The Third Wave of Investing

Big Bets on Artificial Intelligence and Data Science

“Greed clarifies, cuts through, and captures the essence of the evolutionary spirit.”

— Gordon Gekko, Wall Street, 1987

It is this evolutionary spirit that has been guiding the hyper-competitive world of finance for as long as anyone can remember. Brokers and hedge funds are on a constant hunt for the next big thing that will give them a competitive edge over others and ensure their survival. And this time around the big bet is on artificial intelligence (AI) and machine learning (ML).

The confluence of various factors make the time just right for the third wave. For one, recent times have seen an explosive growth in volume of data, with more than 90% of the data currently in existence created within the last few years alone. And to traders, data is as good as gold. At the same time rapid advances in cloud computing have also made storing and managing vast amounts data more accessible and affordable, thus radically lowering the barrier to entry and creating incentives for funds to invest in frontier technologies.

The most significant advance in the past decade has been our ability to make “sense” of all that data. Recent breakthroughs in artificial intelligence and data science have led to the development of new models that can now draw actionable insights from this heretofore unusable ‘data soup’. What’s more, the rise of open source deep learning AI platforms like Google Brain’s Tensorflow and Amazon’s Destiny (DSSTNE) have drastically reduced the time and effort required to build, implement, and maintain machine learning systems.

The arrival of the third wave should not be seen in isolation though. It is but a natural outcome of the ‘evolutionary spirit’ of Wall Street that has itself evolved over the course of time, peaking and ebbing through distinct waves of innovation in the finance sector.

The first wave of investing was about the rise and growth of discretionary funds. In the 1970s charismatic investors like Warren Buffett and George Soros used their innate human decision making and analytical abilities to find new opportunities and maintain a competitive edge in the market. Over time, the proliferation of information technologies like the Internet closed this gap. Opportunities and information have become more readily available and standardized. With everyone working from a similar base set of information and assumptions, there was a need for investment firms to develop a competitive edge.

That all changed in the second wave of investing in the early 2000s. Funds like Renaissance Technologies and WorldQuant* brought a great degree of mathematical rigor to investing and trading by hiring “Quants”. These Quants, or quantitative analysts, leveraged new statistical models and computer algorithms to build systematic strategies that generate consistent alphas (i.e., market advantages) across several markets and financial instruments.

Flash forward to today, things are changing yet again and the funds are catching on. Many are using deep learning artificial neural networks (ANN) to find and exploit complex nonlinear relationships in price movements. Unlike humans, ANNs’s can quickly and continuously adapt to shifts in market regime to provide cutting edge insights. What’s more, they don’t need to rest or take a coffee break.

That considered, a word of caution is warranted. As the saying goes, there are no free lunches, and on Wall Street there is certainly no holy grail of trading. Regardless of what trading method one uses — be it machine learning or more traditional methods — there will always be an infinite number of ways to fail.

Despite its burgeoning use in trading, there are still many questions around the applicability of deep learning algorithms on financial data. Though deep learning can effectively discover patterns in images; whether it can do the same for financial data remains shaky, if not unproven. What’s more, trading has always been a zero-sum game — it matters not how smart a company’s AI is, but only if it is better than the ones it is competing with. Also, as these AI and ML models become commodified, the edges they provide will slowly vanish and start getting priced into the markets themselves.

Hence, AI and ML models are not magic wands, but rather are extremely powerful and strategic weapons that need to be wielded by highly-trained, skillful professionals. Rising investments in AI and machine learning technology is driving non-linear demand for “quants” with these skillsets.

The third wave of investing is here. It’s the quiet hum of a server in a closet and the silent army of quants ramping up their Python and data science skills. As evolutionary pressure tightens its noose around Wall Street once again, the firms who successfully augment their humans with AI tools–and vice versa — will be the ones to thrive.

Ritabrata Bhattacharyya is a professional quantitative trader, an AI enthusiast and faculty member of WorldQuant University, an international not-for-profit dedicated to advancing global education.