Bonded Servitude
Feb. 28th, 2008 12:22 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
peterbirksIn the midst of the various crises in the bond market of the past nine months or so, journalists have had little time for an event which has been gradual and has no immediate obvious impact. This is the return of the value of time.
In January 2007 a 10-year US government bond earned only a fraction more than a 2-year bond. This ostensibly insane state of affairs had persisted for some time -- indeed, for a sufficient length of time for it to be seen as less an aberration and more a new scheme of things.
In the UK, we had actually reached a situation where for short periods of time you got paid less if you tied up your assets for 10 years than if you did so for two years -- the so-called "inverted yield curve".
The reasons for this lunacy were two-fold. One reason was that, in a liquid market, a 10-year bond isn't "really" a 10-year bond, because you can sell it after two years. If the outlook on interest rates is such that you think you can make more by selling on your 10-year bond after two years than you can by holding the two-year bond to maturity, then there is a rational reason behind paying more for the 10-year bond. Long bonds, in other words, benefited from the lax attitude to risk that has been prevalent for quite a few years (see the "lend anything to anyone who wants to borrow" mantra of the same period).
One possible risk is that, while the two-year bond is redeemed by the issuer, the 10-year bond, if traded in at the same time, needs to find a buyer. In the first case there is no risk of a market collapse, only that the redeemer will fail to honour its obligation. But, in a liquidity crunch, there might not be anyone with any money to hand. If you need cash and there are no buyers for your 10-year T-bonds, you can't sell them back to the government. It doesn't work like that. Until recently, a number of traders seemed to think that it did.
The second reason was a piece of accounting madness which, in simple terms, demanded that pension funds and the like match the time aspect of liabilities to the time aspect of assets. What this moronic accounting rule did not allow for was that the supply of two-year assets was far greater than the demand for two-year liabilities, whereas the supply of 30-year assets was far lower than the demand for 30-year liabilities.
That, indeed, was one reason why the whole trading system existed.
So, the upshot of this stupid law was that pension funds had to pay far more than the "fundamental" value of long-term bonds, while short-term bonds were, relatively, rather cheap. Net result, the inverted yield curve.
This aberration also led to distorted expectations amongst savers such as me. We got used to an Instant Access account offering as much as a 10-year endowment bond. Why lock your money up in the latter when you can earn as much in the former, and have immediate access if you want it?
But, it's all change. The two-year US bond has fallen nearly three points (devastating the return on my Schwab account, btw) while the 10-year bond has fallen by just a single percentage point. Investors are finally being rewarded for being willing to tie up their money for a longer time period. Time, once again, has a value.
Although I hesitate to say "this is as it should be", it seems fairly obvious that if the value of time moves to zero, the market is in an aberrant state that is not sustainable long-term.
This return to "normalcy" might even mean a recovery in the market for endowment bonds and other such long-term insurance/savings products that haven't been worth a toss for the past four or five years. Already fixed-term bonds at the building societies are looking attractive -- until interest rates start creeping up in the wake of rising inflation.
In January 2007 a 10-year US government bond earned only a fraction more than a 2-year bond. This ostensibly insane state of affairs had persisted for some time -- indeed, for a sufficient length of time for it to be seen as less an aberration and more a new scheme of things.
In the UK, we had actually reached a situation where for short periods of time you got paid less if you tied up your assets for 10 years than if you did so for two years -- the so-called "inverted yield curve".
The reasons for this lunacy were two-fold. One reason was that, in a liquid market, a 10-year bond isn't "really" a 10-year bond, because you can sell it after two years. If the outlook on interest rates is such that you think you can make more by selling on your 10-year bond after two years than you can by holding the two-year bond to maturity, then there is a rational reason behind paying more for the 10-year bond. Long bonds, in other words, benefited from the lax attitude to risk that has been prevalent for quite a few years (see the "lend anything to anyone who wants to borrow" mantra of the same period).
One possible risk is that, while the two-year bond is redeemed by the issuer, the 10-year bond, if traded in at the same time, needs to find a buyer. In the first case there is no risk of a market collapse, only that the redeemer will fail to honour its obligation. But, in a liquidity crunch, there might not be anyone with any money to hand. If you need cash and there are no buyers for your 10-year T-bonds, you can't sell them back to the government. It doesn't work like that. Until recently, a number of traders seemed to think that it did.
