How Does a Learning Machine Pick Stocks?


What’s the formula for success?

Go to bed smarter than when you woke up.

Being a learning machine has done wonders for those willing to do the work.

Jeff Bezos’, founder of, Inc. [NASDAQ:AMZN], and his concept of ‘Day 1’ comes to mind.

Like their focus on customers, Bezos wants Amazon employees to adopt a Day 1 mentality. Another day to innovate, invent, improve and get better.

In his 2017 letter to shareholders, Jeff wrote:

“Jeff, what does Day 2 look like?”

That’s a question I just got at our most recent all-hands meeting. I’ve been reminding people that it’s Day 1 for a couple of decades.

I work in an Amazon building named Day 1, and when I moved buildings, I took the name with me. I spend time thinking about this topic.

“Day 2 is stasis. Followed by irrelevance. Followed by excruciating, painful decline. Followed by death. And that is why it is always Day 1.”

Not only has this formula proven successful for entrepreneurs, but investment giants too.

Simple, but not easy

How can you turn thousands into billions?

You need two things. A group of high returning businesses you can understand, and a lifetime.

Warren Buffett has built his career on both of these.

Each day, like Bezos, has been a new day for Warren. Another day to learn about the energy industry, or consumers electronics.

For Buffett, reading has been the best teacher. Once when asked how to get smarter, Buffett held up stacks of paper and said, ‘Read 500 pages like this every week. That’s how knowledge builds up, like compound interest.

Simple right? But hardly easy.

It’s far easier to buy ‘the market’ and just sit on the couch watching Netflix waiting for retirement.

But what if it was easy? What if you had a learning machine pick the stocks for you, literally?

Well you can.

It’s called the EquBot, and it could teach us something about stock picking.

Far more than a model

In October last year, the EquBot finally debuted.

The bot is an artificial intelligent (AI) system, designed to pick stocks.

So it’s a quantitative model?

Not exactly.

We’ve had stock picking computers for years already. The EquBot is something different. Rather than picking stocks based on a set of metrics, it picks stocks by learning from its mistakes.

The system uses IBM’s Watson supercomputer technology to analyse more information than humanly possible.

While it’s restricted to the US market for now, the EquBot analyses 6,000 companies constantly. It reads every filing, annual report, articles…anything it can get its hands on.

Of course this information is worthless if the system cannot interpret what it will mean for share prices. And that’s why the AI system holds so much promise. 

It learns from its mistakes, and should be able to learn which stocks are worth buying.

Business Insider Australia wrote last year:

The ETF ranks investments based on their “probability of benefiting from current economic conditions, trends, and world- and company-specific events” and picks those with the best chance at outperformance.

And as I mentioned, you can invest in this learning machine right now. There’s an exchange traded fund (ETF) which mimics EquBot’s portfolio.

The ticker is AIEQ.

So what can we learn from this stock picking learning machine?

A whole new way to pick stocks?

Because there is an ETF tracking EquBot’s every move, we can see its holdings…daily.

What is the AI system holding right now?

At the moment, it’s holding 75 stocks.

The top 10 holdings, shown below, make up a little over 27% of the portfolio.

, EquBot holds a lot of tech stocks. Among the biggest are Google’s Alphabet Inc. [NASDAQ:GOOG], Amazon, NVIDIA Corp. [NASDAQ:NVDA] and Facebook Inc. [NASDAQ:FB]. 08-03-2018

Source: EquBot ETF
[Click to open in new window]

As you can see, EquBot holds a lot of tech stocks.

Among the biggest are Google’s Alphabet Inc. [NASDAQ:GOOG], Amazon, NVIDIA Corp. [NASDAQ:NVDA] and Facebook Inc. [NASDAQ:FB].

So you might think the EquBot focuses on momentum. Meaning it buys stocks which have had great performance in the recent past.

However, the top 10 also includes an office management trust which has dropped more than 30% in the last 12 months. Can’t get more anti-momentum than that!

Whatever the EquBot is doing, it seems to be working.

So far the AI Powered Equity ETF has climbed 6.9% since inception. It beats the S&P 500 (the 500 largest US stocks) over the same time. But not by much.

As you would expect, it will take time for this stock picking machine to properly learn how to sort the amazing investments from the mediocre ones.

Of course, this could just as easily be a disappointing exercise.

What if the EquBot doesn’t perform well over the long term?

Until a couple of years go by, we won’t know that answer for sure.  But I’m excited to see what happens next.

And even if the EquBot does go down in flames, it doesn’t mean you can’t mimic its greatest quality — being a learning machine.

Kind regards,

Härje Ronngard,
Editor, Money Morning

Money Morning is Australia’s most outspoken financial news service. Your Money Morning editorial team are not afraid to tell it like it is. From calling out politicians to taking on the housing industry, our aim is to cut through the hype and BS to help you make sense of the stories that make a difference to your wealth. Whether you agree with us or not, you’ll find our common-sense, thought provoking arguments well worth a read.

Money Morning Australia is published by Port Phillip Publishing, an independent financial publisher based in Melbourne, Australia. As an Australian financial services license holder we are subject to the regulations and laws of Corporations Act and Financial Services Act.

Money Morning Australia