AI threatens the finance industry's perpetual profit machine
Artificial intelligence doesn’t only threaten to put herds of software businesses out to pasture. Anthropic PBC’s schooling of its Claude models in financial modelling has also sent a cold shiver down the spines of bankers and analysts. While I mostly suspect that the banking industry’s talent for self-preservation will defend it from technological change, I do wonder if the extreme version of our fully automated AI future could make financial services as irrelevant as everything else.
Finance has a very long history of always finding new ways to get paid even as the world of money keeps changing. In only the past few decades, collecting peoples’ savings, buying and selling securities and sending money to the other side of the world have all become faster and cheaper. The internet and smartphones have taken out the frictions and delays that were long a source of revenue for intermediaries.
And yet banks, brokers and money managers are still in business and making handsome profits. The process of turning a dollar of savings into a dollar of investment hasn’t gotten cheaper in more than 130 years, according to studies by Thomas Philippon, a finance professor at NYU Stern School of Business. From the age of the telegraph in the late 1880s to the era of high-frequency, algorithmic trading, the all-in cost of intermediation has remained close to 2% of financial assets, Philippon found.
How does that work? For every development that undercuts what financiers did previously, they’ve been able to come up with........
