Ralf Preusser
We believe the rather dramatic moves on Thursday were mainly driven by market technicals. Because there was no economic data released last week that could have triggered such a significant move, we have to look elsewhere to explain what happened.
Indeed, there has been a slow-burn factor of stronger macro data globally and in the US. So, for example, we havenÆt seen a significant spill-over of housing market weakness into the wider economy or even to the consumer. Also, the capex weakness that we saw a few months ago seems to have disappeared again. So some of the main worries that people had concerning the global economic development for this year have been unfounded, and this has been an underlying factor in the grind-higher that we have seen in yields.
But some of the underlying asset allocation patterns that we have seen shifting over the last few months also contributed to the moves that led up to last week. So for example, we are seeing a re-allocation of central bank reserves away from your typical government bond portfolios towards risky asset classes, both in oil exporting countries as well as from Asian central banks.
We are seeing a significant reduction in the number of mortgages that US banks accumulate and buy for their balance sheets. If you look at the growth of mortgage portfolios of US commercial banks, it ground to a complete halt in October/November and has recently even been falling. The reasons are the well-publicised problems in the US housing market: banks are not necessarily happy with the size of their mortgage portfolios in a flat yield curve environment and the risk that poses to their balance sheet as a whole. So by slowing the growth of their mortgage portfolios relative to the growth of other assets, they are effectively de-risking their portfolios. But these are longer-tem issues.
What we think happened specifically last week was a re-steepening of the money market curve - the Eurodollar three-month to two-year sector that was heavily inverted in the first few months of the year started re-steepening and is now basically flat. A mathematical consequence of this is that flattening trades put on further out the yield curve (going long the ten-year sector versus the two year sector, or going long the 30-year sector versus the two-year sector)switch from being positive carry trades to negative carry trades. A lot of these trades were closed out as a consequence last week, and that was one of the legs in the sell-off. (A reversal of a flattening trade implies selling the long-end of the yield curve and buying the front-end: this is exactly the steepening that we saw.)
Another major positioning unwind came from the real money community. DeutscheÆs positioning surveys showed that the international real money community was heavily overweight in the US against Europe, relative to their benchmark. This was one of the main risk positions that they have held. With this first leg of flattening trades unwinding, many were forced to take stop losses on their US trades, leading to another wave of US fixed-income selling relative to European fixed income. This explains why the curve not only steepened, but also why the US underperformed relative to Europe.
Further, two mortgage-related factors are at play here.
In US MBS, borrowers have the right to pre-pay their mortgages at any time, therefore investors in mortgage-backed securities face the risk of early redemption. The speed at which these bonds are redeemed depends upon how quickly borrowers pre-pay their mortgage. When pre-payment speeds increase (i.e. when interest rates fall), the duration of mortgage bonds drops. Similarly, when pre-payments speeds decrease (when interest rates rise), mortgage bond durations lengthen. If pre-payment speeds increase, those investors trying to maintain specific duration targets will need to re-hedge to reduce portfolio risk.
A lot of mortgage products saw a significant lengthening of the duration profile, which investors had to counter by unwinding some their mortgage portfolios, or hedging it by either paying in the swap market or by selling government bond futures against it.
The second is adjustable-rate mortgages. These US mortgages are available with a teaser interest rate, fixed typically for the first two to five years, which is then reset to a market interest rate û typically over Libor. Many of the adjustable-rate mortgages that were issued over the last few years are due for their first interest rate reset. However, the spreads between a post-reset adjustable spread mortgage interest rate and a normal 30-year fixed mortgage are at historical tights. It is therefore in the investorsÆ interest to move from adjustable-rate mortgages into 30-year fixed-rate mortgages. From a nation-wide supply perspective, this means the replacement of floating-rate products into fixed-rate long-dated products, which significantly increases the supply of duration in the market.
If one assumes the demand for duration stays constant regardless of how mortgage borrowers choose to refinance their mortgages, the only way the additional supply for duration can be absorbed is through a steepening of the yield curve, and through wider mortgage spreads.ö
Robert Lind
Yields have been surprisingly low in the last few years, particularly in the US, but in Europe as well. That fall in yields reflected better expectations in inflation, but also lower real yields and a fall in the term premium. So the fall in yields benefited from a reduction in all the major components of nominal bond yields.
Over the course of the last few weeks, we have seen an upward drift in the real yield component of bonds, and that is something that is long overdue. We had seen this gap open up between real GDP growth globally, and the level of real bond yields, where the GDP growth was running ahead of real bond yields. Pessimists on economic growth believe it would fall back towards the levels of real bond yields. We were of the view that real growth would lead to a growth in real bond yields, and I think this is what we are seeing now.
The dramatic move last week was triggered by a combination of better-than-expected US economic data, the latest employment report, the ISM surveys but, alongside that, we had also seen further evidence that the global economy was still pretty robust. WeÆve had higher interest rates in Europe over the last few months, both in the Eurozone and outside, and at the same time expectations of higher interest rates in some parts of Asia and some of the other emerging markets. You could argue the dramatic move last week was a correction triggered by that constellation of factors. And I would say that it was long overdue, since it was more difficult to understand why yields were so low, than why they have suddenly risen. It seems to me that yields now reflect much more the underlying fundamentals of the US economy than they did two months ago when they were clearly abnormally low.
Whenever a market move of this size occurs, there is always a suggestion that it reflects either macro-economic fundamentals or technical factors. I donÆt think itÆs possible to disentangle the technical factors from the macro-economic factors. ItÆs possible that on an hour-by-hour view, some of those factors could have had a role in pushing yields up, but itÆs also clear that investors, who had previously been quite optimistic on the outlook for bonds have obviously changed their mindset over the course of the last month or so.
Irrespective of those technical factors affecting markets, itÆs also pretty clear that if investors had still been optimistic on bonds, those technical factors would have had minimal effect, because as soon as yields rose, investors would have bought back into the market. This has been the lesson of the last few years û whenever we have seen any rise in bond yields, there has always been a willingness on the part of investors to buy back into the market. This time around, I suspect there has been a greater degree of caution about that willingness to buy back into the market, which is why I think the rise in yields has been pretty consistent over the last few weeks. Last weekÆs sharp step followed what had already been a fairly substantial rise in yields over the previous few weeks.
Investors are beginning to re-assess their assumptions on the US and global economies. In order to believe that US Treasuries were appropriately priced when yields were down at 4.5%, investors also had to believe that the Fed was about to start cutting interest rates pretty quickly. There has been little suggestion of that. Those of us who didnÆt expect Fed rate cuts have always felt that the burden was on the bond market to convince us that it was appropriately priced, given that the economic data didnÆt seem to justify that pricing. Therefore you didnÆt need much in the way of better economic data for the bond market position to unwind. I think for bonds to have been right when they were yielding 4.5%, you would have had to see further deterioration in the US economic data, and the fact that you havenÆt had that (if anything, itÆs started to improve) is one of the reasons why yields have actually risen.
¬ Haymarket Media Limited. All rights reserved.
