AI is changing both how much work a trader can produce and which parts of that work remain difficult for competitors to replicate.
Among the businesses I work with, the increase in individual productivity is already tangible. Traders can investigate more opportunities, develop tools more quickly and put more of their knowledge to work. Coding has become more accessible, and scale is no longer purely for the behemoths of the industry. But where that extra capacity is directed matters enormously.
Where firms mine similar data for similar signals, faster research can bring more competition to an already crowded trade. Applied to a specialist's understanding of a market, the same technology can make previously impractical ideas viable. AI can erode some trading advantages while strengthening others.
The trading industry has long understood alpha decay. A profitable relationship becomes better known, more capital follows it, and returns get squeezed. AI adds to this pressure by reducing the time and cost of finding, testing and implementing ideas.
Equity statistical arbitrage offers a useful example. These strategies identify relationships across stocks, often using the same core datasets. When firms investigate similar patterns and apply similar risk constraints, they can end up with overlapping positions. Producing more signals may simply provide more ways of arriving at the same trade.
That doesn't mean the approach has stopped working. Systematic equity long/short funds gained 3.46% in September 2026, their strongest month of the year, according to Goldman Sachs figures reported by Reuters. Equally, a difficult month doesn't prove AI has weakened a strategy. Market conditions, crowded positions and forced selling can all affect returns.
AI's impact depends on what keeps an opportunity profitable. Faster analysis may squeeze an advantage built on spotting patterns first, but have less impact where investors have different objectives or trading constraints.
Sometimes the obstacle is the volume of information. Elsewhere, it's product complexity, investor behaviour or the difficulty of financing and hedging a trade. Better tools help, but don't necessarily remove these challenges.
Consider mortgages. Two securities can look similar while behaving differently because of the borrowers behind them and how they repay their loans. Specified pools provide more detail on those characteristics, with some sacrifice in liquidity. Understanding that information requires product knowledge as well as processing capability.
Convertible bonds combine contractual terms, credit, financing and equity optionality. Identifying an attractive valuation still leaves the question of whether it can become a profitable trade.
These are sophisticated markets where complexity alone doesn't guarantee an edge. But opportunities tied to investor constraints or an instrument's mechanics may persist even when information is widely understood.
An experienced trader may already have ideas they want to explore, but lack the time, development support or research capacity. As those costs fall, smaller teams can pursue more opportunities.
It includes deciding which problem deserves attention, recognising unusual pricing and connecting developments across markets. AI gives that judgement more room to operate.
For FinTech businesses, making fragmented information usable or shortening the path from an idea to a working tool can change the economics of trading. Understanding the user's market and workflow is central to getting that right.
It also shapes what I look for in candidates. Technical ability matters, alongside explaining why a trade should work, how to execute it and what might cause the opportunity to disappear. Those questions become more valuable as research becomes easier to produce.
AI is making familiar approaches more competitive while opening up work previously too costly to undertake. The opportunity is to give people with deep market knowledge more capacity to act. Plenty of good ideas have never made it beyond a trader's notebook. More can now be put to the test.
