Leadership1 publisher2 min readPublished
Wharton panel traces how AI is reshaping quantitative finance
Wharton has put the head of Goldman Sachs's research data strategy team on a panel about AI and quantitative finance, and that job title says more about how these desks will hire than the published transcript yet does.
The Board Room · Leadership desk
What happened
- Knowledge at Wharton has published a Future of Finance discussion on AI in quantitative finance, hosted by finance department chair Itay Goldstein in the series' third season, which is devoted to AI in finance.
- The guests are Wharton finance professor Nikolai Roussanov and Ingrid Tierens, who heads the data strategy team in the Global Investment Research division at Goldman Sachs.
- Roussanov said the field originally meant mathematical modelling of complicated assets, following the option pricing breakthroughs of the 1970s, and that this work came to be called financial engineering.
- He said that as statistical methods and computing power spread, quantitative finance came to mean investing with predictive models that forecast asset returns from historical variables.
Compiled by The Board RoomSomething wrong?How this is made
Why it matters
- capability Once dataset selection has a named owner inside a research division, it becomes a budgeted decision that can be argued and reviewed. Until someone owns it, it is a by-product of whoever happens to be building the model.
- decision The retrain-or-hire call on a quant desk has to be settled on internal evidence, because this panel supports the direction of travel and does not supply a staffing ratio.
- precedent A named master's program built around AI in finance fixes what employers can expect graduates to arrive knowing in two years, and it commits a faculty long before hiring data confirms the demand.
- constraint With one practitioner from one firm on the panel, none of this can be used as a market-wide benchmark for how research divisions are staffing data work.
Job titles are weak evidence on their own, but they show what a firm has decided to own. Neither of the definitions Roussanov gave describes choosing which datasets a research division should buy and trust. Goldman Sachs has a team that does that work, and it has a head: Ingrid Tierens, according to Wharton [2]. She also sits on the advisory board of Wharton's Jacobs Levy Equity Management Center for Quantitative Financial Research [10].
Roussanov said that for many years quantitative finance was understood as using mathematical models to value complicated securities [5]. That is a hiring test as much as a definition: a candidate had to show they could handle the mathematics of a complicated payoff. The statistical era that followed swapped in historical variables and predictive models and kept the same test [7]. Dataset selection is a third kind of work [12].
The published record shows a direction. It does not give a size. Wharton's summary says AI expands the possibilities for quantitative investors while raising important questions about the role of human judgment [3], and it lists the skills the next generation of finance professionals will need among the topics covered [4]. The transcript as published breaks off mid-sentence in Roussanov's history of the field, before the panel reaches that question [11].
One job title and one degree launch do not make a labour market. They are the two places this kind of change appears before compensation data does. The sourcing also deserves precision: of the three people on the panel, one works at a financial firm, so the practitioner view here belongs to a single firm [13].
Goldstein said Wharton "just launched the new Bruce Jacobs Master's in Quantitative Finance program in order to help our students get more familiar with some of these new techniques in quantitative finance and in particular AI in finance" [8]. Degree programs are slow instruments. Naming and staffing one commits a department for years. The faculty making that commitment includes Roussanov, who has been at Wharton about 20 years and advises the MBA quantitative finance major [9].
For someone staffing a quant team this quarter, the choice is concrete. Retrain existing model builders to evaluate vendors and dataset provenance, or hire people whose record is in that work and pair them with the modellers. This quarter's version of that call sets next quarter's problem: a team that retrains keeps its modelling bench intact and learns data work at the pace of its slowest quant, while a team that hires acquires a function someone now has to manage next to the models.
What to watch
- Whether Wharton publishes the rest of the transcript, where the panel takes up the skills the next generation of finance professionals will need.
- Whether other research divisions name their own heads of data strategy; one such title is one firm's structure, and a second and a third would make it a pattern.
- What the first cohort of the Bruce Jacobs Master's in Quantitative Finance is hired to do, and by whom.