WASHINGTON, Dec. 18 (UPI) -- An American Enterprise Institute conference Wednesday, that looked at the "wealth effect" on consumer spending of stock market movements, had dire implications for future consumption.
It showed that a stock market boom or slump has a substantial effect on consumption, but that the effect is lagged, with more than half of it coming after the first year. Thus, a lengthy market decline has an increasingly negative effect as it continues.
The keynote speaker at the conference was Michael Palumbo, a Federal Reserve Board economist. Palumbo has examined the time series of consumption and wealth for the U.S. economy since 1960, and concluded that a change in wealth changes consumption by between 3 and 5 percent, with most recent data pointing to a factor of about 4 percent.
This effect on consumption takes place primarily over the three years following the change in wealth, with about half the change in consumption coming in the first year, and another third in the second year. Palumbo has constructed a model using these parameters, which tracks closely the increase in U.S. consumption's growth rate in 1995-2000, its decline thereafter, and the fall in the savings rate from 6 percent of personal income in 1995 to 1 percent in 2000, and its subsequent partial recovery.
The existence of the effect is confirmed by a further study by Fed economists Karen Dynan and Dean Maki, which showed that consumption increased and savings declined in 1995-2000 in stock-owning households, but not in those that did not own stocks.
The average savings rate among the wealthiest 20 percent of the population declined from 7.5 percent to 1 percent in 1995-2000, while that of the remaining 80 percent of the population remained approximately constant at around 2 percent.
Thus, consumption's rate of growth was around 0.5 percent above the long-term trend of 3 percent per annum in 1995-96, 1 percent above that trend in 1997-99, on trend in 2000, and 1 percent below the trend in 2001-02. The $12 trillion increase in the value of common stocks in 1995-2000 could be expected to have increased consumption by about $480 billion above trend.
Household wealth (including stocks, houses and other assets), which had amounted to between 4.5 and 5 times disposable income in every year from 1960-1994, rose to 6 times disposable income in 2000, and has since declined to 4.9 times disposable income.
There was some discussion about what the expected drop in consumption should be from the 2001-02 drop in stock prices. Although common stocks have dropped in value by $7.5 billion, leading one by the model to expect a drop in consumption, eventually, of $300 million, this has been offset by an increase in house prices, raising house values about $4 trillion above their long-term trend, which would lead to an estimate of a net $3.5 trillion drop in personal wealth, and thus a $140 billion drop in consumption below trend.
Either way, there is still a significant portion of the negative "wealth effect" from past stock price drops still to come, not to speak of any effect from possible future drops in both stock and housing prices.
Kevin Hassett, of AEI, expressed skepticism about the existence of the wealth effect, as he had when a Fed economist from 1993. His model also demonstrates a long-run correlation over the past 50 years of about 4.3 percent between consumption and wealth, but he pointed out that the "fit" of the curve is poor; it consists of a lengthy series of quarterly points above the trend line, followed by a lengthy series below the trend line. If this is corrected for, the correlation drops to 1 percent or less.
There was considerable questioning from both the floor and the panel as to whether the "correction" was appropriate, however.
Chris Carroll, of Johns Hopkins University, the third speaker, also a Fed economist in the middle 1990s, looked at the implications of the wealth effect for policy, and concluded that they were limited in the short term, because the wealth effect on consumption took place only over 3 years or so. Thus a short-term up-tick in consumption produced by a stock market bubble could not be significantly affected by monetary tightening. He also examined the current high level of consumer debt, but felt that it should be of little concern because the increase in debt was largely concentrated among higher income consumers.
Carroll confessed to being a recent convert to the reality of the wealth effect, largely on the basis of the Dynan/Maki study; as he said, "When I was at the Fed I was as skeptical as Kevin (Hassett), and so was everybody else."
There are a number of implications of this research. First, it is pretty clear that Fed Chairman Alan Greenspan's failure to tighten money at the time of his "irrational exuberance" speech in December 1996 was due to Fed economists, in defiance of standard economic theory, at that time believing there was no wealth effect.
I can confirm this from personal experience; in May 1996, I attended a presentation given by New York Fed Gov. Peter McCullough, at which in response to my question, he confirmed to my astonishment that the Fed did not at that time believe in the wealth effect's existence.
