinery, goodwill, or any other real asset? No,
there would be little, if any, change in total wealth, output, or any other indicator of real economic
activity as a result of a simple wealth transfer. This fact has important implications for the role of debt
in economic crises.
Note that Roll concedes that this wealth transfer would impact real estate prices in individual sectors of the economy.
(In this case Omaha and Seattle). The consider what happens when housing market collapsed:
There was a significant net transfer of wealth from consumers and individuals who held the wrong end of CDOs to a small group of speculators who were holding the "Right" end of the CDOs
Historically, the primary vehicle for savings for American consumer was their homes.
So, rather than thinking about 20 billion transferred from Buffet to Gates, think about a whole lot more transferred out of the consumer economy.
Help
In principle one can never perfectly apply the CAPM because the variability of total returns for the totality of all asset classes in unknowable. This problem has been obvious for God knows how long. Never stopped anybody from using it in its imperfect form.