| |
|
 |
The Vast Market for 30-Year Treasury Bonds (and Bond Futures) by Galen Burghardt and William Hoskins
Back in 2000, when budget surpluses were rolling in and we thought we were
flush as a nation, the U.S. Treasury began to reduce its long-term borrowings.
First the Treasury began buying back 30-year bonds, and then, in October
2001, decided to stop issuing long bonds altogether.
This was quite a shock to financial markets. Although many market participants
had already switched to the 10-year note as the benchmark instrument on the
Treasury curve, the 30-year bond still held a central place in the fixed income
market. In fact, the U.S. was the only capital market in the world with an active
30-year bond market, and that market still served as the basis for one of
Chicago's most active and successful futures contracts. So even though all
could see that the outstanding supply would continue for a number of years,
Treasury's decision to suspend issuance seemed to mark the end of an age.
As we now know, those rumors of the long bond's demise were greatly
exaggerated. We are back to running huge federal budget deficits, and the
Treasury needs to borrow. After several months of consultation with market
participants, the Treasury announced in August that it would resume issuing
30-year bonds twice a year, starting in February 2006.
The Treasury's decision was well received all around, and with good reason.
From a pure financing perspective, borrowing for 30 years at 4.5% makes a lot
of sense. Yields are low, and the yield curve is very flat. So not only can
taxpayers obtain low cost funding for a very long horizon, they don't have to
pay a premium to do so. In this vein, much of the favorable press focuses on
the wisdom behind tapping a new funding point on the curve and the broader
diversification of borrowing that will be available to the Treasury.
We think a more interesting story lies in the fact that the federal government
can do the private sector a potentially huge favor by issuing large quantities of
high quality long-term debt. This is because private pension plans may well
find it necessary to obtain better hedges for their liabilities. And if they do, they
will discover the outstanding supply of good hedging bonds is far too limited
for their purposes.
Huge Duration Mismatch
The story starts with the efforts by policymakers
in Washington to force pension
funds to more accurately report their assets
and liabilities. The White House is backing
a proposal to require pension funds to mark
their liabilities, and this proposal has been
incorporated in the Pension Security and
Transparency Act now under consideration
in the Senate. Accountants likely will follow
with their own rule changes to improve pension
security and transparency.
The idea behind these efforts is that disclosing
the true mark-to-market state of pension
plan funding will encourage corporations
to insure that their pension assets and liabilities
track one another more than they have
historically. If they do not match assets to liabilities,
then a host of interested parties—
shareholders, auditors, regulators, and even
plan participants—would see that funding
ratios (that is, the ratios of assets to liabilities)
often swing wildly as the present value of liabilities
goes in one direction while the value
of assets go another.
Given that a lot of the volatility in pension
funding comes from movements in 30-
year interest rates, and given the ultra-long
liabilities of pension funds, pension funds
face a large exposure to interest rate risk at
the long end of the curve. In theory, adding
a significant amount of 30-year bonds to
their portfolios would go a long way to managing
this risk. The reality, unfortunately, is
that the potential demand for long duration
would quickly dwarf the available supply of
30-year bonds.
While it complicates any discussion of
pension fund liabilities, it is important to
note that corporate pension liabilities fall into two types: those that are static—that is,
those whose nominal values are known
today and will not change with time—and those that are inflation sensitive. The inflation
sensitivity comes from the fact that
although standard corporate pensions are "fixed", the total liability is not entirely
known until retirement because it depends
on some form of final salary. Up until the day
of retirement, increases in salary from inflation
(as well as merit) will increase the final
fixed monthly pension.
The importance of the distinction is
shown in the attached graph in Exhibit 1,
which shows the present values of these cash
flows and their corresponding modified durations.
This graph was produced using data
reported for S&P 500 pension plans combined
with U.S. demographics and standard
mortality tables.
As the graph shows, roughly two-thirds of
the total liability is accounted for by the static
cash flows and has a duration of 13.4 years.
The other third stems from the inflation-sensitive
liabilities and has a duration of 22.5
years. The inflation-sensitive cash flows have
much longer duration because they come
from the ultra-long cash flows for younger
workers. These workers' final salary and pension
payment are highly sensitive to the level
of inflation over their working career.
Compare the value and duration of pension
fund liabilities with the value and duration
of the outstanding supply of investment
grade bonds, as also shown in Exhibit 1. This
covers not only Treasury securities but also
agency paper, corporate bonds, mortgagebacked
securities and various other types of
debt, and covers all maturities down to one
year. If corporate pension plans all rushed to
the market to find assets to match their liabilities,
they would come up short because of
the lack of good duration matches. There are
certainly lots of Treasury and other bonds
available, but they are not the right maturities
to match pension liabilities. Pension
plans with current active employees have
very long-dated obligations, so 30-year bonds
are the ideal asset to match the fixed portion
of liabilities. The inflation-sensitive part of
liabilities could be matched by the longest
Treasury Inflation-Protected Securities,
which are currently 20 years. The pension
needs for these long bonds are so large, however
that they dwarf the available supply of
long bonds and long TIPS.
