IMAGE VIA GOOGLE
THE MOST dreaded message for a junior investment banker comes neither as a rebuke from a managing director nor as a client’s sudden change of heart. It is the “please fix”, usually landing late on a Friday, asking for a font-size tweak, a logo realignment on a PowerPoint slide or a formatting correction in a financial model. This is how the industry trains its rookies, by setting them to toil in Excel. Artificial intelligence now promises to sweep away this drudgery. The twist is that to achieve this, machines must first be coached by the same humans they could eventually displace.
Investment banking has, for many years, relied on the exertions of its youngest recruits. During boom periods, such as the dealmaking surge of 2021, junior analysts and associates usually clocked 100-hour weeks. The job extends well outside high finance: much of it consists of clerical labour, assembling pitch books and ensuring that each comma and margin sits precisely where it should. Senior bankers usually justify this system as a vital apprenticeship, cultivating muscle memory and care. Yet plenty of exhausted juniors have questioned how logo alignment conveys insight into corporate valuation.
These are the tasks that new AI tools, developed by firms such as OpenAI (which recently hired over 100 former investment bankers from outfits including Goldman Sachs and JPMorgan Chase), are being trained to handle. The process goes further than feeding in a massive dataset. It calls for seasoned bankers to coach the algorithms, walking them through the finicky rules of financial formatting. The bots are learning which figures merit bold type, where italics belong and how to assemble a discounted cash-flow model. This painstaking tuition reveals an uncomfortable truth: much of banking’s vaunted precision stems less from financial brilliance than from meticulous presentation. It is a craft and its tricks are now being handed to machines.
The upside could be vast. Jamie Dimon, the head of JPMorgan Chase, has spoken about the bank’s multi-billion-dollar spending on AI, forecasting major efficiency gains. If the effort succeeds, the technology could free junior staff from spirit-sapping chores, shifting their attention toward analysis or client engagement. In theory, this could improve dealmaking and soften working conditions.
Wall Street’s past encounters with technological change, however, point elsewhere. “More with less” looks likelier as the guiding slogan. If a group of ten bankers can assess twenty companies in the span once needed for ten, the push will be to widen their coverage, rather than to clock off at five.
Automation also unsettles a deeply embedded generational bargain. Veteran bankers, veterans of their own “please fix” ordeals, tend to treat the experience as character-forming. Complaints about brutal hours strike them as evidence of a softer cohort. Younger bankers, by contrast, have grown less tolerant of what they see as futile exertion. For the old guard, the danger is that the hard-earned know-how they are now transferring to machines, the unwritten conventions, could render their apprenticeship system redundant. The tools they help create may erode the culture they aim to protect.
There is, naturally, a rejoinder: automation brings fresh hazards. A human, even one running on little sleep, might catch an oddity that a formatting-focused bot would overlook. Yet manual work also breeds mistakes. The larger threat may be faith in apparently perfect automated output, where a tiny, deeply buried flaw spreads across countless transactions.
AI’s arrival on the trading floor does signal a closing chapter. The dull assignments never served only to finish tasks; they functioned as a means to train and discipline newcomers. As that arrangement is converted into software and automation, the industry must decide what counts most. Is the objective to mould better bankers through ritualised hardship, or to assemble a faster, perhaps sharper machine? For now, senior figures are tutoring the bots. Before long, they may find themselves taking lessons in return. ■
