A.I. seems like the road to El Dorado. Or your own money printing machine. Using the power of A.I. to create incredible portfolio performance, scanning the markets day and night for tradeable information, and making billions in return…. After I watched a brief YouTube video on A.I., I’m ready to build the most revolutionary trading algorithm of all time…
Unfortunately for me (and nearly everyone else) I have no practical chance of being able to pull it off. However, if you’d like to muse with me while reading a (hopefully) amusing post you’ll need three things.
The three key components of this money machine are:
Knowledge of graph theory and its applications.
A whole lot of data. Primarily on people who invest.
An excellent connection to stock exchanges.
Graph Theory:
If you are like me, you may have heard of graph theory, but never had to understand it. Being a history major at an engineering school, I never had to take discrete math either (the foundation for graph theory), but I have certainly heard horror stories from the programmers and engineers who did. Graph theory is perhaps the single greatest common denominator driving the largest profits in US companies right now. While I was doing a quick google, I learned that graph theory was actively being used to determine what I might be googling. When I went to YouTube to learn more, I found out it was being deployed to send me more videos to watch and even unrelated videos that people like me might want to watch. I’ve learned that Netflix, Spotify, Amazon, Meta, and hundreds of other companies use the basis of graph theory to optimize their suggestions, pick advertisements I might like, and predict what I might want to do or be interested in. A very simplistic way of describing it; these companies identify the relationship between what I search for and what I click on, over time creating a map of people who search for similar things and click on similar things. By applying graph theory they can make accurate predictions from the huge amounts of seemingly unorganized data. So how could this be applied to investment selection? First let’s talk about the data.
The Data:
The word “data” nowadays seems to have a negative connotation around it. Growing up I was warned not to put my address on Facebook and definitely not my birthday, but thanks to graph theory Facebook has correctly guessed all the classmates I went to high school with, what year I graduated, how old I am, and who else might know or influence my decisions. All this was guessed before they updated their user agreement to include locational data, internet cache, advertisement trackers, messenger permissions, Facebook marketplace, scroll monitoring and over hundreds of pages of privacy policy. At some point they said, “how can we make money with this?”, and thanks to graph theory, they found a way to provide an even more meaningful prediction of our behavior and interests. Suppose you know how to get this data, and you know how to use graph theory to make predictions. What could a hedge fund do with it? Could they make educated guesses on the investment ideas people might be interested in? Well, if we use Facebook as an example of what can be done, the hedge fund could make a behavior model for all the users in the data. You could build groups based on centers of influence. Then try to identify some of the most influential users and predict how many other users like them might be interested in their ideas. You could try to predict how quickly the “influencer” impacts their “followers” by tying it to investment data and watching for spikes in referenced ideas with a plan to arbitrage the difference between your model and real outcome in the market.
Most people won’t be able to do this, but one of the benefits of the whole A.I. craze is that companies have started creating the tools that allow nearly anyone to do exactly what I mentioned above. So this is becoming more and more a possibility for people like you and me, assuming you can get the data and no laws prevent RIA’s (registered investment advisors) from using this strategy. However, if it pans out as planned, it would seem that there could be a rapid deployment of this strategy over the coming decade. Which brings us to our next key for success with this strategy, speed.
Speed:
As competition picks up you’ve got to be the fastest one to model the data, make the prediction and place the trade. You’ll need the absolute highest processing power to analyze and crunch the data before your competition does. If your processor is a nano second slower or your fiber optic internet line a block farther away from the exchange, the opportunity to make a profit begins to evaporate. You’ve got to have the best connection to the trading floor, and if you really want to optimize your strategy you’ve got to factor in the lag that the speed of light causes relative to your location to the exchange and your competitors. If they can get the same data, then you’ve got to assume they can come to a similar conclusion and are trying to beat you to it. At this point human decisions are incredibly too slow, so you’ve got to automate most of your trading in order to beat the competition to market and close the position before it's too late.
Whew! You did it, you mastered the markets, conquered the hedge funds, beat the high frequency traders and are on your way to find El Dorado, or…
Did you just open pandora’s box?
Pandora’s Box
A few months or years or maybe only days later your competition has calculated how quickly you are able to make a trade based on the available data and they factor their own graph theory to account for your superior speed. Suddenly your profit-making machine has begun to leak gains back into the market, and you’ve got to get back to work at a new game. This time you’re not in a footrace to the finish line, but a game of quantum chess. Now you’ve got to build a secondary data network designed to predict what predictions your competitor is making about your predictions, and then place an optimal trade based on the likelihood they won’t change their predictions based on your ability to anticipate theirs…
Oh man… If all this sounds like too much, I’m right there with you.
At Blacor we don’t just throw your money into a mind mimicking machine, we identify the companies and investments that do more than generate investor attention. We carry our conviction about the markets to our clients, explain why it may make sense for their portfolio and hold to the principles behind our recommendation through the ups and downs of the market.
We believe why you invest is just as important as what you invest in. Which is why we use a combination of advisory services and investment selection to build portfolios that reflect each client’s unique values. If you are interested in learning what that looks like, feel free to contact us and schedule a time to review your finances. We also offer our free investor risk tool, to help investors identify their own risk/reward profile and enable investors to ask, “Are you getting paid enough for the risks you take?”
References:
Graph Theory: https://en.wikipedia.org/wiki/Graph_theory
Nations, Scott. A History of the United States in Five Crashes: Stock Market Meltdowns That Defined a Nation. William Morrow, 2017.
Securities offered through International Assets Advisory, LLC (“IAA”) – Member FINRA/SIPC. Advisory services offered through International Assets Investment Management, LLC (“IAIM”) –SEC RIA.
Blacor Investments is unaffiliated with IAA and IAIM. The information provided is based on carefully selected sources, believed to be reliable, but whose accuracy or completeness cannot be guaranteed. All information and expressions of opinions are subject to change without notice and are those of Blacor Investments.
Past performance may not be indicative of future results.
This material is for informational purposes only and is not a solicitation or recommendation that any particular investor should purchase or sell any particular security.