Monday, March 2, 2009

Checking on my past forecasts...

Note: not many professional forecasters actually bother to assess how well they've performed in the past. (This was the topic of a William Raspberry column many years ago. Q: "Do you do forecasts?" A: "Oh, yes - we're experts in the field!" Q: "Could I see how well you've performed?" A: "We don't track our performance.")

Me? I figure you don't get better if you don't judge yourself.

Here goes.

Back in October 2008 I asked, "Where's the bottom???" Here was my answer at the time:
A very quick, back-of-the-envelope computation suggests that since the '70s, the largest percentage decline in DJIA has been just under 50% from previous peak.

Using this as a guideline, the bottom would seem to be around 6000.
We're at 8500 today.

The 50% bottom is the WORST since 1970.

The range seems to be 25% to 50%. (Again, these are very hasty, eye-balled figures!!!)
If you bet on 40%, that'd represent about 7000.
30%? About 8000.
Wall Street tumbles anew as financials slide
2 Mar 2009
NEW YORK – Investors' despair about financial companies and the recession has brought the Dow Jones industrial average to another unwanted milestone: its first drop below 7,000 in more than 11 years.
If you believe we're in a meltdown like we've seldom seen, that 50%, 6000 number is looking not too bad.

A quick observation about that headline: Wall Street tumbles anew as financials slide.
One of the things I find particularly disconcerting about daily analysis of the market is that there is always a specific cause cited for market movement. Today it's "despair about financial companies". As a statistician, I've been trained - and probably to some extent came to the job with a pre-disposition - to ignore the specific detail... to look instead at the larger collection of data available: trees don't matter, only the forest is of interest.

I note that my "back of the envelope" calculation was based on a very crude, almost brain-dead, reading of Dow Jones performance over the past several decades. No deep analysis. No consideration of "fundamentals" - just a quick look at how the numbers have behaved in the past: What do deep recessions look like?

What I've yet to get even close to right is timing - WHEN will we hit the bottom?
[boastful note: in my last job, I predicted this recession starting sometime in 2008/9... the prediction was made in Q4 2006.]

Have a nice day.

No comments:

Post a Comment