Thoughts on forecasting...

I thought I'd through in some thoughts on forecasting while waiting for the NFP number and wondering how it is that every single day exactly two people are looking for me on Reunion.com.

I throw out a lot of short-term forecasts on this blog, e.g., saying what I think will happen later this week or even later on the same day. I do this less because that's how I manage money, more because that's the best way to hear what the market is telling you. If spreads are tighter and rates are lower and the curve is steeper, what's that telling you? If the stock market is down severely but bonds are flat and MBS are tighter, what's that telling you? If the CDS are wider on a particular corporate bond, but the cash bonds are not moving as much, what's that telling you?

Some PM's take the attitude that any shift from one day to the next doesn't have much impact on their portfolio performance and isn't likely to change their minds on their economic view. I don't disagree with that. But if you don't watch what's happening every day, then you can't say you understand why the market is doing what its doing. And if you don't understand why, then you can't say you've considered the "market opinion" and either accepted or rejected it.

Anyway, some have observed that I seem to make short-term predictions that are contrary to long-term predictions. Why not? I can see how the market seems to be positioned and make a guess as to how it might shake out. Doesn't change my broad economic opinion.

So, as Steve noted in a comment on Wednesday, my prediction on Wednesday's rally stalling and reflattening started out good, but hasn't followed through. The 10-year topped out on Wednesday around 99-28, and fell off to 99-24 at the end of the day, but the curve stayed at -10bps. Now we've rallied all the way to 100-9. (Steve: I'm looking for your Fed thing.)

Anyway, I'm going to keep making short-term predictions, even if they are consistently wrong, because it helps with the thinking process.