[Part 5] You have to think anyway, so why not think big?

You have to think anyway, so why not think big?
– Donald Trump

Part 1
Part 2
Part 3
Part 4

Over the past 60 years, computers have become a billion times faster.

Computers are so fast, that they can now learn how to drive a car just by observing people drive cars.

A team of computer programmers doesn’t have to write millions of lines of code to cover every scenario the car will run into.

It will drive in the dark. It will drive in the snow. It will understand hand gestures from traffic cops perfectly. It will drive 10x safer than you.

All you have to do is feed enough data to its cameras and sensors.

The same is true with the stock market.

If you have a fast enough computer, it can use machine-learning algorithms to find the best mix of indicators to beat the stock market.

No programming required.

The key is having a fast enough computer.

A modern desktop is about fast enough to find which trading rules work best on a small set of stocks or futures.

But in order to be accurate with this kind of data mining, you have to include delisted stocks – companies that have gone bankrupt or got bought out.

If you don’t, then the strategy will be as worthless as a four-card flush.

In preparation of jumping head first into supercomputing and stock forecasting, we ended up buying a mountain of data that includes delisted stocks.

There’s 46,337 total instruments.

That’s way too much for any desktop computer to handle.

But with thousands or tens of thousands of computers at your disposal – all working on a solution — suddenly the task doesn’t seem all that difficult.

Dan Sullivan always talks about thinking ten times bigger to unleash your inner creativity and tear away the proverbial box you’re stuck in.

Why not think thousands of times bigger?

So not only will we be trying to create something like Shazam for stock price forecasting, but we can boost results even further by allowing the supercomputer to discover new formulas – BY ITSELF.

And of course, if you take this kind of thinking to its logical conclusion, you can also use the technology to forecast stocks globally…at every timeframe from tick by tick to daily data.

If computing power is no longer a constraint, then what else can be accomplished?

Here’s to punching holes through boxes. Cheers!

Click here to view Part 6

Trade smart,

Dan