News

With “Claude Opus 4.6”, AI is pushing deep into the core business of analysts and banks – right up to Bloomberg terminals. Shares in financial research firms are sliding. Is the financial sector really facing disruption, or merely the next fee model? “Historically, the industry has shown remarkable resilience,” says Thorsten Fischer, Managing Director and Head of Portfolio Management at Moventum AM. The real disruption may ultimately be social rather than technological.

AI developer Anthropic is shaking the foundations of the financial industry with the new version of its Claude model. With “Claude Opus 4.6”, the software is expected to take on tasks previously reserved for well-paid analysts, investment bankers and research departments. According to Anthropic, Claude independently analyses corporate data, regulatory disclosures and market information, and connects these via interfaces to professional analytics tools such as Bloomberg. The stock market reaction was swift: shares of established financial research houses came under pressure. For the first time, markets are visibly pricing in the possibility that AI could directly affect the industry’s fee model – not merely its back-office processes.

Yet however spectacular the technological progress may seem, financial intermediation has historically proven surprisingly resistant to disruption. For more than 130 years, the “all-in cost ratio” of capital intermediation has remained stable at roughly two per cent of assets under management – from the telegraph to electronic trading platforms and high-frequency trading. When one revenue source disappears, the sector usually invents a new and often more complex one.

Asset management provides a prime example: despite the global rise of passive index products, the worldwide revenue ratio still stands at around 34 basis points on assets under management – barely lower than two decades ago. Lost income from active retail funds has been offset by new business areas: hedge funds, private equity and private debt vehicles, as well as illiquid private markets charging premium fees.

Investment banks have likewise responded to regulation and technological change in markets by developing new revenue streams. Structured products, complex derivatives and the financing of alternative asset managers have replaced traditional trading margins. “As transparency increases, complexity often rises with it – and so does the opportunity to establish new fee models,” Fischer explains.

Since the 1980s, the global financial system has also expanded dramatically: from roughly the size of the real economy to several times its scale. Deregulation, asset inflation and credit expansion have been the fuel. Against this backdrop, it seems plausible that AI will initially act less as a wrecking ball and more as another engine of efficiency gains – and new products.

“The real disruption may therefore be social,” Fischer argues. “If AI replaces labour on a large scale and the marginal cost of many services tends towards zero, there is a risk of a strong concentration of wealth among the owners of algorithms and platforms.” In such a scenario, models such as a state-guaranteed basic income could structurally alter saving and borrowing needs: fewer mortgages, lower demand for traditional retirement provision, and thus less conventional financial intermediation.

In an extreme case, the importance of the financial sector could shrink if capital allocation becomes largely automated and a greater share of consumption is state-funded. Traditional banking services would then recede into the background. “At the same time, an ironic counter-trend is emerging,” Fischer says. “Money and markets could move further into the realm of entertainment – with prediction markets, options speculation and crypto as the gamification platform of a post-industrial society.”

Despite these scenarios, history points to the remarkable adaptability of the financial industry. Few sectors have created new products, fee models and employment fields as consistently. AI therefore represents a serious structural challenge, but betting against the innovative capacity of banks, asset managers and exchanges has rarely been successful in the past.

The key question is therefore less “Will the financial sector survive?” than “In what form?” Models such as “Claude Opus 4.6” mark the beginning of a new phase in which the balance between humans, machines and the Wall Street business model must be recalibrated – between margin pressure, efficiency gains and the enduring inventiveness of an industry that has spent more than a century finding ways to earn its two per cent.