Recently, trying to forecast interest
rates has been a pretty tricky task.
And there is no reason to expect
that things will become easier soon. In a
global financial world, interest rates - and
this is particularly true of long-term rates
- have become dependent on the interplay
of so many different factors that it has
become increasingly difficult to make a
Even if you manage to understand the
nature of each of the forces at work, your
forecast will be right only if you also correctly
guess what the precise magnitude
of each of those forces will be.
Let us first look at the recent behaviour
of US 10-year rates. To the surprise of
many observers, since the middle of last
year they have been moving within a
fairly narrow range, despite the Fed
funds rate being constantly ratcheted up
at a ‘measured’ but systematic pace by the
US central bank.
The logic of the Federal Reserve’s
move was easy to grasp: having succeeded
in restarting the US economic
engine by lowering the whole yield curve
to levels not seen for decades, the Fed had
to cautiously change gear and slow the
economy down to cruising speed. But
the required deceleration was supposed
to be brought about by an upward move
of the whole yield curve, not just of its
short end. But this is what occurred.
Why? As frequently happens with puzzling
events, conjecture flourished. A
popular theme gave Asian central banks
the lead role. By investing the dollars they
were buying in treasuries they pushed
their price up. The problem with such
‘supply-versus-demand’ explanations is
that they are arbitrary. Why should the
$260bn (€217bn) increase in foreign
official demand for US bonds seen since
2001 have had more influence on rates
than the decrease in the net supply of
housing-related bonds of twice that
amount that also took place during the
same period?
The explanation rested on an assumption
about the maturity of the paper
those central banks were buying that
proved to be largely incorrect. New data
show that foreign official holdings of
treasuries have a much shorter maturity
than those held by private agents .
Chart 1 shows a more convincing
explanation. Since last year, the US bond
market has been consistently responding
to increases in oil
prices by a fall in
long-term interest
rates. The logic of
such a response is
clear: since a rise in
the oil price tends to
slow the economy,
the more steam that
is taken out of the US
engine by the oil
price, the less that
will have to be taken
out by the Fed.
Hence the low bond
rates. The problem is
that this has tied the
long-term rate to the
oil price, which is a
rather volatile variable.
In chart 1 the green
points are for the
period beginning
right after the February
2005 testimony of Fed chairman
Alan Greenspan (when he mentioned
the rate “conundrum”) and ending 11
May 2005. The red points are for the
period 10-21 June 2005 (after
Greenspan spoke about “froth” in housing
markets) and since the release of the
FOMC statement of June 29-30, 2005
(in congressional testifimony on 20 July,
Greenspan echoed its previous “signs of
froth” statement).
So does this mean that nothing can be
said on where US long-term rates are
headed without a clear view on what the
oil price will do? Not necessarily. The
bond market’s response has prevented
the economy from slowing significantly
by aggravating already severe financial
imbalances. To put it in a nutshell, US
households have been paying the extra
oil bill with borrowed money, and so has
the US economy as a whole.
The Fed knows this cannot go on forever.
Greenspan’s recent utterances
regularly pushed long-term rates higher
than implied by the oil price, but only
for a few weeks (as shown in chart 1).
To prevent a further deterioration of
these imbalances, the Fed now has little
choice. Even with elevated oil
prices, it will tighten until US households
borrow at a slower pace. The
more so since it is now clear that the
cost of the post-Hurricane Katrina
reconstruction effort is going to be
huge. A rate increase is necessary to
crowd out some spending that would
have been financed by private borrowing
and make room for spending
financed by insurance reimbursements
and government borrowing. Even
such a monetary tightening will not be
able to handle all the imbalances.
Household borrowing will be curbed
but unless the economy slows sharply
next year, the current account deficit
will stay wide, leaving the dollar at risk.
So one strong view can be held: wherever
the oil price goes, rates should be
higher next year.
However, this does not mean that
long-term rates should skyrocket.
Despite being low, bond rates are not
that far from their ‘normal’ level. During
the last decades, the 10-year Treasury
rate seems to have been anchored around
the average nominal growth of the US
economy over the past 10 years (chart 2).
Come 2006 , this anchor value will be
close to 5.4%.
What about euro rates? Well, here too
interdependencies have become tighter
and rate moves harder to predict. But if
US rates do go up by something like 100
basis points, past correlations make it
hard to believe that European rates could
be unaffected and that the 10-year Bund
yield remains at 3%, as at the end of the
summer. Chart 3 shows that, for European
long-term rates too, past average
nominal GDP growth gives a good
anchor. The anchor value right now is
slightly below 4%.
Can we think of forces that could prevent
euro bond yields from moving
closer to this value when US rates
increase? We can. First, changes in
accounting and prudential rules are forcing
various European financial institutions
to increase their demand of fixedincome
long-term bonds at a time where
issuance of such bonds is very scarce.
Since 2004, the issuance of bonds by the
private sector has been mainly at variable
rates (driven by the securitisation of
adjustable rate housing loans made in
Spain and Ireland).
Only governments issue fixed income
debt securities. But the amounts they can
issue have not increased much recently,
being strictly constrained by the EU’s
Stability and Growth Pact. If the euro is
not made weaker by internal disputes,
the ECB doesn’t have any reason to
tighten its rates soon and there are no
threatening inflationary pressures.
As Euro-zone growth picks up slightly
next year this force will progressively vanish,
leading to higher European bond
rates. But even more than in the US,
there is hardly any reason to fear a bond
market crash. This could only be triggered
by an inflation scare.
Given the present state of the European
economy and the vigilance of its central
bankers, this seems pretty unlikely. So, in
a world of uncertainties, there are still a
few things about interest rates that look
almost assured. At least, so it appears.
Anton Brender and Florence Pisani
are economists with Dexia AM in
Paris. They are joint authors of
‘America’s New Economy’ published
by Economica in 2005