In Exhibit 2 we take a closer look at this
problem. The table shows three possible pools
of bonds for hedging pension funds' static cash
flows. Of these, the best duration match
would be longer-term government securities
(shown as 20+ governments, which include
Treasury and agency bonds over 20-year
maturity). The next best would be 10+ government
and corporates. The worst would be
the third category, which includes all types
and maturities of investment grade bonds, as
represented in the Lehman Aggregate index.
With an average duration of less than five
years, the supply in this category would a poor
match for pension fund liabilities.
Based on our estimates of the likely
impact of the pending mark-to-market proposal,
pension plans would need $753 billion
of the 20+ government bonds to match the
duration exposure of their liabilities. This is
more than three times the available supply. If
the pension funds turned instead to the second
pool, the supply would be larger but still
well short of their needs, and the tracking
error widens. As for the third pool, there certainly
are plenty of investment grade bonds
to go around, but the tracking error between
assets and liabilities would be much, much
larger than the better hedges would provide.
To hedge the inflation-sensitive cash
flows, the best hedge would be provided by
longer term TIPS (shown as 10+ TIPS).
Plans would need $467 billion of these bonds
for hedging purposes, which is more than six
times the current outstanding supply.
The picture painted here actually tells
only part of the story. Corporate pension
plans represent only one group of hedgers.
The actual shortage of long bonds likely is
even worse because public pension plans, life
insurance companies, and aging baby
boomers (such as Galen Burghardt) need
long bonds as well (and especially long TIPS
for cost-of-living hedging).
A Land of Opportunity for the U.S. Treasury
The Treasury should view this long-bond
shortage as an opportunity to meet the legitimate
risk-management needs of its domestic
investors. To be sure, foreign central banks
have been the biggest customer for Treasury
bonds and notes, but they do not want 30-
year maturities. As every corporate treasurer
is aware, the best place for your securities is in
the hands of long-term investors who want
the risk and credit characteristics of the security. A distant second is to place them with
investors who are parking their dollars for the
short term. Pension funds are stable longterm
investors, while foreign central banks
may change their minds about Treasury bond
holdings at any time. Catering to your longterm
customers is always a good strategy
whether in business or in bond issuance.
We might note, too, that if the Treasury
fails to fill this gap, it could serve as a financing
windfall for other countries. Germany or
Japan, for example, could issue long-term
dollar denominated bonds and hedge the
proceeds back into their home currencies.
On the other hand, corporations cannot fill
the gap. As experience has taught us, no corporation
can promise high quality credit for
30 years. The mighty have fallen left and
right, so it is left to sovereign issuers that
have the full backing of their tax paying
publics to maintain their credit standing.
Implications for Treasury Futures
As Galen Burghardt and Terry Belton
chronicle in The Treasury Bond Basis, the
long-term Treasury bond contract was
eclipsed in recent years by the 10-year
Treasury note contract. Depending on how
you keep score, the 10-year contract passed
the bond contract either in 1999, when its
open interest exceeded that of the bond contract,
or in 2002, when its daily trading volume
passed that of the bond contract. (See
Exhibit 3.)
The reintroduction of the 30-year bond
will almost certainly breathe some new life
into the bond futures contract. Not that the
new issues will be especially relevant to the
deliverable set for some years to come, but
the presence of newly issued bonds will
almost certainly increase liquidity and trading
in the long end of the market.
In addition, if the proposals now under
consideration in the Senate become a reality,
or if other regulations force corporate pension
funds to take duration hedging more
seriously, there may be a surge in the use of
bond and note contracts for outright hedging
purposes. It is well worth noting that while
pension funds must cope with vague IRS regulations
on "unrelated business income tax"
to use leveraged cash bond positions as a way
of adding duration to their assets, they are
explicitly allowed to use swaps and futures.
That is, they can receive fixed on long-term
swaps, or they can go long bond futures.
Many pension plans also have very good
relationships with their current managers
and like their asset allocation mix. These
plans can simply add bond futures or swaps to
increase the duration of their assets to match
their liabilities.
This would be huge for the Treasury
futures market, although it seems unlikely
that it would restore the long-term Treasury
bond contract to its original dominant place
in the Treasury futures universe. Bond
futures may be useful for creating duration,
but as Magritte might have said, they are not
bonds. Those who short bond futures to pension
plan hedgers would need a place to
hedge. And until the Treasury has had a
chance to roll out a few hundred billion dollars
of the 30-year bond, the chief hedging
market will remain the market for Treasury
notes 10 years and under.
It is also worth pointing out that if pension
plans are required to more closely match
their assets and liabilities, it seems highly
likely that the resulting buying pressure will
flatten the Treasury yield curve for years to
come. The Treasury curve might well
become inverted, with yields on longer term
securities actually falling below shorter term
securities. That is what has happened in the
U.K., where hedging by pension plans is
required.
Exhibit 1
Source: Mellon Capital Management
|
Exhibit 2
Estimates: Mellon Capital Management
|
Exhibit 3
 <
Source: Futures Industry Association |
Galen Burghardt is senior vice president and
director of research for Calyon Financial in Chicago.
William Hoskins is director of fixed income
research for Mellon Capital Management in San
Francisco. As colleagues at Dean Witter Institutional
Futures, they wrote widely on Treasury and Eurodollar
futures, including the seminal The Convexity Bias in
Eurodollar Futures. |
|
