An (almost) 100% accurate forecast by Paul King

Paul King of writes:

If I was a meteorologist, people would definitely hate my weather forecast. It would look something like this:

Weather today in Middlebury Vermont as of 8:54 AM EST is cloudy. Rain so far today is 0.5 inches, snow or other precipitation is 0.25 inches. Temperature is 25 degrees Fahrenheit. Five day forecast is from 0 to 10 inches of rain, 0-24 inches of snow, temperatures between -5 degrees F and 65 degrees F.

This forecast is a) all made up, and b) would be based on historical weather for this day in history plus 5 standard deviations from average (just to be on the safe side). How useful is this as a forecast to determine whether you should wear snow boots, take an umbrella, or if it's safe to swap your snow tires for summer tires? Not at all. But neither is an absolute prediction that's wrong most of the time....

What does this all have to do with trading? Well I go about trading the same way that I would construct a weather "forecast". All we know is what price is doing right now, what it's done in the past, and then expect it to do something similar in the future based on current estimated volatility. This tells you the likely range of price of a move (but not the direction).

Picking the next "big winner" is not effective, or even possible for a chaotic system like a market of stocks. So if stock-picking isn't the answer, what is it that makes money? Having a good exit strategy that keeps losers small relative to winners, and position-sizing to prevent you exceeding your maximum desired draw down. That's what makes the money.

Then, assuming you trade liquid instruments, it doesn't really matter what happens next - you'll capture some part of things that trend and keep your losers as small as possible. Unlike weather prediction, it doesn't matter if you're wrong more than you're right as long as the size of the winners is bigger than the losers overall and you can tolerate the length and duration of your draw-down periods.

Luckily for meteorologists there is no formal feedback process that tells them how good (or bad) a job they are doing unlike trading where your monthly brokerage statement doesn't lie.

What exactly does an 50% chance of rain mean - you should take half an umbrella out with you?

Read more about Paul King here: