Friday, September 11, 2009

30y US Bond yield sends encouraging signal for government bonds

Yesterday's 30y TBond auction according to Reuters had the 2nd highest bid-to-cover since 30Y were brought back in 2006. Even more than this surprisingly positive demand, I think that the trading behaviour of 30y UST yields merits a closer look. Given talks that Asian central banks which continue to buy Treasuries do so largely at the short end, it should be the least distorted part of the curve. Furthermore, given its long-term nature, its trading behaviour should provide insights about long-term prospects for the US economy and its institutions. Amid the high economic uncertainty, ongoing inflation fears, increasing supply pressures and waning credibility of US institutions and the USD as the major reserve currency once should expect that 30y US bonds should fare badly. However, as I will show below, I cannot find ANY evidence that this is indeed the case. Rather to the contrary, 30y US Treasury yields are at an important support level and if breaking through that would be a very significant medium-term bullish sign for US government bonds. On the other side, it would need a move above 4.70% to send a bearish sign for government bonds.

In order to come to such a conclusion we need a bit of background and history: Usually, yield curves bull-steepen and bear-flatten. This is because short-end yields are more volatile than their longer-term counterparts (as during normal times central banks conduct monetary policy via changes in the level of short-end yields and because long-run expectations are more stable than short-term expectations). However, over the longer-term this relationship is anything but stable. As the first chart below shows, the relationship between the steepness of the US yield curve (I use 3y and 10y constant-maturity Treasury yields due to data availability) and the level of short-end yields has shown significant changes over time.
During the inflationary period between thee late 60s up to the start of the 80s, the yield level shifted higher across the curve/the yield curve steepened structurally (i.e. for every given level of short-end yields, the yield curve was steeper than it would have been otherwise).
The Great Inflation: yield level shifts higher across the curve, breaking the cyclical relationship
Source: St.Louis Federal Reserve, Research Ahead

The reason for this has to be seen in the fundamental economic environment. As the next chart shows, during the same time period the so-called Phillips cure exhibited significant structural shifts towards higher unemployment and rising inflation. Theory as well as empirical research suggests that these structural shifts in the Phillips curve are down to rising longer-term inflation expectations.
Worsening inflation-unemployment trade-off amid rising long-run inflation expectations
Source: Bloomberg, Research Ahead

During the disinflationary 80s and 90s, both trends reversed. The inflation-unemployment trade-off improved as the US Fed moved to an implicit inflation targeting regime which brought long-run inflation expectations lower.
The disinflationary 80s and 90s improved the inflation-unemployment trade-off
Source: Bloomberg, Research Ahead

Once again the strucutral shifts in the Phillips curve were mirrored by the behaviour of the bond markets. Yields moved towards structurally lower levels/curves towards structurally flatter levels over the past 25 years.
The past 25 years: Structurally lower bond yields
Source: St.Louis Federal Reserve, Research Ahead

Now, if we bring 30y yields into the picture, the story is the same (but available history is not as long). If we look at the steepness of the 3-30y curve vs. the level of 3y yields we see exactly the same behaviour over the past 25 years: a movement towards structurally lower yields/flatter curves.
30y yields behave in the same way
Source: St.Louis Federal Reserve, Research Ahead

Given that it is long-run inflation expectations which are a key driver for long-term yields, therefore, 30y yields do not currently send any danger signals. Rather, their relationship with 3y UST (and again it is the short-term part of the curve where Asian central banks are most active these days), is extremely stable, baring a brief exception late 08/early 09 during the worst part of the financial crisis. If inflation would indeed be such a problem down the road why then is it not showing in the asset least distorted and most at risk of such a development? To me it is a) because and as I have stated on numerous occasions inflation fears are overdone and b) because the underlying economic developments do not send such a signal as is evidenced by the stable behaviour of the Phillips curve over the past 2 years.
Finally, the technical situation of the 30y US yields is extremely interesting as the below chart shows.
30y US yields remain in long-run downward trend but are at an important support
Source: Bloomberg

30y US yields have reached a low of 4.173% in June 2003 and traded 3 times to 4.192% in June 2005. In July this year they reached a temporary low of 4.194%, however, and in contrast to their 10y counterparts they have last week tradeed below this July-low and moved to 4.121%. Currently they trade at 4.20%. Should they overcome this 4.10-4.20% area, then that would be a significant long-term bullish signal for the entire UST market! On the other side, the multi-decade downward trend is running at roughly 4.70% at present. I think that as long as this downward trend has not been broken, the odds for an ongoing constructive bond-market environment remain favorable.

No comments:

Post a Comment