Monday, November 9, 2009

Rates Strategy: Sticking to longs

Last week I upgraded my tactical bond market outlook back to bullish. Since then the 10y Bund future has dropped by roughly one point and the US TNote future has lost some 6/32nd. Therefore, the move back to bullish seems ill-timed. However, I maintain my view.
Technically, 10y Bund yields look trapped in a range of roughly 3.20-3.40%, a range where they spent most of the past months with only short-lived deviations on both sides. 10y US Treasury yields are still trading within a slowly falling donwnward trend channel (see chart). Arguably, Bunds look a bit more bearish than USTs but still do not give a clear sell signal from the technical side.
10y US Treasury yields remain in downward trend channel

Source: Bloomberg

Positioning-wise, not much seems to have changed if judged by the CFTC data for US bond futures. Non-commercial accounts remain positioned on the short side with significant shorts at the long-end of the curve. However, given partially offsetting longs in 2y and 5y futures, overall net shorts are considerable but still only roughly half of what they were at the end of May, i.e. just before yields hit their highs.
Cross-market wise I frequently use the chart below to compare 10y govie yields with the development of equities and commodities. The combination of equity and commodity performance should give a good indication about the changing outlook for real growth as well as inflation and therefore for nominal growth. Given that it is nominal growth which should be the key driver for nominal bond yields, government bonds should exhibit a close correlation to the combined equities and commodities performance.
Government bond yields vs. equities and commodities

Source: Bloomberg, Research Ahead

To quantify the cross-market impact, I regressed 10y Bund yields on the ESTOXX and the CRB index. With this regression I can calculate an implied value for 10y Bund yields. The chart below compares this implied value with the actual 10y Bund yield and also plots the difference (labelled Error Term) in basis points. Furthermore, it gives the 1.5 standard deviation bands for the error term (corresponding to roughly 20bp). Unfortunately what I missed at the start of last week is that government bond yields were looking expensive based on this simple model. In fact, the error term was at 24bp corresponding to 1.9 standard deviations. This means that 10y Bund yields were trading 24bp below the level implied by equities and commodities. In the meantime, however, this expensiveness has corrected significantly and moved down to 5bp at present. In turn, current 10y Bund yields were expensive at the start of last week (highlighting that my change in the tactical outlook was indeed ill-timed) but have since moved back to a level which is not deviating to a statistically significant extent from the implied cross-market level.
10y Bunds have moved back from overly expensive level
Source: Research Ahead

Fundamentally, I stick to my shorter-term view. That is, I think that seasonally adjusted data risks coming out rather on the weak side as the typical seasonal strength going into autumn should be rather less pronounced than usual. Friday's employment report confirmed this assessment. While the not-seasonally adjusted payrolls according to the establishment survey increased by 641k, the seasonally adjusted payrolls decreased by 190k, meaning that usually employment in October increases on average by 831k. However, according to the household survey, unadjusted employment increased by only 9k, resulting in a seasonally adjusted decrease of 590k. These numbers are weak. And remember that the establisment survey excludes self-employed as well as the job-losses by business deaths and instead uses a statistical method to account for net business births/deaths which added 86k in jobs during October. Given this method and given my belief that amid the lack of credit availability for small firms, the recession for small firms is far from over, the establishment survey should paint too rosy a picture of the employment situation.
Therefore, in light of all this, I maintain my tactical bullish outlook.

No comments:

Post a Comment