There are two elements here... Firstly, a longer-dated bond is mechanically going to have less sensitivity to various short-term phenomena, such as central bank actions, spot inflation fixings, money mkt rate fluctuations, etc. Obviously, since all these things can go in any direction, this can either be a good or a bad thing. Secondly, I believe what you might be alluding here is the significant convexity of long-dated bonds. There is undoubtedly value there, but that's value for which the mkt is normally going to demand a certain price. Will the price in this case be right?
This is confusing. Suppose an investor wants to have 20% of their portfolio, or X dollars, invested in bonds. They would need to buy X dollars worth of bonds, regardless of maturity. They can't buy "a little bit of huge duration bonds to free up cash", since they need to spend a fixed amount. The resulting portfolio could have a lot of duration risk or a little, depending on what bonds are bought. Obviously, if your goal is to have a portfolio with a particular duration, the procedure will be different.
Is your expectation that in a non-captive market these things would still trade at a premium to lower duration bonds? PS. I think French had consols trading up until very recently, no?
I am not aware of anything like that in France. The UK DMO redeemed the last "war loan" perps a couple of years back. There was talk of them issuing new ones, but it never went very far. It's a tough question. On the one hand, we've had a few century bonds that went like hot cakes (e.g. Mexico, Ireland, etc). In most cases, they're either private placements and/or very small size, which would suggest that they were driven by reverse inquiries. In the UK, the long-dated gilts are sold mostly through syndications. So there's definitely some appetite out there. The problem is that the US treasury isn't likely to want to operate like that. Can they have a regular auction program with $10-15bn 100yr bonds a pop? Who will have the capacity to take that down? How will the dealers handle this? So I don't know if these bonds will trade at a premium or discount, since it really depends on demand. If there is lots of demand and the Treasury doesn't abuse it in the name of liquidity, these bonds will fly. Otherwise, they will struggle. Finally, one thing I would add is that, to the best of my knowledge, the US pension funds, unlike their counterparts in U.K. and Europe, don't really want ultra long-dated paper, at least for now.
Kinda makes you wonder about the embedded credit risk (well, devaluation risk), doesn't it? Once you start thinking multiple generations, risk of an event becomes totally real IMHO. Ps. Mart, did you see my PM?
I'm talking about the latter. My point is that its a way to guarantee duration (vs futures) without being exposure to a upside surprise in the average repo in the long-run. Plus you save all the commissions/bid ask spread/time waste during roll overs
Right, so if your hypothetical investor is targeting a particular duration, are they hedging a specific liability? If so, is a century bond really what they need? It's one thing that a pension fund, which either has an expectation of future inflows or is desperately underhedged, needs this long-dated paper. But hard to imagine same from a regular investor... I am not sure why you're comparing these bonds vs futures. The relevant comparison is vs 30y bonds, for instance, or a bond index ETF.
I'm comparing it to futures because they are both a levered play on rates. Someone could always use futures and 'free up' cash for other purposes. But that exposes the person to upside surprises in the repo rate (plus the commissions, spreads). The huge duration bond is more protected in that sense. A regular investor could use it as part of a 'all-weather' type allocation that balances exposure to rising/falling rates, rising/falling growth, rising/falling inflation. It might also make sense for foundations that follow a similar approach
Sure, but at least in the US and Germany you have 30y bond futures. Their sensitivity to short rates isn't going to be that much greater than that of the ultralongs.
I now realize a mistake I had made. Deriving leverage from the bond should be pretty close to deriving from the futures (although cash bonds still have benefits) so there should be no savings (in terms of yields) or difference in interest rate exposure. Hence, there is no reason to expect much from the 50y or 100y A portfolio with 20% 10y bonds Can be achieved by a 10% 30y bond 10% free cash 0.3% final yield OR 10% 30y future (futures exposure) 10% T-Bills (cash bond for futures margin) 10% free cash 0.3% cash bond yield - repo rate + t-bill rate (which usually will cancel out the repo) = 0.3% So the final yield should be the same (hence no savings) and the interest rate exposure will be equal (since the T-bill hedges out the chance of a repo rate increase, that's what I was missing) But using the cash bonds instead of futures still offer benefits in terms of saving commissions, spreads, time (no roll over to deal with) and depending on the person, maybe even taxes (also, didn't in 2008 the repo rate went weird? So maybe that's another benefit of cash vs futures). But it doesn't look very likely that the 50y or 100y would provide much in terms of additional benefits over what is already avaliable in the 30y cash bond (especially given the 30y STRIPS already exist and they provide a crap ton of duration)