Forecasts are for fools

I noted yesterday that my Twitter feed seems to be full of bears apoplectic with anger while bulls are generally calm and chill. Of course bears have been calling for the end of the world ever since we hit near double digit inflation in 2021 and can’t understand how the economy has managed to remain buoyant and worse - how the markets have crashed.

The recession was supposed to come at the end of 21, then start of 22 then end of 22 then start of 23 then... Well you get the idea.  Like a broken clock that is right just by luck twice a day the bears may yet be proven right. Or maybe not. Maybe  today’s bears are just the ghost of Robert Prechter reborn to amuse us with their ridiculous stubbornness that borders on religious fanaticism. (Yea I know Prechter isn’t dead - but if you followed his trading advice your account certainly is)

All of this plays into my thesis that Forecasting is for Fools. Yesterday it rained all day long in New York city. Not just a drizzle but an endless downpour from 6AM until midnight. Now imagine if I came to you in the spring of 22 and asked you to tell me the exact day a year from now when it will pour endlessly  in New York city. You would look at me like I am an idiot, maybe smile and do what? Make. A. Guess. 

That is what all financial forecasts are. When it comes to long term weather forecasting we understand that intuitively, but when it comes to finance we tend to venerate the assessments of any analysts if they were lucky enough to guess correctly a year ago and better yet if they were lucky enough to guess correctly  two years in a row. Then they can walk on water  as Joe Granville once actually did in front of investors before plunging many of his followers into near bankruptcy with his idiotic forecasts in late 1970s and 1980s.

Back when I used to do the morning show at CNBC I would drive the anchors crazy by saying that I really did not have strong  long term views and that at best I just tried to anticipate the next 72 hours. That wasn't a number that I just pulled out of my hat. That data point came from a one time visit to a quant fund in New York that was operating out of an opulent early 20th century beaux arts skyscraper with more computer power than the Pentagon. I found the sharp contrast between the ornate brickwork of the building  and ultra modern luminescent decor of their offices to be so jarring that the visit stayed in my memory.

The guys at the quant fund told me that they ran thousands of tests on currency data against hundreds of technical and fundamental parameters and that the forecasting value of the data degraded by about a half within 24 hours and then disappeared entirely into noise 72 hours thereafter.

I know that we all like to think markets have memory and that squiggle on a chart 10 years ago will resonate with investors today - but that is just our mind playing pattern recognition tricks with us from our  days on the African plain. The market is much more like the character from the movie Memento - a lost man who has no short term memory and stumbles forward in the day looking for what is next.

The “Memento”-like nature of the markets is the reason I day trade. Just as with weather I have far more confidence in what will happen within the next hour than I do  the next year. I focus on the timeframe that gives me a modicum of control - but even that control is far from certain. Ask any New Yorker who has ever ducked into a subway on a nice sunny day only to pop out 15-20 minutes later to a torrential rainstorm that drenches all his clothes.

In my opinion forecasting is for fools but what’s even more foolish is to stick to your forecast in the face of all evidence to the contrary. If you stop trying to do the former you may be fortunate enough to avoid doing the later.

---