Click here for Mike Amey's biography.

Over the last 12 months bond markets have been on something of a rollercoaster ride. The Fed, the European Central Bank (ECB), the Monetary Policy Committee (MPC) of the Bank of England and even the Bank of Japan have pushed official rates higher, and higher than the markets expected (in some cases ourselves included).

In response, bonds weakened significantly over the first half of 2006. By mid-2006, European and U.S. intermediate yields were 75 basis points higher. Japanese yields were some 50 basis points higher, as were intermediate U.K. yields. In contrast, long-dated U.K. yields were unchanged. Subsequently some cracks in the U.S. economy appeared, the Fed paused, and inflation appeared to peak (in the U.S. and Europe, at least), pushing bond yields back down. Intermediate bonds have now recovered around two-thirds of the sell-off. But what about the long-dated U.K. bonds?

Whilst they did not participate in the sell-off, they have participated in the rally. Sounds like a pretty attractive bond – it does not participate in sell-offs but does participate in rallies? Time to buy? Not exactly.

Long-dated U.K. bonds have been the prime beneficiary of two key events in the last 12 months, neither of which is likely to repeat itself over 2007 and beyond. Meanwhile, long-dated U.K. yields are low relative to both shorter-dated bonds and overseas bonds, which makes the hurdle for long-dated U.K. bonds to keep outperforming incredibly high. In our view, the prospect of long-dated U.K. bonds remaining the best performing part of the bond market is low, but first of all, the history.

Late 2005 and early 2006 witnessed a rapid appreciation of long-dated gilts led by a combination of long-dated, asset-liability flows at a time of little supply. Widespread underweights to long gilts by end investors added fuel to the buying fire, and all this at a time when the central bank was sounding increasingly hawkish (i.e., very much anchoring long-dated inflation expectations). As such, by early January 2006, long-dated bond yields fell sharply to a low of 3.8%. A heady combination of asset-liability matching, end investor positioning and hawkish rhetoric from the central bank all played their part in the impressive performance of long-dated U.K. bonds, both on an outright and relative basis. So what’s to stop this from happening again over 2007?  After all the asset-liability flow is still very much in play, as is a hawkish central bank.

From a domestic perspective, two key developments suggest we should not expect 2007 to replay 2006. First of all, whilst the flow of funds into long bonds will undoubtedly remain, the end investor is much less underweight in long bonds than they were this time last year. Industry estimates for 2006 asset-liability flows average around £25 billion, roughly double the 2005 number. To put this into context, the cash value of outstanding U.K. investment-grade bonds, both fixed and index-linked, is £825 billion. If we assumed an average market underweight of 5% by end investors, that would imply a £41 billion underweight, which needs to be closed (i.e., twice the size of the 2006 asset-liability flow). This is obviously a gross approximation, but the point is that whilst asset-liability flows are clearly large, it is only when you put these flows together with the buying by end investors to close underweights that the demand starts to exceed long-dated supply (which over the year has been around £40 billion in gilts, £60 billion in corporates).

All of this clearly ignores the swaps market. However, the basic point still stands: the market may be able to meet the asset-liability demand, but it cannot meet the additional demand from the unwinding of unprofitable short positions. As I said earlier, whilst the asset-liability flow will likely continue, the swing factor of short covering is not likely to be repeated in 2007.

The other differentiating factor is that at the end of 2005 the market expected MPC policy rates to be unchanged to slightly lower, whereas today the market expects the MPC to raise rates at least one more time. In other words, the market expects the MPC to be hawkish, and has priced those expectations into long-dated U.K. bonds.

If the key catalysts for the 2006 rally in the long end are no longer in place, what about market levels? After all, is the fact that long bonds have performed so well a reason to buy them or sell them? To my mind there are good reasons why you would use this as a time to sell long bonds in favour of other U.K. or overseas bonds.

The first point to note is that if 30-year gilts yield 4% and 10-year gilts yield 4.50%, implicitly the 20-year gilt yield in 10 years time is 3.75%. If you went further on and looked at the 10-year rate in 20 years time, the current level of the market would imply a yield of 3.45%. In the context of medium-term inflation expectations of 3.15%, this implies a “real” yield of 0.3%. This may all sound very theoretical, but forward rates below 3.5% in nominal bonds (or 4% in long gilt yields) do generate interest from corporate issuers and market participants to sell long-dated U.K. bonds. This is what happened in January 2006 (when the gilt market made it onto the front page of the FT!) and more recently in October. 

We can do a similar analysis by looking at long-dated U.K. yields versus other parts of the U.K. market or the overseas markets. The charts consider how you would have fared in an equally weighted portfolio of long gilts, long U.K. non-gilts and index-linked gilts relative to an allocation of one-third into either the all-stocks U.K. bond market, all-stocks U.S., all-stocks European or all-stocks Japanese bond market. As you can see, long-dated U.K. bonds go through protracted periods of over and underperformance (numbers below the zero line represent over performance of long-dated U.K. bonds). The periods of protracted outperformance do not last forever and often finish at the point of greatest market conviction that long-dated U.K. bonds are indeed “bulletproof”.

As we can see, we have just been through a period of very strong outperformance and many argue that long-dated U.K. bonds will remain the strongest part of the bond market. However, we can also see from Chart 2 that the yield difference between long-dated U.K. bonds and the other markets is close to historic highs. Effectively, there is a point whereby the yield differential against other markets gets sufficiently attractive that forward-looking investors start to move back out of long U.K. bonds into other markets. In turn, as the market sees that low long-dated U.K. yields are not always the best place to be, so more investors turn to better value securities and so the performance cycle turns against long U.K. bonds. We are now approaching those levels. Recall that because long-dated U.K. yields are so low relative to shorter maturities, to avoid being better off in cash at 5% you need long-dated yields to keep falling to levels, which have seen strong resistance in the past. Put another way, if the one-year gilt yield is 5.2% and the 30-year gilt yield is 4.2%, I need to generate a capital gain on the 30-year gilt to make up for the yield difference (and recall that relative to last year when one-year rates were 4.3%, I have a lot more yield to make up for this year). As the yield difference between short- and long-dated U.K. bonds drifts wider, so the burden of outperformance becomes ever greater.

So, we know that the best place to be in the last 12 months has been long-dated U.K. bonds. We also know that persistent demand for long bonds and asset matches should keep a lid on any rise in long yields. But does this mean they will maintain their status as the bulletproof bond that never really sells off, but does participate in rallies? Clearly, it will take a pretty amazing set of circumstances to create a big sell-off in long bonds, but we should also remember that it is now reflected in the price – that is why long-dated yields are fully 1% below short rates, 0.5% below U.S. long yields, and only 0.5% above European long yields (where nominal short rates are materially lower). With the central bank interest rate cycle likely to turn and inflation having peaked, I wouldn’t mind betting that in the search for yield, one or two of us will seek assets away from long-dated U.K. bonds with very fruitful results.

Mike Amey
Senior Vice President