wilmott.com

Recently the US Treasury decided to stop issuing 30-yr bonds, and as a result many fund managers were nonplussed. Some were even at a loss for what to do! No need to worry, for perhaps a synthetic 30-yr bond could be created.I imagine a contract based on 10-yr stock with a 2x roll provision. It would become feasible if liquidity develops. No doubt this roll bond would entail some quantmagic in order to succeed. The rest is up to you fellows; any ideas?

but would it be backed by the full faith and credit of the US government...? that could be crucial.

I think you have something in mind a depository receipt holding $2 of 10-year, $1 of 7-year and short $1 of 5-year and $1 of 3-year treasuries for each $1 face value. That had better than a 97% correlation with the weekly return on the 30-year over the last 7 years, with the same mean. One one hand, a quant could refine the holding by making the portfolio weights more precise and possibly making them dynamic. Also, instead of a periodic rebalancing you could smoothly buy and sell smaller quantities to keep the match close. On the other hand, simplicity is a virtue, and I don't know how much the market would care about a 98% versus 97% correlation. Someone who really wanted to guarantee a fixed 30-year return could buy one of these with a forward or option to cover the small residual risk. Of course, the hedge would have counterparty risk, but it is such a small part of the economic value, most investors would be willing to treat the receipt combined with a AAA-legal vehicle hedge as riskless.The receipt itself could be structured to be essentially riskless as long as the net coupon flow is positive, e.g. that 2 times the 10-year plus the 7-year rate is greater than the sum of the 5-year and 3-year rates. That has never come close to happening, even in the inverted yield curve days of the early 1980's.Personally, I can't see that people would pay a lot for this. It's too easy to replicate. People who want a high duration have other ways to get it. But I've been spectacularly wrong before about such things.

Although most theorists state that the lower government yield is because of perfect "credit," this doesn't really hold water. In reality, Government bonds are extra valuable because of their utility as collateral. Any synthetic would have to possess this same utility, which would be unlikely.To obtain the same sort of interest rate exposure, one could go long a 30 year swap, or a combination of a 30yr swap plus an appropriate amount of the 10 year treasury spread ...

I think Lenny is asking how best to 'extrapolate' into uncharted territory, where no related contracts exist. But as Aaron says there may be some stats to point you in the right direction.P

there might be an issue with the liquidity premium. benchmark givernment issues tend to be the most liquid; even for such a long bond where the turnover is smaller, it may be more liquid than a comparable swap (correct me if i'm wrong on that!). but i would definitely think that a 30yr T-bond would have more liquidity than an OTC synthetic product. this could be important with regard to the collateral issue pat mentioned.

Although most theorists state that the lower government yield is because of perfect "credit," this doesn't really hold water. In reality, Government bonds are extra valuable because of their utility as collateral. Any synthetic would have to possess this same utility, which would be unlikely. >>In that case, I think you could design a synthetic floating rate 30-year treasury that was a pretty good substitute for the real thing. GNMA would be the natural issuer, since it has full faith and credit of the US government, just like treasuries. In a pass-through structure, the security buyer owns the underlying collateral, but the servicer can distribute principal and interest pro rata. So you would have the same prinicpal protection as a real treasury (because you have $1 of real treasuries underlying each $1 investment; and you own the collateral), and you would receive whatever interest the underlying treasuries threw off. The pool would run for 30 years, paying out interest and reinvesting principal payments, then liquidate.If GNMA didn't want to do it, FNMA or FHLMC have the financial sophistication and quasi-government backing. Other agencies might want to do it, or even a major bank. But I think it would be important that if there is more than one issuer, the securities trade interchangably. Otherwise you wouldn't have the necessary liquidity.

a) swaps are more liquid than treasuries by orders of magnitudes ... there are only a finite number of bonds (which is why they can be squeezed and can go on special) and one issuer, but anyone with a decent credit rating can issue swaps.b) people are looking into agencies as substiture for the long bond; also looking into using GC (general collateral ... ie off-the-run bonds). But I don't think there are a lot of (non-callable) agency bonds issues at extremely long maturities.c) how do they do it in Britain? People routinely strip the swap curve out to 40 to 60 years.

swaps are more liquid than treasuries by orders of magnitudes ... there are only a finite number of bonds (which is why they can be squeezed and can go on special) and one issuer, but anyone with a decent credit rating can issue swaps. >>There are so many different definitions of liquidity, it may be time to retire the term. While swaps are certainly more flexible and easier to trade quickly in large amounts; they cannot compare with treasuries for interchangability and risklessness. In a liquidity crisis treasuries are a safe haven, swaps are not; and swaps with different counterparties will not be interchangable (even if the different counterparties have the same credit ratings).There is a need for treasuries, just as there used to be a need for banks to have some gold and silver in the vault. You need long-term ones for many applications. But I don't know if you need 30-year bonds, 10 years is a long time.

Yes. You cannot (easily) use the PV of a swap as collateral. But if I need to go longer or shorter some given maturity ot manage my interest rate risk, I can easily do this through new deals instead of interchanging my swaps.During liquidity crises, the US swap spreads can blow out, but usually treasury prices soar, depressing the yields, while the swap rates remain more-or-less unchanged.