?

This time — no preludes. A small promised piece with numbers and statistics.

📌 According to Mercer, AI is used in at least one investment process by 55% of asset managers. It is most often used to accelerate information search and analysis, as well as to reduce costs. 74% of respondents use AI for automation, 69% as an analytical assistant. That is, it does not replace humans but enhances them, leaving decision-making to them. Only 6% of managers use AI directly for making investment decisions.

📌 The main financial effect of AI adoption so far is that the investment process becomes cheaper and faster. The technology has a weak direct impact on profits: only 8% of managers noted an increase in returns, and 8% noted a reduction in risks. Our world is accelerating: the number of transactions and the volume of data are growing, so decision-making speed is becoming an increasingly critical parameter.

📌 A good example is J.P. Morgan. Its analytical platform SpectrumIQ receives about 7,000 broker reports daily, and the built-in AI assistant Smart Monitor reduces the time from manual information search to a ready-made conclusion by 80%.

📌 However, there is another opinion. A study published in the Review of Financial Studies showed that hedge funds actively using generative AI achieved 2–4 percentage points higher annual returns, adjusted for risk and market factors.

📌 At the same time, the best results were shown by funds that already had their own AI specialists, data, and established processes. I have talked about this — strategy, management, and people are many times more important than AI tools. This study once again confirms that thesis.

📌 The market itself is also changing. Trading is gradually becoming almost round-the-clock: Nasdaq is preparing to switch to a mode of 23 hours a day, five days a week. The launch is expected in December 2026.

📌 This is partly why the financial market is becoming increasingly automated. In the next 3–5 years, a trader will manage not only a portfolio of assets but also a system of AI agents. One tool will read reports and news, another will look for patterns, a third will write code and test strategies, a fourth will monitor risks, a fifth will help execute trades and automate routine operations, and so on.

📌 Thus, according to a J.P. Morgan survey, the share of trades executed through electronic infrastructure is estimated at about 60% in 2026, and the forecast for 2027 is already 70%. AI was named the main technology of the next three years by 51% of respondents (professional market participants). 15% named the development of APIs, routing systems, and algorithmic order execution as the key technology.

📌 According to McKinsey, by 2030, about $2 trillion of traditional financial instruments will be traded in tokenized, i.e., programmable digital form. This primarily refers to bonds, funds, and loans. This will open up more opportunities for automated systems, including those based on AI.

In the coming years, AI will become a new infrastructure layer of financial markets. Analysts, traders, and other professional participants will manage assets through systems of AI agents that can read reports and news around the clock, test hypotheses, monitor risks, and help execute trades. At the same time, not the most sophisticated models but unique data and knowledge will acquire exceptional value. Humans will retain a key role: they will determine strategy, allocate capital, set acceptable risk levels, and make final decisions.