Khaleej Times - AI is good as a research assistant, not as an adviser

Artificial intelligence isn’t new. But its most recent manifestation — and we are talking about generative AI and its popular subset: conversational AI such as ChatGPT — have piqued the interests of people from all backgrounds and even prodded many to seek investment advice.

While ChatGPT doesn’t readily give a recommendation on a portfolio of stocks to invest in, suggesting its limitations “as an AI language model it cannot provide specific investment advice or recommend individual stocks”, the platform does offer suggestive stocks when asked with precise conditions.

On May 9 this year, we asked ChatGPT for its recommendation on 10 stocks that were trading at low valuations on the Indian bourses, and 10 stocks of undervalued companies listed on the UAE bourses. Almost 12 weeks on, both the portfolio of stocks have generated positive returns.

The portfolio of 10 Indian stocks were a mix of large and mid cap companies and generated 16.85 per cent returns during May 9 to July 31 period, compared with 7.5 per cent on BSE Sensex and 7.99 per cent on NSE’s Nifty during the same period. Only two stocks in this portfolio are in the red, and the rest eight are in the black.

The portfolio of UAE stocks is up 3.58 per cent during the same period. However, of the 10 stocks recommended by ChatGPT, two are not being traded on the bourses any longer.

Experiments like these have been conducted in Western markets too over the last few months and most of them have yielded similar results. More sophisticated AI is powering investment platforms of firms across the world. A majority of Wall Street firms are trying to develop their in-house AI tools for serving their client base. As for retail investors, many are still apprehensive on how and if at all they should seek investment advice from freely available platforms like ChatGPT.

What do finance experts say?

Software such as ChatGPT cannot, at least at the moment, overshadow personal finance advisers, says Vijay Valecha, CIO, Century Financial. “Users need to understand the significant limitations of ChatGPT for investment advice. One of the foremost is the knowledge gap in providing real-time updates. Typically, these AI tools rely on large data sets for training and modelling. As with any computer program, the data is fed and modelled by a particular date. ChatGPT currently prompts the user to inform of its major limitation of the old knowledge base as of September 2021. While new versions will solve this problem, the knowledge part in an investment world must be real-time, iterative, and without gaps.”

He adds: “The world of financial investments has always relied on human feelings and intuition. Major market veteran investors and traders have repeatedly highlighted the human aspect and importance in the financial investment space. Ultimately, human fear, greed, emotions, grit, intuition, and risk-taking have always dominated the global financial market landscape.”

Noor Sweid, managing partner at Global Ventures, agrees and noted in one of her blogs: “AI’s output is entirely dependent on the dataset it is pulling insights from. That data could be outdated at the very least, completely biased at most and in between, unable to keep pace with the rapidly shifting and dynamic world of technology and innovation. Despite its transformative abilities, AI will not replace the power of human intuition, expertise and the value of personal connections forged in the world of investing.”

Agrees Sonam Srivastava, CEO and Founder of Wright Research. “AI provides powerful tools for analysis, decision-making, and automation. However, personal finance involves human aspects such as understanding personal goals, emotional responses to financial situations, and the ability to interpret complex financial information, where human advisors have an edge. The future will likely involve a blend of AI technologies and human advisors for the optimal financial advisory experience.”

How are industry players using the technology?

The number of firms vying to tap artificial intelligence-powered solutions for investments is rising fast. BlackRock, for instance, has partnered with and invested in Clarity AI to integrate its in-house platform Aladdin system for investment professionals. Others, such as Goldman Sachs and JP Morgan, have also started harnessing AI algorithms for their prop system and retail advice.

Sonam Srivastava, CEO and Founder of Wright Research, is among the industry players who is leveraging the power of artificial intelligence and machine learning to make gains in the stock markets. An expert in quant finance and algorithms, she constructed her portfolio in 2019 and over the years has conducted rigorous testing. On beating the market consistently over different time scales, her firm has recently launched Portfolio Management Service in India. “We leverage technology, expertise, and AI to make investing smarter and more intuitive. We use AI to analyse large amounts of data quickly and accurately, eliminating human biases and providing a systematic strategy. The AI helps us dissect each stock into numerous factors such as momentum, value, growth, and quality. We use this information to anticipate their behaviour in various market regimes. Our data-driven approach allows for quick analysis, automated risk management, making investment outcomes more consistent, reliable, and predictable.”

During the testing stage, Wright Research’s portfolio primarily focused on Indian equities. “Our aim was to tailor solutions for unique opportunities and risks within the Indian market. We included some debt instruments in 2019 when the market was volatile. The timeline of our portfolio varied as we evaluated performance across multiple timescales to gain a robust understanding of our strategies. As for the returns, our strategies have consistently outperformed the benchmark, delivering a significant advantage over traditional passive funds,” Srivastava says.

How can ChatGPT and similar software help investors?

While ChatGPT and other AI chatbots may not be the best tools for building a portfolio, many of these certainly can be useful to investors in myriad ways.

Use it to demystify concepts and educate yourself: The technology can be a helpful tool for learning. Retail investors could use it to look up definitions of financial terms they may be unfamiliar with or for gathering data when researching a company that they are considering investing in.

Use it as a research assistant: It's tricky asking ChatGPT for information or an explanation that requires subjective reasoning. But AI could serve well if used as a research assistant to gather initial information on the health of companies you consider investing in. For instance, you can ask ChatGPT to name undervalued stocks in a market, or stocks that give more than X per cent dividends to begin with. Even here though, remember, ChatGPT doesn't always draw its answers from the most up-to-date sources. It has information until September 2021. You can, however, build on this information and alongside use a number of free resources such as Google's TensorFlow, PyTorch, and RapidMiner to build basic machine learning models for further research.

Besides, there are paid software as well that provide more advanced capabilities. “It's important to note that effectively leveraging these tools requires a strong understanding of finance, statistics, and machine learning,” said Srivastava.

Artificial Intelligence has the potential to change investing, and it is already doing so in many ways. Free software such as the ChatGPT can become powerful tools for the retail investors, only if used responsibly.