How Big Data Can Unseat Big Players in the Stock Market

Have you heard of Fibonacci trading? It’s an investment strategy based on the Fibonacci sequence. A stock trader who favors the Fibonacci ratios ” a high volatility or low volatility Fibonacci trader ”  will sell or hold their position based on the ratios. The interesting thing about this strategy is the way in which it mirrors nature, which is an anomaly. For whatever reason, nature decided to organize structures according to the pattern the Fibonacci sequence describes. In turn, traders can base their strategy on a mathematical anomaly that corresponds with nature.

Here’s the thing: before day traders could take advantage of advanced automated trading software, a trader who tried to manually employ the Fibonacci ratios was at the mercy of their own emotions. At times when a Fibonacci-based strategy is working, manual day traders can fall prey to either the gambler’s fallacy or the hot-hand fallacy. They can decide it’s time to change strategies because of the basic logical errors to which humans are prone. Now, automated trading removes the emotional irrationality and makes it possible for small-fish traders to employ multiple strategies at the same time.

In short, humans can tend to have a bias towards a strategy like Fibonacci trading, as well as towards irrational moves, a bias that automated, algorithmic trading can remove. And the introduction of AI to investing could soon take this to another level.

In the past, high-frequency algorithmic trading was the domain of pension funds, mutual funds, and other investment firms with access to supercomputers. Now, any investors can potentially make high-frequency trades because they can access big data and the software to analyze and execute trades.

This data economy isn’t neutral, says Fortune’s David Z. Morris, it gives atomized workers and small entrepreneurs a huge leg up. They can increasingly mimic the informed decision-making of firms that use legions of humans to process data for a small cadre of leaders.

Big data for investment is no longer just a big firms’ game. Investors still need to know the ins and outs of the stock market, but as New Jersey Institute of Technology points out, when it comes to big data and competition, Sophisticated analytics can substantially improve decision-making. According to NJIT’s researchers, 13,000 exabytes of the digital universe will have big data value by 2020.

As we’ve seen lately, even the president’s tweets possess big data value for investors. When Trump tweets, algorithms analyze the nature of the tweet and make trades accordingly. There’s a stock trading bot powered by Trump tweets that uses sentiment analysis to determine whether his opinions are positive or negative toward publicly traded companies.

Surprise! The bot’s creator, Max Braun, turns around and donates the profits from its automated investments to Planned Parenthood, the American Civil Liberties Union, and the National Resources Defense Council, respectively. According to Braun, The simulated fund has an annualized return of about 59% since inception. He wrote the code on a flight to Europe, and it’s open source.

Now, anyone who knows how to code can write an investment sim that scans the internet for applicable data and manages a fund. Imagine applying machine learning to this endeavor. This would work particularly well for high frequency trading because there’s a ton of data for the machine to work with. An individual with coding skills and knowledge of the stock market can compete with any firm because AI can do all of the legwork that legions of employees used to do.

BlackRock, the world’s largest asset management firm, uses AI to play a research role, which includes social media monitoring and search engine monitoring. The firm’s AI also looks for relationships between securities that are hard for humans to spot. Then, human managers make the investment decisions. AI Powered Equity ETF, on the other hand, actually selects holdings based on data from 6,000 companies, as well as one million articles and filings per day.

The latter uses EquBot, an investment bot that runs on IBM‘s Watson, to make decisions. An ETF is an exchange-traded fund that tracks an index, commodities, bonds or a basket of securities. The intriguing possibility is that day traders who know how to code their own bots could create AIs that analyze ETFs, mutual funds, individual stocks, and market indicators. If, for example, a custom AI were to analyze AI Powered Equity ETF, which is non-diversified, alongside a diversified mutual fund, such a program could make for a highly effective hedge machine.

Overall, big data combined with AI has the potential to revolutionize the stock exchange and truly make data the new money. Increasingly, big firms will have to look out for young investors for whom coding is a second language.