Getty Images; Isabel Fernandez-Pujol/ BI
Wall Street bosses used to turn their noses up at the idea of open source.Open source is non-proprietary software that developers can view and modify collaboratively.Now, finance firms are finding the upside of open source and are embracing it like never before.
Software developers at Morgan Stanley are accustomed to pushback when pitching to build something new.
“I tell technologists all the time, if you’re building something and it is not a source of competitive advantage, you have to wonder: ‘Is this already outside? Should I be using it if I’m not using it? Is there something or a gap that I should be improving and contributing back my improvement?'” Dov Katz, a managing director helping oversee the bank’s tens of thousands of developers, told Business Insider.
“And only then should you build something net new.”
Katz is talking about prioritizing open-source technology, which is non-proprietary software that can allow developers from all over, including rival companies, to view, collaborate on, and modify the code behind it.
The open source-first approach and collaborative attitude is a departure from the hyper-competitive and proprietary nature that Wall Street has come to be known for. But Katz’s stance is part of a sea change in the industry as banks, hedge funds, and asset-management firms find more reasons to open up.
While revealing their secrets is alien to the Wall Street DNA, finance firms have been shedding those conservative notions over the past few years. In that shift, financial institutions are growing to become contributors to open source, not just consumers. As such, the amount of intellectual property flowing to open-source hosting platforms and the number of contributions from financial services professionals continue to rise.
In 2024, financial services workers contributed more than 750,000 commits, or changes, to GitHub, one of the main platforms for hosting open-source projects, according to the Fintech Open Source Foundation (FINOS), an open-source nonprofit for financial services. That’s a 55% increase from 2021 levels. Around the same time, financial companies from JPMorgan to BlackRock and Man Group also released internal platforms for open source.
Wall Street’s embrace of open source could come at a good time. Global technology spending in banking ballooned to $650 billion in 2023 — roughly the GDP of Belgium or Sweden, according to an October report from McKinsey. Despite the steady 9% a year increase on average in spending, bank shareholders and analysts are still questioning when they’re going to see quantifiable value, the report said. Scrutiny comes as finance firms carve out even more resources for AI, which can be a costly endeavor.
One of the main upsides to leveraging open source is cost efficiency: sharing the burden of development while freeing up engineers to work on other tasks. It could allow firms to be more nimble in the long run, cutting down on their yearslong roadmaps because outside contributors can make fixes to codebases and introduce new features.
“Otherwise, you’re just building tech debt forever. You’re going to have to support something forever if you’re building it yourself,” Katz said.
Goldman Sachs’ chief data officer, Neema Raphael, has witnessed this cultural shift over his more-than two decades at the bank.
Attitudes went “from almost very secretive, we’re going to keep everything within,” to accepting the benefits of open source with some caution, he said. “Then finally, to like, ‘OK, we should actually be a player in this space in the sense that we should also contribute to open source'” and get leverage from that, he said
There’s momentum at senior levels, Katz said, for engineers to work together to solve common industry problems and that more diverse contributor pools can lead to better solutions. The reasons are not all altruistic, of course.
Open source can lower the cost of development and ownership, help with scalability, and even provide some regulatory insulation. It’s helped banks speed up integrations with technology vendors and made it easier to share data with clients. There are even ripple effects to engineering retention and job satisfaction, Raphael said, adding that, “contributing to open source and having your name out there in open source is almost like table stakes.”
The rise of open source on Wall Street
It’s hard to talk about Wall Street’s adoption of open source without talking about the cloud.
Wall Street largely held out on adopting the public cloud, with regulatory uncertainty and security concerns being two big reasons. But that started to change around 2019 and 2020, when the cloud became the place for data and analytics, faster experimentation, and a breeding ground for launching new businesses.
To reap the benefits of speed and cost efficiency, companies had to adopt a different toolkit and approach compared with developing software on physical servers. That brought many finance firms face to face with open-source technologies, like Kubernetes and Docker, which are cornerstones of the cloud-native software development.
