With Gemma Hagen, Director of Technology, Aspect Capital
What is your professional background and role/responsibilities at Aspect Capital?
I look after data, software development and production. My background is software development having studied computing at Imperial. My career began at Deutsche Bank doing FX HFT trading. This is my second stint with Aspect. In between I spent a few years at Pilot FX, a small HFT firm. I’ve worked on building all systems involved in running an HFT platform, all C++ based, covering the full stack.
While at Pilot I specialized in execution, particularly FX. I left Pilot because Aspect was embarking on a significant execution project integrating with a third-party EMS to improve FX trading. I moved to Aspect to lead execution development and was appointed Director of Technology in 2021.
How would you characterize the current state of technology on the buy side? How modern (or not modern) is it?
It’s a mix across the industry. More established companies carry the weight of legacy systems more than newer companies, which are built using modern techniques.
At Aspect we have a lot of legacy systems, but we still deliver huge amounts of business change; having a legacy system doesn’t necessarily preclude quick delivery.
There is an industry-wide modernization trend, much of which is driven by advances in the technology space over recent years. Investment firms are more comfortable migrating to the cloud, and there is a huge rise in the number of vendor platforms and industry standard tooling that’s available. I think we’re in a state of significant change. Many more established firms are embarking on significant modernization projects, while still carrying lots of legacy systems.
How old is the legacy technology that’s out there — is it from the 1990s? And how useful does it remain?
We have systems at Aspect that were established back when the company was founded in 1997. They’ve been enhanced and they’ve been changed, but the premise remains largely unchanged.
That’s an interesting aspect of legacy technology. It’s easy to conflate legacy with not useful, but often legacy systems are a sign of good development as they’ve stood the test of time. If it’s not broken, don’t fix it – otherwise they would have been swapped out or decommissioned long ago.
What are the pros and cons a firm needs to assess regarding a technology modernization?
Projects are typically long-running, significant and challenging to complete. Ultimately any decision is driven by business value.
What is motivating you to consider this? You can divide these into a few different themes, one of which is productivity improvement. As systems get older, they become more difficult to develop, maintain, support, and test, so you can improve that by modernizing the stack. Another consideration is, are they still fit for purpose? Decades is a long time for technology to match a company. Your stack reflects your business strategy, so it must align with your mission.
There are opportunities for technology to be a competitive advantage on the buy side if you can be quicker, more scalable or more agile. Often you can’t do that with legacy systems. For instance, data centres constrain your ability to scale quickly, whereas cloud offers infinite scalability at your fingertips.
While modernization presents opportunities, there is a cost. Your technology teams will be contended, so, when you allocate resource to a migration, that comes with an opportunity cost of not delivering on other business priorities. You do have to consider your business and where your focus should be. Is modernization the right focus? Another consideration is, these are critical systems and high-risk to migrate, so you must be methodical, understand and mitigate the risks migrations pose.
What does a significant technological modernization project entail?
You need to understand why you’re doing it and your end goal. You need to be confident that the investment is worth making. Then it’s a case of strategizing – what does modernization look like? Which problems that we’re trying to modernize are generic, whereby we could benefit from an evolving, cutting-edge product that doesn’t hinder our ability to make proprietary aspects of our systems work?
A good example is the data platform revamp at Aspect. We are using a third party, cloud-native data platform that manages the segregation of storage from compute and gives us infinite independent elasticity. You must think through where it’s worth spending development effort, against where you should use a provider. One thing I find that isn’t often done well – and at Aspect we’re learning to do well – is proof of concepts. When making an outsourcing decision, you build a premise for why you think a system is right for you, but you need to test that rigorously.
Taking the time to devise a POC is critical. When you pick a provider you partner with them for a long time. You can’t afford a mistake in that decision. Equally, do not be afraid to fail at that stage.
These projects often involve cloud migrations. Therefore, you need to think holistically about your entire stack. It’s not just about the software system at the top, or moving from data centre to the cloud or the platform that you’re running on. These things work closely together. If you want to move to the cloud and get the cost benefits of that, that means you also may move to Linux for example. Always keep in mind the goals you set at the beginning.
What is the end benefit of a technological migration, and what is the risk of putting it off?
The benefit should be your original goal. However, there are unexpected benefits too. Modernization projects are great for teams. Technologists love learning and working with new technologies on challenging projects. Those opportunities make a difference in retention because people feel engaged and challenged. The more legacy technology you have, the harder it can be to hire too, especially for junior roles. People want to do the fun, new, shiny stuff.
Often with migrations, you uncover unexpected opportunities. We’ve been investing in execution research over the past few years. The further we go we are discovering opportunities across research = opportunities to diversify strategies and to improve signal generation. Opportunities do crop up unexpectedly and you should apply the benefits across your business.
You will be left behind if you don’t modernize. The longer you leave it, the harder it is. You can’t be in a position where you didn’t migrate to the cloud. Suddenly you need more compute resource, and you can’t get that because you need to rethink your data centres. You can’t risk standing still.
What does the future hold for tech modernization on the buy side? Is today’s state of the art technology tomorrow’s legacy technology?
On the latter, yes. Technology is the fastest evolving field and changes at lightning speed.
One strong theme for the future is cloud infrastructure. Firms will run their research and infrastructure in the cloud. Without that you are going to be at a disadvantage in terms of speed, agility, and breadth of research. I wonder at what point exchanges, brokers and liquidity providers start moving to the cloud rather than data centres.Infrastructure in finance is built around proximity. A lot of money has gone into squeezing the most out of every microsecond – think of cross connects and even microwave links. It’s a lot to give up for people to say, “Actually, we’ll move all that to the cloud.” What does that mean for high frequency trading for instance? There are lots of big questions in the execution space.
Listen to Gemma Hagen discuss tech modernization on the buy side on the GlobalTrading Podcast.