From here to eternity

Buttonwood

May 31st 2005
From The Economist Global Agenda

In pushing ultra-long bonds, governments are looking after their own interests more than those of pensioners

SOME years ago, Buttonwood was going through an old trunk and came across a letter from her great uncle. With it were enclosed some bonds, and the hope that when your correspondent came of age “the dollar would be worth something more like 100 cents on the dollar than it is today”. The coming of age has come and gone, and the dollar buys considerably less than it did. The bonds, however, made a modest contribution to college expenses.

Might Buttonwood have enjoyed a few more under-age whisky sours if Great Uncle Alfred had slipped in General Motors shares instead? Are bonds definitely the way to go, where long-term liabilities are concerned? As populations age, and financing their ever-longer sunset years assumes crisis proportions, a lot of governments clearly think so.

Last week, the British government sold £2.5 billion ($4.6 billion) of 50-year bonds, the longest debt it had issued since 1960. This followed the French, who launched their own 50-year euro-denominated bond in February. Germany is talking of following suit (and Telecom Italia has issued 50-year paper, to mixed reviews). Across the Atlantic, the Americans said earlier this month that they might bring back the 30-year bonds they got rid of in 2001.

At first sight, it seems strange that long bonds should be seen as the wave of the future just as the bond bull market looks like becoming a thing of the past. But it is easy to figure out why governments like them. They want to lock in current low borrowing rates for as long as possible.
They have other reasons, too. In America, for example, higher budget deficits have coincided with aggressively shorter public debt maturities, so more and more debt has to be rolled over, exposing the Treasury to ever-greater refinancing risk. Long bonds could also help to pave the way for a partial privatisation of Social Security, a cherished item on President Bush’s agenda. And it is possible that Alan Greenspan thought he saw a way to solve the low-interest-rate conundrum and talked the Treasury into it.

Ultra-long bonds are also useful for pension funds and life-insurance companies, which have payments to retired folk and annuity-holders stretching out for decades in front of them. The stockmarket collapse in 2000 scared many money managers away from shares, even though they traditionally return more over the long haul than other assets. And countries where retirement schemes rely heavily on government bonds have begun to feel the need for longer versions of them.

Accountants and regulators are telling providential institutions to match their assets more closely to their liabilities. Long-term bonds seem to do the trick better than shares, because the value of assets and liabilities broadly move together. As interest rates rise, the value of lower-yielding existing bonds in their portfolio falls—but so too does the present value (discounted by bond yields) of their future liabilities. And vice versa. The appetite of investing institutions for what long-dated paper exists is one reason why long-term interest rates have stayed stubbornly low.

A pig in a poke?

From the pensioner’s point of view, though, does it really make sense to buy an asset near what looks like the top of the market? Obviously, as people get older they grow keener on investments that produce a reliable stream of income (eg, bonds from creditworthy issuers) than on investments that take years to produce capital growth (eg, most shares). But are ultra-long bonds the right ones to buy?

There are broadly three issues: income, vulnerability to a rise in interest rates (which cuts the value of lower-yielding bonds) and exposure to erosion by inflation (which reduces the value of both income and capital). Long bonds usually have higher coupons than short ones, as investors demand a premium for exposure to the risk of interest-rate rises and inflation.

Last year, Harry Kat, a professor at City University’s Cass Business School in London, and three others used American stock and bond data from 1970 to 2003 in thousands of simulations measuring risk and return on ten-year Treasury bonds, alone and in more diversified portfolios. They found that it made sense to buy long bonds in almost all interest-rate scenarios but one: when rates are low (so the risk that they will revert to the mean, thus undercutting the value of the bond, is high) and the yield curve—a line drawn through the interest rates on government securities with different maturities—is flat (so there is little extra reward for taking that risk).

But this, of course, is exactly the scenario that prevails today. Interest rates are at recent historic lows in America, Europe and Japan. The yield curve is flattening in America (the gap between the short-term federal funds rate and yields on ten-year Treasuries is only a smidgeon more than one percentage point) and in Britain it is inverted (long bonds yield less than shorter ones). So long-bond buyers in today’s market run the risk of purchasing something that is extremely likely to lose value while gaining nothing or next to nothing in yield pick-up. And if that analysis was correct for the ten-year Treasuries that Mr Kat and his colleagues looked at, it is even more true for 30- or 50-year bonds which may, in addition, be less liquid.

Buttonwood has been a fan of ultra-long bonds until now, warmed by the image of a generation of oldsters wandering off into the sunset clutching their perfectly tailored, income-generating, supercreditworthy bonds. She is now wondering whether debt-crazed governments are peddling a pig in a poke, and using the growing panic about pensions to do so. It is interesting that the British 50-year bond, though successful, attracted bids equal to only 1.6 times what was on offer, compared with demand equal to three times the total when France launched its ultra-long bond earlier (though different circumstances explain some of that gap).

On a cheerier note, though, Stephen King, chief economist at HSBC, questions the conventional wisdom that bond yields must head back up. He believes that worldwide structural changes (increased mobility of capital and labour, the greater credibility of central bankers as inflation-fighters, the preference of ageing baby-boomers for fixed-income securities) have altered the equation, and that top-quality bond yields are, if anything, headed down. On his reckoning, the yield on ten-year Treasuries is likely to fall from today’s 4.1% to around 3.5% in 2006. Now that would represent a triumph for bonds which would warm the cockles of Great Uncle Alfred’s heart.

Send comments on this article to Buttonwood (Please state whether you are happy for your comments to be published)

Read more Buttonwood columns at www.economist.com/buttonwood

Copyright © 2005 The Economist Newspaper and The Economist Group. All rights reserved.