While it’s tricky to distill the exact number of commits financial services workers made on GitHub — company policies and restrictions push developers to use their personal accounts rather than ones tied to their work emails — FINOS’s low-ball estimate shows a steady increase. The latest data shows a 26% jump in the number of commits between 2023 and 2024. Meanwhile, FINOS last month topped 100 members in a group that includes the four biggest US banks, the hedge fund Point72, American Express, as well as vendors like the public-cloud providers and Nvidia.
At a company level, finance firms have also been open-sourcing platforms that were originally developed internally. JPMorgan open-sourced its design platform, Salt, in 2022, Morgan Stanley contributed its Morphir platform that helps business and technology teams collaborate in 2020, and Citi open sourced GitProxy, a risk and controls tool, the same year. Other examples include Goldman Sachs’ key data platform, Legend, and Aladdin, the crown jewel of BlackRock’s tech.
“Eventually, people, they evolve toward a standard and competitors realize the benefit of having the standard outweighs being different,” Katz said.
In many cases, what’s being open-sourced are software-development utilities, like methods for writing log files, ways to connect to networks, formulas for linking data sets, or frameworks for testing applications. Katz described it as “trivial commodity capabilities” that aren’t core to business or drive value.
“You want to test your application? Your IP is the application. How you test it? Much more commodity,” he said.
Open source hits private equity
At Blackstone, a project is underway to build an open-source library for its portfolio companies in the realm of data management, data engineering, and generative AI, the firm’s chief technology officer, John Stecher, told BI.
They were areas where the PE giant was already spending a considerable amount of its own engineering resources for its own data and AI ambitions.
The aim for the project, which started about a year ago, is to give portfolio companies a head start on the engineering legwork while allowing them to pivot engineers to spend time on more meaningful work. Blackstone also benefits since portfolio companies can contribute back to the effort by fixing coding bugs and developing new features and functionalities, Stecher said.
While there are no direct savings for Blackstone, it should allow portfolio companies to be more effective with fewer engineering resources, he said.
The ethos of redeploying engineering talent is one that Stecher prioritizes.
“As we go through discussing open-source, the more I can pivot engineers away from building functionality that is non-proprietary and duplicative on the Street and just have a smaller set of those engineers build that collaboratively, I can have them focus on delivering much higher value work for our clients at the end of the day,” he said.
Using open source doesn’t come without risks, however. One challenge is around licensing, Goldman’s Raphael said. Open-source software is released under specific licenses that have their own set of requirements, like disclosing modifications or not being able to use it for commercial purposes.
“We have a full formal governance of what projects are allowed to be in, what vetted licenses are allowed, to make sure that we’re not using things inappropriately,” Raphael said.
Another issue is because open-source libraries are really available to anyone, it’s possible for people to try to sneak in bad code that could lead to far-reaching cyberattacks in what’s known as supply-chain attacks.
“We take the supply chain of what’s actually being used where, and which libraries we’re using from open source, tracking that also really at a detailed granular level to make sure we know where all the software providence comes from,” Raphael said.
The next open-source frontier
Now, finance firms are even looking to bring the open-source approach to the latest technology frontier on Wall Street: AI.
The intersection of generative AI and open source has been on full display ever since the Chinese AI lab DeepSeek rolled out its AI models that are as good, or even better than, the best products from OpenAI. DeepSeek’s offerings have sharply undercut OpenAI on price, shaking up Silicon Valley and Wall Street. The budding rivalry between the two companies has thrown a spotlight on open-source models (like DeepSeek’s) versus proprietary approaches (like OpenAI’s).
Last year, FINOS launched an AI readiness group that’s attracted firms like Citi, Morgan Stanley, and the London Stock Exchange. The effort is designed to create a collaborative environment to develop frameworks and policies around generative AI, related to concerns like mitigating bias in models and adopting new tech.
“That’s very much an industry problem, not an individual company problem. What better thing than to mutualize it and try to tackle it together,” Katz said.
+ There are no comments
Add yours