Bonds just don't want to sell off. While shorter maturity issues were somewhat weaker on stronger than expected economic data, the long end of the yield curve continues to defy gravity and be a conundrum for the Fed. Each piece of bad news was greeted by a negative initial reaction in bonds, only to be reversed time and time again. As pointed out last week, the importance of inflation data seems to have diminished. By the end of the week 10 year notes remained stuck at 4.25%, while the long bond actually improved slightly on the week. Positions remain heavily skewed to the short side both in account land and spec country. Put/Call ratios on TLT, Long Bond futures and 10 year Note futures also indicate an overwhelmingly bearish bias to bonds. With the Dow down 6%, S&P down 5% and the Nasdaq off 11% on the year, bonds earning their coupon in 10 year notes and actually better than that in the long end does not seem to be a bad place to invest so far this year. The fixed income market is still insisting that there is something afoot to hold yields at low levels, but the mainstream - lead by the cheerleaders at the Fed - chooses to ignore the wisdom of the bonds.

NOTEWORTHY: We had a busy week on the data release front. The week started off on a positive note, with PPI and Housing starts reported weaker than expectations. This provided more evidence to the previous week's theme that there was a slowdown ahead. The rest of the week was a different story. The Weekly ABC/Washington Post Consumer Confidence bounced a couple of points to -16. CPI was reported substantially higher than expectations both at the overall and core level, Weekly Claims dropped below 300k and the Philly Fed Manufacturing Survey jumped 14 points to 25.3. Lost in the shuffle was the fact that Leading Indicators for March was reported at -.4%, the 7th decline in the past 10 months. The story on the Leaders was that this metric is losing its predictive power and will soon have its components revised by the wise men at the Conference Board. If it ain't working the way it's supposed to, we better change it. Next week will have more miscellaneous releases, with Durable Goods and Q1 GDP being the headliners.

INFLUENCES: As the bond market improves, fixed income portfolio managers become more bearish. (RT survey was down 1 point for the second consecutive week to 40% bullish. This metric is somewhat bullish from a contrarian perspective.) The 'smart money' commercials are still long 389k contracts (off from last week's 413k). The Commitment of Traders data is still solidly bullish. Seasonals are negative into the second week in May. I suggest disregarding the seasonals at this point. On the technical front, the tone remains positive as long as 4.42% holds on the US 10 year notes. As mentioned above, positions are still skewed to the negative side, and bonds are trading very well even in the face of bad news, which are also supportive.

RATES: US Long Bond futures closed at 114-00, up close to half a point on the week, while the yield on the US 10 year note was practically unchanged at 4.25%. My bias is neutral. The Canada - US 10 year spread was unchanged at -8 basis points. We are officially neutral on this spread at this point, but leaning towards selling Canada to buy US bonds. Dec05 BA futures closed the week 83 basis points through Dec05 EuroDollar futures, which was also unchanged from last week's close. Political upheaval is building in Canada and it would not be surprising to see a new election called before the end of the spring. We have not seen much in terms of bond market reaction, but risk premium for Canada should increase. This should translate into higher rates and a lower currency. At 62 it was an official trade recommendation to buy EDZ5 to sell BAZ5. The belly of the Canadian curve outperformed the wings by 2 bps last week. Selling Canada 3.25% 12/2006 and Canada 5.75% 6/2033 to buy Canada 5.25% 6/2012 was at a pick-up of 50 basis points. Assuming an unchanged curve, considering a 3-month time horizon, the total return (including roll-down) for the Canada bond maturing in 2011 is the best value on the curve. In the long end, the Canada 8% bonds maturing on June 1, 2023 continue to look like very good value.

CORPORATES: Corporate bond spreads were wider again last week. Long TransCanada Pipeline bonds were 4 wider at 128 (out almost 20 basis points in the past month or so), while long Ontario bonds were actually in 1.5 to 53. A starter short in TRAPs was recommended at 102 back in February 2004. Auto-paper land was choppy in a two-way trade. Shorter maturity, quality corporates should be favoured over lower rated issues as I believe corporate spreads will continue to be under pressure. Any credit that is connected with the consumer and discretionary spending should be avoided. Corporate spreads have widened considerably, so a pause and perhaps a bounce is due during the next 4-6 weeks. Look to sell the bounce.

BOTTOM LINE: Neutral continues to be the operative word on bonds. I would like to see a pullback to 4.4% to consider buying bonds. As the most negative seasonal of the year winds down heading into mid-May, there will be an additional reason to buy the market. An overweight position in the belly of the curve is still recommended for Canadian accounts. This trade has worked quite well for the first quarter as it has narrowed somewhat as well as provided around 50 basis points of extra carry for the quarter. Short exposure for the corporate sector was advised since February 2004. After a short pause, this sector is expected to move substantially wider going forward. Sell BAZ5 to buy EDZ5 at a pick-up of 62 bps or better was recommended a few weeks back.