So bonds have rallied like hell for a while, especially in the days leading up to the jobs number. All I've been thinking the whole time is bonds are a sale, they're too expensive regardless of whats going on (keep in mind I'm not a position player, I'm a scalper so this isn't longterm)...and they're bound to pull back. Like alot of other guys I was dreaming about how great it would be if the jobs number came in super high and out of line and I would have an excuse to sell like crazy. well, the jobs number was estimated at 120K, ADP came in low and there were some revisions down, some banks saying as low as 100K and some whisper numbers even 50K. Myself and I think alot of other bond traders pretty much disregarded this, ADP is kind of a bullshit number (ADP is a jobs report that comes out 2 days before non farm). I'm talking with other traders beforehand and there's a consensus about what'll happen depending on where the number comes in, its pretty much decided if it comes in low we'll rally pretty well just because thats what markets do, right? And if it comes in high we'll all get rich selling. Number comes out...50K, ridiculously out of line and low...exactly 10 seconds after the number I'm buying (checked the execusion history) offers are gettng pulled and I'm worried I won't even get filled. Then suddenly, I do. Then the revision for last month comes in kinda high, I hear another trader say, "watch for the geek out" and a second later, on CNBC in the back ground I hear Santelli shout, "We have a U- turn in the pits!" Instantly I'm smoked, barely managing to get out losing only half my week.It was alot worse for some guys I knew. I managed to come back because the day was a great trade in my market but i was still down fees. This all happened in exactly 14 seconds mind you. So whats the lesson? Its not to wait for bullshit revisions on last months numbers...its the obvious one that we all know, livermore wrote about it well....The whole fucking market was long and everybody knew it, and obviously a weak number was a chance to get out. We were gonna break on friday regardless...because people needed to sell. It was just a crazy moment...ask any bond trader whether they wanted to be long or short friday not knowing anything else, they'd say short. Ask them then what would they do if non farm came in way low....they'd buy. Basic stuff. Which was the trump card? Selling...because when everybody is already long, you have to be a seller. Sometimes we forget this basic stuff. Just a simple trade that illustrates one of the most basic lessons in trading. Cheers.
the whole jobs number is a farce....the back months numbers jumped large and the annual state supplied data resulted in 810,000 jobs added.... amazing that so much is bet on these borked data..... since it's lotto time, me bet that Benny and the Jets jacks the rate 25 more this month..... I can't decide if I should trade bonds or go bowling..
I usually stick to option lotto ticket style plays on NFP, especially during the first hour or two of trading afterward. I have had the opinion for a little while the market is a bit overextended long (and based upon some mortgage flows this week heading into the number had a good idea who was gonna have to sell a little bit on Friday). So I figured the 10yr Nov 107 puts were a good play into the number. I was either gonna be right or wrong by the end of the day (not instantly after the number). Basing my opinion more on market positioning into the numbers rather than actual data levels always seems to help as well. Bought the puts for 4 ticks and sold them at 8 ticks. Knew the risk and didn't get chopped up trading outrights. I have definitely learned the hard way that NFP is hyped up so much I am probably at an information disadvantage right after the figures are released, so I take advantage of the information I do know, which is market position into the number.
Ok, but you are aside the treasury pits...how can we analyze the market position off the screens (apart MP theory)?
You don't need to be in the pits to know market positioning. These position develop over weeks and sometimes month's time. Definitely get a premium news service like MCM Capital Markets (on telerate website I think), which usually reports specific account (hedge fund, real money, swap desks/dealers, etc.) activity. After this analyze open interest at the beginning of every day, which will tell you generally who was active the day before. Finally, watch the committment of traders reports for any irregularities or sizable positions leaning one way.