The next generation of robo-advisers might veer into the world of short-term trading and stock-picking.
A recent study by researchers from the University of Chicago has shown that AI, specifically large language models that have been trained on vast amounts of text and can generate natural-sounding responses, can analyze financial statements as well as, if not better than, human analysts. But before you ask ChatGPT to suggest your next trade, there’s more to the story.
The researchers provided GPT-4 with financial statements – with company names and identifying details removed – and asked it to predict the direction of future earnings. The results showed that GPT-4 outperformed human analysts with more accurate predictions of earnings changes. Moreover, trading strategies based on GPT-4′s predictions yielded higher returns than those based on other models.
So, what does this mean for the future of financial analysis? In the short term, AI’s ability to process vast amounts of data and generate insights quickly and accurately is undeniably going to have an impact on the financial services industry. Ultimately, for investors, AI will strengthen the argument for a passive investing approach.
AI can analyze trends, compute key financial ratios, and provide narrative insights about a company’s future performance faster than humans. Professionals will add AI co-pilots to their tool kits, much as software developers are already.
However, the long-term implications for retail investors are more nuanced. As more firms adopt AI technologies, the initial edge provided by AI will diminish. Just as high-frequency trading firms, who rely on speed for a trading advantage, competed with one another to reduce the time it took for their orders to reach an exchange (one high-frequency trading firm once paid US$14-million for a field next to the Chicago Mercantile Exchange so they could put up antennae to route their orders one microsecond faster), so too will AI-enhanced financial analyses develop into an arms race.
In other words, if an edge exists, the market will devour it until no sustainable advantage remains.
This is where the writings of American economist and Nobel laureate William Sharpe demonstrate their timelessness. In his 1991 paper The Arithmetic of Active Management, Prof. Sharpe argued that in a world where all investors are actively trying to outperform the market, their collective efforts will cancel each other out. For every investor who beats the market, there must be another who underperforms.
Therefore, after accounting for costs, the average actively managed dollar will underperform the average passively managed dollar (because the costs of active management are higher than for passive management). Prof. Sharpe’s insight is grounded in the idea that markets are generally efficient, meaning that all available information is already reflected in stock prices.
Combined with the rise of AI-powered financial analysis, Prof. Sharpe’s logic leads to a compelling conclusion: the best way for most investors to benefit from these technological advancements is through a low-cost, passive indexing approach.
By investing in broad market indexes, investors can gain exposure to the market as a whole, benefiting from the collective insights of all market participants, including those using sophisticated AI tools. Crucially, they can do so while minimizing costs, which are a key drag on long-term investment returns.
For individual investors, this means focusing on low-cost index funds that provide broad market exposure. It also means resisting the temptation to constantly tinker with one’s portfolio in an attempt to outperform the market.
The current generation of robo-advisers generally use low-cost index funds, but some may be actively managing the exposure to different passive products within their portfolios. The range of five-year annualized returns of growth portfolios for Canadian robo-advisers reported by The Globe last fall was between 3.67 per cent to 5.98 per cent, in the period ending Sept. 30, 2023.
Research has consistently shown that the more frequently investors trade, the lower their returns tend to be. This is often owing to transaction costs, taxes, and the psychological pitfalls of trying to time the market.
The key to successful passive investing is to stay the course and avoid reacting to short-term market fluctuations. This approach aligns with the principles of passive investing, which emphasize a buy-and-hold strategy and minimizing trading activity. By doing so, investors can take full advantage of the market’s long-term upward trajectory without being derailed by short-term volatility.
However, it’s important to acknowledge that staying passive is easier said than done. One of the hardest things about passive investing is maintaining discipline during market downturns. Investors often believe they can remain unemotional and stick to their plan, but it’s precisely during these times that they are most tested.
The behavioural challenge of staying the course cannot be underestimated. It requires a strong understanding of the psyche of the investor.
Artificial intelligence could be used to provide insights and alerts to help investors avoid common pitfalls such as panic selling during market downturns or overtrading in an attempt to beat the market. For example, the use of AI could help properly match an investor to the most appropriate risk profile for them.
It’s far too easy for pure DIY investors to assume higher risk owing to overconfidence. Predictive analytics could be used to help better understand investor psychology and guide them to portfolios they are more likely to stick to in the first place.
While AI has the potential to revolutionize financial statement analysis, its greatest contribution to investors might not be in picking stocks but in helping investors maintain their discipline.
Preet Banerjee is a consultant to the wealth management industry with a focus on commercial applications of behavioural finance research.