Richard RollThe annual meeting of the IAFE in New York on 2010/06/18 featured UCLA's Prof. Richard Roll , who was a colleague of Chair at the University of Chicago many years ago. He started out by warning the audience that his explanation is different from everybody else's, cannot really be considered proven, and may be hard to accept. Nevertheless he urged the audience to keep an open mind, if only because if this explanation is correct then the current remedies may actually be harmful.

First he dismissed the popular idea that inappropriately low interest rates caused a bubble in real estate prices, which then crashed. Although nominal interest rates were low, the more relevant real interest rates (as shown by the yield on TIPS) were actually rising during the period 2004 to 2007.

Also, defaults in the debt or derivatives markets cannot have been at the root of the crisis, contrary to common opinion. The net amount of debt in the economy is zero (someone owns each debt and someone else owes it) as is the net amount of derivatives (for every long there is a short). Default on debt simply involves a redistribution of wealth, not a destruction of wealth. For example if a borrower defaults on a $300,000 mortgage, and the house is now worth 200,000, the result is essentially as if the bank had given a $100,000 "gift" to the borrower. One is better off and the other is worse off, but the net national wealth is unchanged. These are just redistributions with no (or little) system wide effect.

So what happened? Roll believes that the root of the crisis was a reduction in wealth, and specifically a drop in the value of human capital. Recall that human capital is the present value of all future labor income streams for all persons. It is very difficult to measure because it requires knowledge (or estimates) of the future; but it must be a very large number, perhaps the largest component of national wealth. Roll believes that the value of human capital is correlated with the value of real estate and, to a lesser extent to the value of the stock market. The correlation can be seen, for example, in the fact that people who expect to have a high income in the future live in expensive houses; the value of someone's house is to some extent an estimate of that person's future income.

According to Roll a sharp drop in the value of human capital took place in 2007-2008. This immediately, or perhaps with a short lag, caused a drop in the value of real estate and (to a lesser extent) a drop in stocks (because if the people's future income is expected to be lower, the revenues of corporations will also be lower). We cannot measure the drop in human capital directly, but the drop in real estate and stocks is a clue that (according to Roll) the value of human capital dropped.

The only remaining question is why the value of human capital fell. Roll's controversial explanation is that the market correctly anticipated that government intervention would greatly increase in the years ahead, and that his would cause a permanent lowering of the rate of growth of labor income. Economists have found that past a certain point, a decrease in the share of GDP generated in the private sector leads to lower growth; conversely "liberalization" or an increase in the private share typically leads to higher growth.

This explanation was contested by a member of the audience, who said he had worked in the mortgage securities field and who felt that enormous problems developed in the mortgage market which the Professor was leaving out and which were essential to understaning the crisis.

Another member of the audience pointed out that the professor's explanation is similar to the theory of Amity Shlaes as to why the Great Depression lasted a long time.

Another critique was made by derivatives textbook author John Hull, who felt that the human capital explanation was on the right track, but disagreed about the cause. He felt that the markets began to realize that the US was increasingly unable to compete with China and could not easily restructure itself because of weaknesses in education and skills of the US population. Roll replied that this explanation was too specific to the Us, and did not account for the fact that other countries, for ex. Great Britain, also experienced a severe financial crisis.

Steve Ellison writes:

One possible reason for a decline in human capital is the aging of the population. As the average age of the population increases, the value of future income decreases.

Rocky Humbert writes:

Roll says, "Default on debt simply involves a redistribution of wealth, not a destruction of wealth. And that wide-spread defaults have no system-wide effects."

His argument is like saying "Muggers and bank robbers simply involve a redistribution of wealth, not a destruction of wealth." He ignores the costs and consequences that wide-spread mugging and bank robbing would have on behavior and economic activity. He also ignores that bankruptcy and reorganization imposes significant costs on all of the stakeholders (and by extension, society as a whole.) That's one reason why the value of an enterprise declines as it enters bankruptcy protection. Defaults is not a zero sum game with the value moving from shareholders to creditors. It's a negative sum game.

Professor Roll's argument falls down when one considers that "human capital" is a balance sheet item, but "human income" is on the cashflow statement. A country, company and individual with a negative net worth (negative human capital) can function without any problems — but it's when the cash flow cannot support the expenses that the problem causes a crisis. Hence, human capital is like goodwill on a corporate balance sheet. It's an accounting fiction. It's the human cash flow that matters.

Gibbons Burke writes:

Abortion and contraception have taken a heavy toll on human capital. Since Roe v. Wade was decided, over 50 million potentially productive human beings have been murdered in the womb in the U.S. alone.

