This AI-powered stock portfolio beat the S&P 500 and left market pros in the dust

© MarketWatch photo illustration/iStockphoto


Load Error

In the ever-evolving landscape of finance, artificial intelligence (AI) has been making steady strides. One development that recently caught my attention was the performance of OpenAI’s ChatGPT in stock trading.

The advanced AI model selected a portfolio that outperformed the S&P 500 as well as some of the most popular investment funds in the U.K. Over an eight-week period, the AI-powered portfolio of 38 stocks gained 4.9% while the S&P 500 rose 3%, and top-10 investment funds in the U.K. saw an average loss of 0.8%.

ChatGPT’s stock portfolio included Microsoft Netflix and Walmart using criteria such as low levels of debt and a track record of growth. Without a human hand to guide it, the AI analyzed tons of data, made its choices, and outpaced the market, leaving seasoned fund managers and algorithm-based indices in its wake.

ChatGPT’s success is not just about the impressive gains it made, but also about the unique advantages that AI brings to the complex world of trading. AI systems can process vast amounts of data at lightning-fast speed, analyze patterns and trends without bias, and operate continuously without the need for sleep or coffee breaks.

So, does this mean that AI will soon be able to replace human traders? To answer that question, I’ve reached out to several industry professionals:

Matthew Tuttle, CEO/CIO at Tuttle Capital Management, believes that AI could eventually replace human traders. He praises AI’s ability to remove emotions from trading and process more information than human traders, but he also points out that trading is about adapting, and emotions can sometimes keep traders out of trouble.

On the other hand, Giuseppe Sette, co-founder and president of Toggle AI, claims that AI will not replace humans but will instead empower traders to ask more in-depth questions. According to him, AI’s biggest strength is its ability to pull together large amounts of information across different fields to discover hidden trends in the market.

Global Predictions’ CEO and co-founder Alexander Harmsen points out that algorithms have already replaced human traders to a large extent, and he sees AI, like ChatGPT, as a natural extension of this trend. AI’s ability to analyze text data is particularly useful in trading, as it can quickly sift through SEC filings, company financial reports, and news-based sentiment analysis.

Safeguards are necessary to prevent AI from causing large market moves. 

However, these experts also acknowledge the drawbacks of AI in trading — from the limitations in modeling and mathematical capabilities to the fact that AI can sometimes “hallucinate” or generate false information — and suggest that safeguards are necessary to prevent AI from causing large market moves. 

Experts make a good point. Before you start figuring out how to get your own ChatGPT trade-bot running, it’s important to recognize that we’re still in the early stages of this technological revolution. The financial markets are inherently unpredictable, influenced by myriad factors ranging from political upheavals and natural disasters to shifts in investor sentiment. This poses a significant challenge to AI, which, despite its advanced capabilities, may struggle to factor in such unpredictable variables.

AI falters when faced with new, unfamiliar market conditions.

One of AI’s key limitations is the risk of overfitting. This occurs when an AI model becomes so attuned to the historical data it has been trained on that it falters when faced with new, unfamiliar market conditions. It’s akin to studying past exam papers and expecting the same questions to pop up on the final exam. When the actual test contains questions you’ve never seen before, you’re stumped.

Moreover, while the absence of emotional bias can be an advantage, it also means that AI lacks human intuition and creativity. When facing an unpredictable situation, an experienced trader can pick up on subtle market signals and respond accordingly with an innovative solution. This is a skill that AI, in its current state, struggles to replicate.

Experts I’ve interviewed are clearly aware of said limitations, as well as possible advantages. What about an average investor?

A recent survey conducted by Finder UK has revealed that 8% of U.K. adults have already used ChatGPT for financial advice, and another 19% said they would consider doing so. This indicates a budding curiosity and a willingness to embrace a new technological advancement.

However, the survey has also revealed a significant apprehension about using AI for financial decisions. A sizable 35% of those surveyed said they would not consider using a chatbot like ChatGPT to guide their financial choices. The resistance could be attributed to many factors, such as the lack of understanding how AI works, concerns about its accuracy, or simply an unwillingness to trust a machine with such important decisions.

It’s important to ensure that AI conducts trading in transparent, accountable and fair ways.

It’s only a matter of time before we see more and more AI models involved in trading. As AI becomes more integrated into trading, it will be handling vast amounts of sensitive data.

Regulations are needed to ensure the privacy and security of this data and to prevent misuse. This includes guidelines on how data should be stored and processed, as well as measures against cyber-threats.

It’s also important to ensure that AI conducts trading in transparent, accountable and fair ways. This includes addressing issues related to the accuracy of AI predictions and the risk of overfitting. Regulations could mandate the use of robust testing and validation methods for AI models before they are used in live trading. This would guarantee that the AI models are reliable and that they perform as expected.

Issues related to market stability should also be addressed. As AI becomes more prevalent in trading, it could cause large market moves, particularly if many AI systems make similar trading decisions at the same time. Regulations could help mitigate risk by limiting the size of trades AI systems can make.

Striking a balance between leveraging the advantages of AI and addressing its limitations will pave the way for a future where AI and human traders can coexist, empowering investors with more robust and insightful financial decisions.

More: This is how AI conquers humans: trapping us in its robotic ‘echo chamber’ with no escape

Also read: Will AI save money-losing tech stocks from the short-sellers?

Continue Reading