July 9 , 2003
The word "bubble" is used so much that I thought
I would discuss what economists call a bubble and why it is such an economic
Bubbles are inconsistent with efficient markets or with
economic equilibrium. Indeed, half
the finance schools in this country teach courses that do not allow students to
understand how a bubble develops and why it is dangerous.
Nevertheless, bubbles do exist and they are dangerous
largely because they are inconsistent with optimizing economic behavior.
The South Sea bubble had investors forgoing their own land to develop
enough purchasing power to invest in the "new opportunities."
During the tulip mania, investors borrowed enormous sums to buy a single
Bubbles are not rapid price increases for assets, although
all bubbles show such increases before they collapse. Sometimes, the rapid price gains are justified by economic
conditions. For example,
exceptionally low interest rates mean that the value of enterprises (which earn
income streams well into the future) are worth more even if the income streams
are not improved.
Housing prices should rise when mortgage rates fall.
Indeed, the failure to see the value of housing increase as the cost of
financing is declining probably
would be a violation of equilibrium economic conditions. (I cannot say
precisely, because other things could be happening.)
So, if we cannot look at rapid price increases to identify
a bubble, what can we use?
Probably, the best clue is that we begin to use alternative
yardsticks to justify asset prices. During
the height of the technology bubble (and that was a bubble), analysts were
talking about ratios to sales to determine enterprise value instead of earnings
projections. Any change in the
yardstick should raise a caution flag about the value of what is being measured.
Bubbles also attract financial assets.
Land prices actually fell in Britain while they were booming in the South
Seas. Tulip mania was financed by a
surge of debt. If anyone cared to look, many stocks failed to rise while
technology stocks were booming. When
relative asset values change dramatically that could be a clue.
Lately, I have been using changes from normal behavior to
determine if a bubble is present. In
single family housing, household
holdings are about the same significance to all assets that households held for
the past quarter century, at least in the U.S.
By contrast, equity holdings never constituted more than 25 percent of
household wealth until 1998. They
then surged to 50 percent before the bubble burst.
Now, they are back down to slightly more than 25 percent.
Hyman Minsky used to blame "excessive gearing"
for bubbles. By this he meant that
liquidity flowed to the hot hand in investing.
In other words, the winners got more resources with which to invest while
the losers got less. Naturally, the
psychology of the winners dominated the game, just as the winner in the weekly
poker game tends to increase the size of the bets.
If the monetary system is not producing more liquidity, the
relative values of assets begin to widen until obvious values are exposed in the
less favored investments. However,
without going into all the details, Minsky and others argued that most monetary
systems would expand their liquidity to accommodate the hot hand.
As a result, excess value distortions periodically surface.
Thus, is there too much liquidity in the economy, or is the
liquidity being concentrated into to few varieties of assets?
Money growth currently is sufficient to support asset
bubbles, as prices are growing almost 6 percentage points less than the growth
of most measures of money. Therefore,
one condition for developing bubbles is present. (Current money growth is slightly less after inflation than
during the technology bubble, but only slightly).
However, unlike the technology bubble, most asset values
are growing appropriately relative to prevailing economic conditions.
Enterprises are not priced too high for prevailing interest rates,
although they are not cheap either. Housing
is appropriately priced for prevailing mortgage rates, although some excesses
have developed in California and the Mid-Atlantic states.
What has not been priced correctly are bonds.
After avoiding bonds in 1999-2001, investors have embraced them in recent
months. As I have mentioned in
earlier columns, riskless overnight money, such as the federal funds rate or
treasury bills, should yield enough to compensate for current inflation and the
cost of saving instead of spending resources.
With prevailing inflation at slightly more than 1.25 percent and a presumed time preference of 1.5 percent per year, this would mean a yield of 2.75 percent. Yet, such rates currently are 1 percent or less. In another column, I will discuss why this discrepancy is causing distortions in mortgage rates and corporate borrowing rates that have caused some analysts to incorrectly declare that bubbles exist in housing and equity values.