The second reason was a piece of accounting madness which, in simple terms, demanded that pension funds and the like match the time aspect of liabilities to the time aspect of assets. What this moronic accounting rule did not allow for was that the supply of two-year assets was far greater than the demand for two-year liabilities, whereas the supply of 30-year assets was far lower than the demand for 30-year liabilities.
That, indeed, was one reason why the whole trading system existed.
So, the upshot of this stupid law was that pension funds had to pay far more than the "fundamental" value of long-term bonds, while short-term bonds were, relatively, rather cheap. Net result, the inverted yield curve.
This aberration also led to distorted expectations amongst savers such as me. We got used to an Instant Access account offering as much as a 10-year endowment bond. Why lock your money up in the latter when you can earn as much in the former, and have immediate access if you want it?
But, it's all change. The two-year US bond has fallen nearly three points (devastating the return on my Schwab account, btw) while the 10-year bond has fallen by just a single percentage point. Investors are finally being rewarded for being willing to tie up their money for a longer time period. Time, once again, has a value.
Although I hesitate to say "this is as it should be", it seems fairly obvious that if the value of time moves to zero, the market is in an aberrant state that is not sustainable long-term.
This return to "normalcy" might even mean a recovery in the market for endowment bonds and other such long-term insurance/savings products that haven't been worth a toss for the past four or five years. Already fixed-term bonds at the building societies are looking attractive -- until interest rates start creeping up in the wake of rising inflation.
![[identity profile]](https://www.dreamwidth.org/img/silk/identity/openid.png){kind=small}
no subject
Date: 2008-02-28 01:39 pm (UTC)Ten years is two and a half presidential administrations away. It is a year or two past peak oil for those sitting on the fence. For us believers oil production peaked in 2005/6.
Ten years is too far to be certain of anything about the US or western economies.
People may have pumped money into stocks and bonds in the past. The clever money is piling into commodities. Any commodity.
US bonds are a prop for a bygone era. They represent a very bad investment. The world has moved on and sovereign investment funds are looking elsewhere to stash their cash dollars.
no subject
Date: 2008-02-28 01:55 pm (UTC)What you have written explains why the yield on a 10-year bond should be higher (i.e., the price should be lower) not why the yield should be similar (i.e., the price should be about the same).
What market did you work in again?
PJ
no subject
Date: 2008-02-28 02:08 pm (UTC)By uncertainty I do not mean risk or any other metric that would increase a T-bond yield to junk bond status.
What I meant was T-bonds not being the popular choice to sink dollars into. In the past the US treasury has printed dollars like no tomorrow. Americans have bought oil and junk from the mid-East and China. Now the mid-East and China have stacks of dollars and the US has a valueless nation. In the past you bought T-bonds with your euro-dollars to prop up the US so that it kept buying oil and gadgets. What is the point of buying T-bonds from a bankrupt economy?
Two years, five years, ten years, thirty years. It makes no difference which bond you buy because US PLC has a net value of zero so all bonds may as well carry the same value.
I guess my world view is a lot different to yours. Western gov. bonds are a way of buying into our way of life. Il est banco rupto!
PS - Saw Master Channing last night on Five in the PartyPoker 24 hour cash game. That was a lot more watchable than those tournament programmes. Mr C was second fish to Master Feldman though.
no subject
Date: 2008-02-28 03:08 pm (UTC)no subject
Date: 2008-02-28 09:21 pm (UTC)matt
no subject
Date: 2008-02-29 06:50 am (UTC)I agree that, if supply is in sync with demand, it's the sensible way to go. But supply is not in sync with demand. Take an esxtreme example. Suppose there was only one small product with a 30-year maturity. It would be "accounting madness" to demand that all pension funds with 30-year liabilities buy this product. The gain achieved by matching maturities would be more than wiped out by the price disconnect that would result from the vast level of demand swamping the limited supply.
In essence, that was the situation in the pension market a couple of years ago (and it was so severe that 30-year bonds began to be issued again). If the bonds had been issued at the same time as the new accounting rules were put into effect, that would have made sense.
I'm certainly an opponent of borrowing short and lending long, (in the assumption that you will be able to go on borrowing short). But the pension funds were borrowing long (from employees) and lending short (buying short-term bonds). This too has potential drawbacks, but they aren't the same as those that did for the Northern Rock. To compel them to lend long (buy long-term bonds) at a worse rate than lending short (buy short-term bonds) is "accounting madness".
PJ