Reading the GPT Leaves
“You can have data without information, but you cannot have information without data.” — Daniel Keys Moran
“Predicting the future isn’t magic, it’s artificial intelligence.” — Dave Waters
ChatGPT changed everything.
Over the past six months, we have all come to the collective realization that artificial intelligence (AI) is going to reshape how we work and live. We need not look further than the parabolic spike in “chaptgpt” and “AI” search interest via Google Trends.
Like mainframe computers in the 1950s, PCs in the 80s and 90s, and the emergence of mobile/cloud computing in the most recent decade, AI has enormous potential to be the next world-changing technology platform.
Many of the early adopter organizations that embraced AI have experienced cost savings, increases in revenue, or both – the holy grail for any enterprise.
Transformative as it may be, AI is not without controversy. As this technology accelerates exponentially, questions remain regarding the displacement of human workers and the potential for nefarious and unethical use.
As the public discourse continues to seek answers to those emotionally charged topics, the world of finance is exploring a question of its own:
Will AI change the way we invest?
No matter where you look these days, there are signs that AI is infiltrating financial markets and perhaps one day, our portfolios.
J.P. Morgan recently unveiled an AI-powered tool designed to uncover trading signals by assessing the tone of Federal Reserve speeches. The CFA Institute has created a Data Science for Investment Professionals Certificate program, distinct from its flagship Chartered Financial Analyst, or CFA, designation. Even well-known quantitative asset manager AQR has been exploring the potential role of AI in its investment process.
Investors are already inundated with ways to trade the AI theme. Several ETFs exist that aim to own stocks of companies that derive a meaningful portion of their revenues from AI and Robotics-related activities. The challenge here is that not many “pure play” publicly traded AI companies exist today, so you end up getting a fair amount of exposure to other – perhaps unintended – factors. We’ve written on the pitfalls of thematic investing before. Thus far, these products have done a better job at gathering assets than providing market-beating returns.
But let’s take it a step further. What if, instead of investing in AI-related companies, investors leveraged the AI tools themselves to identify winning stocks or time the market?
In other words, what if ChatGPT could help us read the tea leaves?
Turns out there are already a few ETFs that attempt to do just that. Unfortunately for investors, the results here have been equally as disappointing in the relatively short span of time they have existed.
More like artificial lack-of-intelligence, am I right? (I promise that will be my only dad joke in this post.)
As for asking ChatGPT for hot stock tips yourself: don’t hold your breath. The chatbot (today) lacks the ability to process real-time market data. Not to mention is has been specifically designed not to give investment advice.
That doesn’t necessarily mean generative AI and large language models (LLMs) like ChatGPT can’t be useful arrows in an investment manager’s quiver, as Institutional Investor recently discussed with Bryan Kelly, AQR’s head of machine learning and a finance professor at Yale School of Management:
Put simply, generative AI tools like ChatGPT have allowed portfolio managers to process news or other financial documents more efficiently. “We capture the meaning from a whole bunch of different financial-related text documents, and then we intercept these document-level representations,” Kelly said. “We’ll pull them outside the GPT and plug those representations into our models, [which have] things that we care about on the finance and portfolio management side.”
The pursuit of improved performance and/or better risk management from AI-informed investing is intriguing, no doubt. Some may indeed discover ways to augment existing systematic or discretionary processes in such a way that new avenues for “alpha” generation are unlocked. Yet a strong counterargument can be made that the incorporation of AI into statistical and quantitative investment strategies will have the ultimate effect of making markets more efficient, not less.
Some of the brightest minds on Wall Street will continue to tinker and experiment with ways to use AI to their advantage. Whether the investment industry experiences the same AI revolution as other businesses remains to be seen. Financial markets are complex, adaptive systems from which even the most advanced models may struggle to separate the signal from the noise. As the Wall Street Journal notes:
“Unlike languages, markets can change quickly—companies alter strategies, new leaders make radical decisions and economic and political environments shift abruptly—making it harder to make trades using models reliant on historic, long-term data trends.”
For now, the best form of intelligence for successful investment outcomes will continue to be good, old-fashioned human emotional intelligence.
This is intended for informational purposes only and should not be construed as personalized investment advice. Historical performance results for investment indices, benchmarks, and/or categories have been provided for general informational/comparison purposes only, and generally do not reflect the deduction of transaction and/or custodial charges, the deduction of an investment management fee, nor the impact of taxes, the incurrence of which would have the effect of decreasing historical performance results.