Kim Zussman adds:

You wrote "the drop in real estate and stocks is a clue that (according to Roll) the value of human capital dropped."

Then it must have been true that human capital (anomalously) increased, causing the housing bubble in the first place.

In the attached chart (Case-Shiller real house-price data, 1890-2010), the bubble peaking in 2006 DWARFS all other housing price peaks over 120 years.

Perhaps a surreptitious rally in human capital occurred, manifested by the unprecedented housing bubble?

Rudolf  Hauser comments:

It is not quite fair to criticize Prof. Richard Roll without having heard his presentation, but based on Alex's summary thereof, I will do so nonetheless. I agree with some of what he has to say but differ in many respects. The main failure is to make any reference to the discount rate. Wealth may be the present value of future income but that is both a function of those future income streams and the rate by which they are discounted. I also find the focus on the value of human capital a strange form of analysis which ignores some market realities as to what actually happened.

Let me start by a very simplified explanation. Real wealth in the capital stock is created when someone labors to produce it. That includes amounts spend in developing human capital via education and training. That labor means that consumption will be less than the amount of production by the amount of that capital investment. But that tells us nothing about how existing wealth is valued. Assume A is moving from NY to LA and B is moving from LA to NY. Both purchased their homes for $100,000. They now decide that their homes are worth $500,000. They trade homes. All they have is capital transfer. It hardly matters what the actual value is in that it is an even exchange. Now in reality A will sell his home for that amount to someone, just a B will, and they will both buy their respective homes from others. The ignoring the intermediate transactions, that is in effect what you have. Now in the past people and lenders would base their decisions largely on their expected future incomes. But in the sort of bubble situation we had in housing, people were expecting home price appreciation to bail them out. In essence, expected future appreciation was part of their anticipated income stream. Now when it was finally realized that this assumption was a chimera that it was decided that the two homes were only worth $300,000. Now wealth has declined in value. The only question is who bears the loss. Given the mortgage amounts the lenders might well find that they must bear some of that loss. There is no reason to refer to "gifts" to the borrower. It is the decline in the value of the property, of wealth that causes the loss. It makes no sense to call that loss a gift.

Then we get to what brings about the additional losses of real wealth. That happens when capital, either physical or human, becomes useless in producing future income. That can happen when lenders refuse to renew loans and/or revenue declines cause bankruptcies. Long periods of unemployment destroy human capital. Physical capital decays from neglect. And that can happen because firms are driven into bankruptcy because of debt defaults and their ability to refinance themselves.

As to the argument that this was a drop in the value of human capital, it is first of all something that cannot be measured. The wealth we can measure has cyclical tendencies. While the recent recession was one that might logically influence future expectations, for the most part recessions are and should be expected as they have always happened from time to time. There is no reason why they should change long-term expectations other than emotions. What does happen is that there is a need in the time of crisis to have more liquidity. That increases the risk premium on longer-term and less liquid assets. Part of that increase in returns is an expectation of capital gains when the liquidity crisis/recession ends. That is why we often see an inverted yield curve leading into such declines. Logically, the yield curve should become more positive, not turn negative, because risk premiums should rise more on the longer term assets. But what matters is total return and that included anticipated capital gains. When those are not great enough the yield curve does not invert, as was the case in the 1930s. Well if the declines in wealth were due to changes in future expectations that is not what you would see. Rather it is because of what I would call risk liquidity premiums rising. If Roll's argument were valid with regard to government intervention, how does he explain the increase in stock values of the past two years-a period when by all logic the changes in government should be increasing fears of greater intervention?

What happened was that the markets finally realized that with all the complicated debt and derivate structures that depended on counterparties many transactions away to deliver was in doubt and that no one's balance sheets could be trusted anymore. With that lending dried up and all values where put into question. That cause a large increase in risk liquidity premiums that was only mitigated by the Fed belatedly pouring in large amounts of liquidity and the government offering guarantees for parts of the financial system.

Another point, although a bit more trivial. When Alex writes that Roll said that "The net amount of debt in the economy is zero." he ignores the fact that some of that debt is owed to foreigners. That is the statement is only true in an international sense.

What you had in this period was an increase in the foreign inflow of savings. Net fixed non-residential investment by business relative to GDP was significantly lower than it had been in the 1990s. In essence there was too much savings wanting to earn higher returns relative to the business investment opportunities leading investors to finance a housing boom instead.





Speak your mind

1 Comment so far

  1. Sam Humbert on June 22, 2010 2:35 pm

    The US stock market fell ~50% from the time the big-government party candidate — the leftmost of 100 US senators — pulled ahead in the betting markets till the time he took office. Fits with Prof. Roll’s thesis, at least to a first approximation.


Resources & Links