Retail investors turn to ChatGPT, AI tools as robo-advisory companions

OpenAI’s artificial intelligence large language model ChatGPT can provide you with investment advice, much like the one in 10 small investors who have turned the chatbot into their “potentially profitable” stock selector.

Through AI, anyone can select stocks, monitor performance, and do what was previously limited to banks: run market analysis. 

The so-called “robo-advisory market,” which includes fintechs, banks, and wealth managers who have keyed in their advice into the algorithm, has grown exponentially over the past 12 months. Data firm Research and Markets projects revenues will grow by roughly 600% to reach $470.91 billion by 2029, up from $61.75 billion last year.

Analysts against paywalls are using AI for trading

Jeremy Leung, who spent nearly two decades analyzing companies for Swiss bank UBS, began using ChatGPT after losing his job earlier this year. He told Reuters that he now applies the chatbot’s responses to manage his multi-asset portfolio.

“I no longer have the luxury of a Bloomberg terminal, or those kinds of market-data services which are very expensive,” he said. “Even the simple ChatGPT tool can replicate a lot of the workflows I used to do.” 

Even though ChatGPT can read the internet, no AI chatbot can access subscription data behind paywalls like Bloomberg’s stock aggregator, so when using LLMs, you might miss some important investment details.

Leung uses prompts such as “assume you’re a short analyst, what is the short thesis for this stock?” or “use only credible sources, such as SEC filings.” He believes context improves the quality of answers, but admits they come with risks. 

“If people get comfortable investing using AI and they’re making money, they may not be able to manage in a crisis or downturn,” he said.

According to a survey done by eToro, 13% out of 11,000 retail investors use AI LLMs like ChatGPT or Gemini for investments, while around half said they would consider them. Another one from UK-based comparative site Finder saw 40% of respondents say they prefer chatbots to agencies when it comes to personal financial advice. 

When Finder asked ChatGPT in March 2023 to select stocks, it produced a list including Nvidia, Amazon, Procter & Gamble, Walmart, alongside 30 other entities. The portfolio has gained nearly 55% since, beating the average return of the UK’s 10 most popular funds by 19 percentage points.

Do not blindly rely on AI

“AI models can be brilliant,” said Dan Moczulski, UK managing director at eToro. “The risk comes when people treat generic models like ChatGPT or Gemini as crystal balls.” 

He is warning investors against using tools that are not specifically trained on markets, noting that general models can misquote figures, misdate events, and issue outdated price trends.

When deployed as advanced search engines, LLMs can find information quickly or issue good tips on budgeting and credit. However, when users seek precise answers to complex questions about money, health, or law, some researchers find that they come up short.

“It is quite dangerous to seek advice [from AI] of pretty much any sort, legal, financial, or medical,” said Andrew Lo, director of the Laboratory for Financial Engineering at MIT Sloan School of Management. “All three of those areas have very large dangers if they’re not done well.”

Chatbot predicted S&P 500 index gains, but is not entirely reliable 

The US stock market S&P 500 index is up 13% in the last 12 months, while the pan-European STOXX 600 is up almost 10%. The rally has made AI-powered selections look prescient, but according to naysayers, chatbots have only made the right picks because the market is doing well overall.

According to Lo’s research paper on generative AI in finance, ChatGPT may recommend Microsoft as a strong investment option. A human advisor might also do so, but would be expected to disclose that Microsoft has invested more than $13 billion in OpenAI, taking up almost half of ChatGPT’s parent company’s profits. 

The chatbot does not automatically reveal that connection, which is a clear conflict of interest.

“There are winners and losers every time you introduce new technologies,” Lo said. “The early adopters are the ones that may end up making a lot of the mistakes, but they’re also the ones more likely to innovate with the technology, as opposed to pushing against it.”

Your crypto news deserves attention - KEY Difference Wire puts you on 250+ top sites