How AI is changing capital markets jobs
Equity and debt capital markets (ECM & DCM, respectively) bankers are going through some interesting times. On one hand, IPOs (initial public offerings) and bond issuances are booming like never before, especially in the USA. On the other hand, they’re raising money for the AI data centre boom which may eventually make them redundant. Double-edged sword, really.
Click here to join the bubble by eFinancialCareers, our new anonymous community. ✍️
Of course, one of the biggest impacts of AI on capital markets is the push to market of several colossal AI companies. The SpaceX IPO earlier this year raised $75bn on a $1.8tn valuation; WSJ said that the IPO raised $500m for ECM bankers, with lead underwriters Goldman Sachs and Morgan Stanley earning around $100m from the deal, each. Anthropic is planning a $2tn IPO in October, and that will likely pay out a similar sum. OpenAI might IPO this year too, with a $1tn figure being thrown around by Yahoo.
It’s not just ECM bankers enjoying the boom. CoreWeave, an AI cloud provider, issued $10bn in bonds in the second quarter, and SpaceX issued $25bn soon after IPO, with maturity dates between 2031 and 2056. Based on the usual 0.5% bond proceed math, that’s around $200m raised between the two.
That’s not to mention the huge sums raised by Google, Amazon, and other hyperscalers building huge volumes of computing capacity. Meta alone raised $30bn back in October last year according to Bloomberg, a day after CEO Mark Zuckerberg announced aggressive AI spending. Morgan Stanley estimated $500bn of AI-related debt would be issued in 2026, a four-fold increase on 2025, Reuters reported.
AI is changing the actual jobs that capital markets bankers do too. As we noted when looking at the impact of AI in M&A, AI has a fantastic ability to build pitchbooks and models. Pitchbooks are the materials that banks use to pitch clients and models the valuation calculators that decide on the price of a security. AI is good and only getting better at doing that – it’s a key focus of both Anthropic and OpenAI.
As early as January last year (before the release of GPT4.5 or Sonnet/Opus 4, mind you!) Goldman Sachs CEO David Solomon said that AI could draft 95% of an IPO prospectus, with the last 5% mattering most because “the rest is now a commodity”.
Solomon has not since updated the figure, but there are some indications of the direction of travel. Private equity firm CVC, selling Greek e-commerce firm Skroutz earlier this year, did not use investment bankers to run the sale – it used an AI agent, the Wall Street Journal reported. That’s a sale, an M&A deal – that’s not a capital markets deal – but private equity firms have traditionally relied on IPOs as investment exits. Either way, it is not an inspiring trend if you’re a banker.
There are some differences between how M&A bankers and capital markets bankers work, especially for DCM bankers. DCM clients issue securities much more often than ECM or M&A clients do, and bonds are much more templated than a merger or IPO. Bonds also have a smaller percentage fee on issuance than an IPO – around 0.7% vs 7%. Blockbuster IPOs, such as SpaceX’s, are outliers. Generally, the bigger an IPO, the less it pays as a percentage of the deal.
Either way, capital markets bankers have a certain level of protection from AI. Sumeet Chabria, former global tech & operations COO at Bank of America and founder of C-Suite consultancy ThoughtLinks, told Business Insider earlier this year that while large parts of investment banking could be automated by AI, final IPO and syndicate (pricing securities through a group of banks jointly arranging distribution) pricing – aka, ECM and DCM bread and butter – will remain human-led. Chabria said that price-setting requires banker judgement, market feel, and live investor feedback. Winning mandates and maintaining relationships will also remain human-led, of course.
In the short-term, it’s likely that banks will hire fewer capital markets bankers, especially as juniors. Debasish Patnaik, head of McKinsey’s AI consulting arm QuantumBlack, told Fortune in June that junior analyst classes would be cut by as much as two thirds due to AI. That’s been a threat for more than two years at this point, however. It doesn’t seem to have come true. Goldman Sachs analyst class sizes have plateaued in recent years, not shrunk.
Have a confidential story, tip, or comment you’d like to share? Contact: +44 7537 182250 (SMS, WhatsApp or voicemail). Telegram: @SarahButcher. Signal: sarahbutcher.22 Click here to fill in our anonymous form, or email editortips@efinancialcareers.com.
Bear with us if you leave a comment at the bottom of this article: comments are moderated intermittently by human beings. Sometimes these humans might be asleep, or away from their desks, so it may take a while for your comment to appear. You must take sole responsibility for comments you post on this site. We will take reasonable steps to weed out anything that we consider to be offensive or inappropriate.
