Google’s Robot Traders Will Conquer Wall Street
- Four out of five of the world’s largest funds use automated decision-making.
- Google’s new supercomputer is more sophisticated than anyone expected.
- With the rise of Big Data, robots will be the asset managers of the future.
In the late ’80s, I took an eighth grade field trip to the New York Stock Exchange.
It felt like hallowed ground. Like walking into center field at Yankee Stadium, praying at the Vatican or standing on the 50-yard line at Lambeau Field.
I remember almost being knocked over by the hordes of shouting traders who were moving quickly between trading posts underneath the vaulted ceilings.
Giant TVs hung in clusters on the walls, displaying stock quotes and news headlines. The floor was littered with paper trading tickets.
This was the stock exchange’s heyday. It was the epicenter of Wall Street.
That’s because 30 years ago, if you wanted to buy shares of IBM, you had to dial up your broker at Merrill Lynch.
He would then call down to his broker on the floor of the stock exchange, and that broker would run to the post where IBM was traded and execute the order.
With all those hands in the cookie jar, trading was expensive.
The average retail investor paid a $45 commission for a simple stock trade. Larger orders would cost hundreds, sometimes thousands of dollars.
If you were a frequent trader with a small account, the brokerage fees ate away your profits.
But then the internet arrived.
Information was democratized. Investors could directly access the exchange.
Now, floor traders have been replaced with internet communication. Last year, I visited the New York Stock Exchange floor for an interview with a news network, and there were more cameras than traders.
Once humans were removed from the equation, trading commissions began the slow march to zero.
It’s great for the individual investor … and not so great for E-Trade, Charles Schwab and TD Ameritrade, which all lost a revenue stream recently when they offered $0 commissions.
But this was the first phase of technology’s Wall Street takeover. The next wave of tech aims to get rid of the humans who make the buy and sell decisions, and you need to stay in the loop on this.
Robots Are Automating Investment — No Humans Required
In the old days, the decision-making process was purely a human endeavor.
Berkshire Hathaway CEO Warren Buffett is the perfect example. He reads and digests six different newspapers and a pile of Form 10-Qs while swigging Coca-Cola.
When software arrived in the ’80s, so did quantitative funds, which automated the decision-making process by following a set of investment rules.
For instance, a “quant” fund may be designed to buy stocks that offer a cheaper price-to-earnings multiple than their peers.
Quant funds use sophisticated software to scour the market for investment ideas — and then they alert a human about the opportunity.
These funds are wildly successful. Four out of five of the world’s largest funds are dedicated quant funds.
And now, as technology becomes more sophisticated, quant funds are eliminating humans altogether.
They’re letting the computers pick which investment ideas have merit. And they’re letting them execute the trade.
Now, even more automation is on the way…
Google’s New Supercomputer Is a Chess Prodigy
In the above scenario, you at least need humans to develop the strategies and program the machines.
However, a new branch of artificial intelligence called “machine learning” allows computers to learn from data by identifying patterns and creating their own strategies.
Google is a pioneer in this field, famously training its machine learning programs how to translate by feeding them the same books written in different languages. The computers identified patterns in the texts to become multilingual.
In 2017, Google also debuted AlphaZero, a computer that taught itself how to play chess after simply being given the rules.
After a few hours of training, it beat Stockfish, the best chess machine programmed with human strategies.
What was even more incredible about AlphaZero is that it looked like it was making mistakes, but these were really a more sophisticated strategy than anyone had imagined.
As you can see, Big Data is quickly revolutionizing the world.
Big Data Fuels the Market-Beating Strategies of the Future
With the rise of Big Data, I don’t believe it’s a stretch to say the biggest asset managers in the future will be artificially intelligent algorithms.
They’ll gather massive amounts of information from satellite photos, location data and social media to stay a step ahead of investors.
They can then slice and dice all this information into market-beating strategies.
This would reverse the investment process as we know it.
Typically, a human investor starts with a hypothesis such as: “Target’s earnings will be better than expected if credit cards show more spending at Target and satellite photos show full parking lots at Target.”
The human then feeds this data into the machine to see if his hypothesis can be validated.
A machine learning program does just the opposite.
It starts with the data and then “mines” the information for a hypothesis, determining the best market-beating strategies by figuring out what worked in the past.
And this presents a problem.
The 1 Big Problem With Robot Traders
In a perfect world where machines replace human investors, there would be no booms and busts. Capital would be allocated in a perfectly efficient manner.
And you could rest easy.
But if machines are all trying to discover the same ideas, they’ll wind up engaging in the same behavior as humans. They’ll crowd into one “next big thing” after another.
To keep us from these mistakes, machines will eventually need to quantify the irrationality of human investors and avoid it.
It’s a brand-new world that’s bringing lots of changes for the future. And that means investors need to keep up with breaking developments if they want to keep their edge.
I’ll definitely be keeping an eye on this trend. In the meantime, you can watch my new video on how the Big Data revolution is shaking up every industry.
Editor, Automatic Fortunes
P.S. My colleagues Jeff Yastine and Brian Christopher have put an incredible amount of time and money into developing their own trading algorithm. In Saturday’s Market Insights video, they’ll explain how they came up with the idea for their patented trading system, and how you can use it to score gains of 1,000% or more over and over again. So check your email Saturday morning to learn more about Jeff and Brian’s secret trading strategy!