Naturally, if the wealth effect did not exist, then there was no need, in spite of the stock market being "irrationally exuberant" for Greenspan to tighten money in December 1996 to head off a consumer spending binge.
Of course, the extraordinary 1995-99 bull market had in December 1996 been in full swing for more than two years, and had carried the Dow to 6,400, more than twice its early '90s level, so there had already been plenty of time for a consumer spending bubble to take hold.
Rapid action at that time, which could have deflated the bull market at around Dow 6,400, would have prevented a huge bubble followed by the collapse of which, at Dow 8,447 Wednesday, we have still by no means seen the bottom.
John Maynard Keynes wrote that practical men were all slaves to "defunct economists." The U.S. economy since December 1996 has been held hostage to an appalling monetary error, produced by the misconceptions of economists very much alive -- indeed, in 1996 almost in British Chancellor of the Exchequer Nigel Lawson's famous category of "teenage scribblers."
The other implication of this analysis is for the years ahead, and it is a grim one. There are a number of factors tending to depress consumer spending in the near future, some of which have not been present in the last couple of years, which together suggest that consumption may drop very substantially in 2003-04, with of course a huge negative impact on the economy as a whole.
First, if only half the wealth effect of a stock market drop is felt in the first year, and the rest thereafter, then there is a great deal of concealed deflation in the stock market drop we have already had. The Wilshire 5000 Index, a proxy for the entire U.S. stock markets, dropped 1,470 points (equivalent to about $1.7 trillion in value) in 2000, and 1,319 points ($1.5 trillion) in 2001, but has so far (to Wednesday's close) dropped 2,276 points ($2.6 trillion) in 2002.
If you assume a 4 percent wealth effect coefficient, and the wealth effect to occur over the three years following the price drop, then the wealth effect on consumption in 2000 was somewhat positive (because of the market rises in 1998-99), in 2001 it was roughly negative $32 billion, and in 2002 negative $62 billion. In 2003, even assuming the market remains flat, it will be negative $64 billion, with another negative $40 billion to come in 2004 and thereafter.
However, if as Palumbo suggests, about half the wealth effect in 2001-02 was masked by rising house prices, the deflationary drag will be more pronounced going forward.
Second, the argument that the high level of consumer debt does not matter because it is concentrated on the wealthy overlooks the fact that these are the people whose stock portfolios have been decimated. Their leverage has thus shot up, in many cases far beyond what they can comfortably carry, and the sharp rise in consumer credit delinquencies in 2002 suggests that many of them are "tapped out" on their borrowing limits. Of course, this has further negative implications for their spending going forward.
Third, short-term interest rates are close to as low as they can go, and long-term rates also appear to have stopped declining. This has resulted in the beginnings of a fallback in the unprecedented boom of mortgage re-financing in 2001-02, in which $1.2 trillion of mortgages were refinanced in 2001 and a further estimated $1.25 trillion in 2002 -- a total of approximately 40 percent of all mortgage debt outstanding.
According to a study published Wednesday by Economy.com and Homeownership Alliance, approximately one-fifth of the economic growth since late 2000 is directly attributable to this refinancing boom, with $170 billion in cash raised in 2002 from cash-out re-financing. Of course, this has contributed mightily to the relatively healthy trend in consumer spending in 2001-02 -- according to an estimate by Fannie Mae over half this released cash, $85 billion in 2002, is being used to finance more spending.
This boost to the economy, which in 2001-02 has masked the deflationary effect of the stock market decline, will do so to a much lesser extent going forward.
Finally, there is the possibility of a further stock market decline. According to traditional stock market valuation metrics, an appropriate value for the Wilshire 5000 Index is in the region of 5,000 giving an additional 3,431 points (equivalent to $3.9 trillion) of downside potential even before the possibility of "overshoot" is considered. In addition, there must be some possibility that the housing bubble that undoubtedly exists in the major East and West Coast centers will start to deflate. Both these factors would greatly worsen the deflationary effect of declining wealth on already weakening consumer spending.
In summary, expect consumer spending in 2003-2004 to be very much weaker than in past years, or than is currently forecast. Far from being an "engine" of the economy, as it has been in 2001-02, it will become a major drag on it, causing a second recessionary "dip" much longer and deeper than the first.
