[Project 88 Part 7] Forecasting Apple…

Here are links to the first 6 blog posts of our crazy idea to forecast stock prices with a 2000-core supercomputer at Microsoft's secret datacenter:

If we were to try to predict the future of any stock with Project 88, it would be AAPL.

Its $220 billion in sales over the past 12 months rivals the GDP of countries like Vietnam, Portugal, and New Zealand.

While this image might not seem like much, it graphs the returns of every stock (after sifting through nearly 50,000 of them) that's ever matched Apple's pattern from August 30th, 2016.

What happens AFTER the pattern is detected (AAPL… August 30th, 2016)

While this is just a single day's projection of a single stock (and unadjusted for volatility), it's getting close to full blown production.

The giant hurdle is speed.

It takes about 6 seconds to make this projection…if we multiply that out over a ten year period and 3000 stocks, it would take at least 1.5 years on a typical desktop to run a single backtest.

Lucky for us, we already tackled distributed computing in the cloud (go team!).

It's like having 2000 workers at your beck and call whenever you need them 24/7.

They never call in sick, never take smoke breaks, and can continuously work as long as you need them to.

If one “dies”, another takes over.

All you have to do is keep feeding the meter.

No money…well, then you have a problem.

The hypothesis of the project is that we can forecast stock prices based on past price patterns, and only jump into those stocks projected to go up the most.

It's like taking a snapshot of the beach, or a train, and telling the computer to create a 5-second video of what it thinks should happen next.

Which researchers at MIT recently did:

ML Prediction

 

Right now, simple backtesting can tell you what the probability of a stock making money is.For example, your odds of winning at most casinos when you play basic strategy blackjack are around 46%.

Like the lottery, if you play long enough, you'll lose all your money with 100% certainty.

Forecasting stock prices not only tells you your odds of winning, but how much money you'll likely make.

It's like playing blackjack and only betting when you're dealt an ace…the best starting hand possible.

That's not gambling, that's speculating.

I don't know about you, but I'll play at that casino all day.

By the way, we've had 311 responses to the name the supercomputer contest.

Some were good.

However, many had about as much thought put in as a goldfish with ADHD.

So I'm going to leave the contest open.

I suggest you don't use any of these:

There will be no placement trophies for any snowflakes that snuck onto this list.I will however be giving away something cool to the winner (it's a secret).

So go ahead and enter your suggestion if you haven't done so already. The deadline is extended to this Friday at midnight:

Click here to view Part 8

Trade smart,

Dan