Monday 28 November 2016

"Pastcasting"?

We need a new word for something that I did for much of my career as a business economist.  A big part of the job is to try to estimate what key economic numbers will be before they're actually released by the statistical agencies.  This may sound pretty futile to the non-specialist -- why not wait until the real number comes out, and then you'll know for sure? -- but in fact it's one of the key ways an economist can contribute to the bottom line, especially in a dealing room.

How is that possible?  Well, markets don't just react to economic data in a vacuum.  Traders try to set themselves up ahead of time, in order to profit from the data if they've figured it out correctly.  This means that the market doesn't react to the data per se -- it reacts to how the data differs from the market consensus.  A business economist who can estimate the data releases better than his or her peers will therefore set the traders up to make more money than the competition.

You may have noticed that I have been trying hard not to use the word "forecasting" for this function -- because it's not, is it?  You'll regularly read things in the business press like "economists forecast that GDP grew 1.2 percent in the third quarter of the year; official data will be released today", even though the quarter for which they're "forecasting" ended two months earlier.  There's an old saying that "forecasting is difficult, especially when it's about the future"; the customary inaccuracy of the "analysts' consensus" for economic releases often suggests that "forecasting" the past isn't much easier.

What brought on this mini-rant was this story from the Toronto Star today.  It appears that the OECD is going out on a limb and "forecasting" that the Canadian economy "will" grow by 1.2 percent -- THIS YEAR!  By my count there are 33 days left in the year, so for the most part, whatever is going to happen has already happened.  It's not exactly brave, and it's not going to cause anyone to change their behaviour.

So, as I said, we need a new word for this vital but slightly bizarre activity -- but "pastcasting" probably isn't it.

No comments: