It’s hard to say exactly what bank technologists do. The term “technologist” is rather generic, covering dozens if not hundreds of roles in a bank – everything from infrastructure, to cybersecurity, to the cutting edge of research, including quantum computing and AGI.
For the most part, what technologists do is support the rest of their company. Banks are huge employers of technology talent and they spend accordingly. JPMorgan, for example, spends $7.6bn in technology in total and $2bn on market making technology for its sales and trading business alone according to its most recent investor letter. Of that $7.6bn, $4.5bn goes to platforms and user experiences while a further $3.1bn goes on modernizing technology, maintaining existing technology and technology to protect the firm and its customers. Banks have large “legacy” code systems to support their operations.
They’re trying to retire old applications and to move as much as possible to the cloud. They’re also trying to automate and to use AI as much as possible. On a day-to-day basis, what this means for you also varies. Working as a technologist supporting the trading floor will be very different to working as a technologist creating a user interface for a wealth management app, or deep in an ancient payments system.
Banks have historically been quite bad at allowing technologists to remain at the coal face, coding, and have tried to promote them into management roles. However, this has changed with the introduction of “distinguished engineers,” which are roles allowing banks best developers recognition for their development skills.
There are a number of roles available in a bank’s technology team.
A lot of people want to work in this space when they first start out. Front office technology is sexy. It’s closest to the bank’s sources of revenue – and in fact, the “front office” is usually defined by the roles in the bank that are in themselves revenue-generating. Think commodities, equity derivatives, rates, portfolio managers in asset management, or maybe even a banking team.
As a front office technologist, you’ll be developing tools like trade and position blotters or creating pricing engines in partnership with quantitative research (see our section on quant careers) and market data teams.
Sounds great, right? Well, the bad news is that front office technology teams often run very lean. They can have arduous support rotas and a lack of investment in adopting strategic frameworks and renewing tech stacks. The most important work is given to trusted individuals, and work tends to be very atomized. Front office technology can be both stressful and boring.
Core technology teams are where the long-term strategic work is done. This is about long-term planning rather than day-to-day functioning. A core team might embark on a technical strategy for a user interface and develop a framework that starts to be used by business-aligned teams. If you’re someone who mostly wants to interact with other developers and be able to keep up with industry trends, then core teams are the way to go.
If you work in the middle or back office, you’ll find a whole host of teams doing different types of work. There are compliance and regulatory technology teams who often have to work to externally mandated deadlines. You also have market risk, payments and settlements, valuation control, and capital management.
If you work on middle- or back-office technology requirements, your clients will be teams in operations, in financial control, in compliance, or in any other non-revenue generating team. Things tend to be a bit more relaxed than front office, and you’ll be more likely to use industry standard tools and languages. Some of the roles in compliance (RegTech) can be interesting and involve the use of natural language processing (NLP) Artificial Intelligence (AI), or Quantum Computing.
Infrastructure technologists work on the technological framework that underpins the bank. This includes cybersecurity and cloud computing. If you work in infrastructure technology, your clients are other developers. There will typically be highly skilled specialists hired in this area – with ‘fellows’ or ‘distinguished engineers’ more common than elsewhere in banks’ technology teams.
Business analysts are the people who intermediate between the business and developers. This is a role that’s going out of fashion – banks want ‘T-Shaped’ developers. What they mean by this is that they want to cut costs and have developers talk to the business directly – pretending that complexity and interaction over multiple silos doesn’t exist. There are fewer people working in business analysis now than before: graduate tracks for business analysis no longer exist at some banks.
Project managers can be some of the most infuriating people for the average developer to deal with – they put meetings into developers’ calendars to talk about project timelines, why things are late, and delivery milestones. Project managers are often used to arbitrate between different teams for complex projects. To a large degree they are little more than secretaries, but the senior ones are paid like technical architects. If you’re a developer, this can be particularly infuriating.
Data science jobs arguably need an entire section of their own in this guide because, as data proliferates, banks and funds are building armies of data specialists. Hedge funds especially are increasingly trying to get an edge by looking at “alt data” sources, like information on footfall in shopping malls, or sentiment on social media.
Broadly speaking, there are three types of data-related roles in finance: data analysts, data engineers, and data scientists. Data analysts interface with the business to find out what the data needs to do, and develop visualizations to show that information. Data engineers take raw data and prepare it for analysis – cleaning it, moving it, tagging it, and so on. And data scientists create the models that are able to extrapolate from the data in a way that is presentable to the business.
Data science has become more important in the age of AI and machine learning. Large Language Models (LLMs) rely on accessible data in the cloud and banks have been constructing “data lakes’ for this reason.
For a long time, the biggest challenge that any bank employee in a support role could face was offshoring. This is when jobs are moved from London and New York to the likes of Bangalore, where Goldman Sachs has its office employing the most people. Nearshoring, or moving staff to lower cost offices away from expensive capital cities is also popular. Goldman also has technologists in Warsaw, Birmingham, Dallas, and Salt Lake City.
The biggest change to banks’ technology jobs, however, is artificial intelligence (AI). Goldman Sachs’ CTO Marco Argenti told CNBC that “developers are already using some of the assisted coding technology,” and that at the time (March 2023), developers were already managing to write 40% of their code using generative AI.
Ex-Goldman Sachs tech MD Kwame Nyanning said earlier this month that AI would replace junior coders, with more experienced technologists guiding and debugging the process based on holistic needs. The “pipeline of hiring people and maturing them is going to be dramatically impacted,” Nyanning said. That means that there might be fewer technologists in the future – and those that are left will be more business-minded than before.
Because Large Language Models are general and financial services firms use a lot of specialist terms and proprietary data, some firms are large language models of their own. Bloomberg for example, has developed “BloombergGPT”, which will “assist Bloomberg in improving existing financial NLP tasks, such as sentiment analysis, named entity recognition, news classification, and question answering, among others”, in its own words.
Not all financial services firms are taking that path, however. Goldman Sachs indicated last year that, while it was exploring potential AI use cases, the company had no intention to build its own LLM from scratch, according to Venture Beat, a tech-focused news site. Goldman announced in June of this year that it had developed something called “GS AI Platform”, based on ChatGPT (OpenAI) and Gemini (Google) frameworks, which uses Goldman’s proprietary data within existing LLM model frameworks.
The rise of LLMs has created a new genre of job for “prompt engineers,” experts in the hyper-specific form of English that is required to efficiently extract data from an AI. Argenti recommended his daughter study philosophy as well as engineering, for example, and said that “ambiguous or not well-formed questions will make the AI try to guess the question you are really asking, which in turn increases the probability of getting an imprecise or even totally made-up answer.” As such, the leading engineers of the future won’t be asked if they can code, but rather “can you get the best code out of your AI by asking the right question?”
Not all AI is related to LLMs, though. JPMorgan has an AI called “Moneyball”, which the Financial Times reported as being a tool that shows portfolio managers how they reacted in similar situations to similar news, as well as “help them correct for bias and improve their process.” Moneyball is part of JPMorgan’s “SpectrumGPT” portfolio management platform, which, however, is based on OpenAI’s GPT-4.
In the future, there may be more jobs in banking for quantum computing engineers. JPMorgan and its head of global technology applied research, Marco Pistoia, are very much at the forefront of this new wave of innovation. Many of the quantum computing jobs in banks relate to quantum cryptography, or systems to prevent banks’ security walls being penetrated from a new generation of quantum computers.
You might imagine, with the variety of jobs available in a bank’s technology team, that the qualifications you need to get a job there vary rather wildly. You’d be wrong. When we analyzed Goldman’s tech analysts (juniors), there was one degree that outnumber all others: computer science.
There were some exceptions, but mostly for software engineering graduates – computer science was the clear king of that particular hill. A lot of the junior class held master’s degrees, too. Master’s degrees in financial technology, such as those offered by Imperial College London and the University of Hong Kong, could be particularly valuable here.
Financial engineering, despite usually being associated with quantitative roles, is also a strong option, especially in data roles. Our ranking of financial engineering courses put Baruch College, in New York, on top, followed closely by Princeton and Carnegie Mellon. Imperial College London topped the UK rankings.
In the age of AI, however, that might not be enough to make the big leagues. Some bank roles, such as the most cutting-edge machine learning roles, can require a PhD. These will mostly put you in the “labs” that banks have set up to develop their AI capabilities.
It’s also possible to get directly into a banking role before your bachelor’s degree even starts. Some banks and universities, mostly in the UK, have partnered to offer degree apprenticeships, which are a combination of formal university education and work experience. They’re very difficult to get into, however; a participant in the scheme that we spoke to last year said that 4,000 people applied to the 28 places on his course, indicating an acceptance rate of 0.7%, less than most investment banking summer internship roles, including Goldman Sachs and JPMorgan’s.
Additional qualifications for technology are harder to pin down, but depend on what you want to specialize in. In a lot of cases, they simply aren’t necessary – but if, for instance, you want to work in network engineering, it can help to have a Cisco-issued qualification such as a CCNA or CCNP.
Another entry point for a bank technology role is via the cloud. Banks works with different and varied cloud computing providers – such as Goldman and Amazon, or Deutsche Bank and Google - and specializing in their respective cloud services isn’t unheard of as a way to get into the industry. Goldman, for example, hires a lot of people from AWS, including its current Chief Information Officer, Marco Argenti, who joined the bank from AWS directly back in 2019.
All major cloud providers also offer qualifications and certifications to prove knowledge, and there are internships that also offer them in the process.
When we analyzed the most popular coding languages in financial services – based on analysis of job listings and their requirements – the most popular languages were, by quite some margin, SQL, python, and Java. SQL was top, with nearly a full quarter of technology financial services roles advertising for specialists in the language. Python and Java were sought in 18% and 14% of applications, respectively. All other languages were below 6%.
SQL: Although SQL might not seem like the obvious choice for the financial services industry’s favorite coding language, it makes perfect sense with a bit of thought. It can deal with huge amounts of data. It integrates beautifully with Excel. It allows for complex queries, data analysis, and reporting. It might not be as useful in some front office capacities (and therefore likely not as well paid) as Python and Java, but it’s a silent workhorse.
Python: Man Group, the biggest hedge fund in the UK (and Europe), once said that its first language was English – and that its second language was Python. The language’s influence in finance can’t be understated, and with good reason: it’s incredibly intuitive, easy to use, and has a huge ecosystem of libraries that can be tapped into. It also integrates easily into other languages.
Java: Java has its own uses in the finance ecosphere. For one, Java (in certain iterations) and especially its “grandfather”, C++, are considered very low latency languages. That means that you can build code that executes its instructions faster than other languages – and that’s an extremely valuable trait in fields such as high-frequency trading, where microseconds can cost millions or even billions of dollars.
Python and Java aren’t the only languages, however. There’s a host of them: Rust, R, C#, Q, KdB, and so on. Generally, these are niche languages, with specific uses. Some big banks even have their own internal programming language, most famously Goldman Sachs’, called “Slang”, which is an internal proprietary coding language similar to Python, and underpins much of Goldman’s front-office technology. It was built to work with SecDB, Goldman’s risk and pricing engine, and it’s very good and that – even if Goldman is moving away from it, and towards Python and Java.
Technology roles in banks (and financial services in general) don’t pay as well as front-office roles such as those in M&A or sales & trading, but they pay very handsomely regardless, as figures below from our 2024 salary and bonus survey show.
As with many other support roles, such as those in risk or compliance, the majority of a technologist’s compensation is in their salary. A technologist can expect a bonus of around 25 to 50% of their annual salary on average for the majority of their career, until they reach Managing Director (MD) level.
Because banks are generally quite protective of the MD title, very few technologists reach this rank. However, those that do are extremely well paid – although that also usually comes with extensive leadership and management responsibilities. The further away you are from the code, the more money you make.
One benefit of working in bank technology, as opposed to at a fully-fledged tech firm (or in a bank’s revenue-generating roles), is that the hours are usually pretty good, and the pay is higher than other support roles – meaning that bank technologists have some of the best lifestyles in the business.
Have a confidential story, tip, or comment you’d like to share? Contact: +44 7537 182250 (SMS, Whatsapp or voicemail). Telegram: @SarahButcher. Or email editortips@efinancialcareers.com. Signal also available.
Bear with us if you leave a comment at the bottom of this article: all our comments are moderated 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. Eventually it will – unless it’s offensive or libelous (in which case it won’t.)
