In a new mChats, Scott Rofey, Global Head of FIC and Cross Asset Risk, discusses what effective risk management looks like at Millennium, and how AI is shaping that work today.
What first drew you to your role at MLP, and what continues to excite you about it today?
I started my career on the sell side, and it was obvious to me from that seat that Millennium had an amazing platform for operating at scale. The chance to become part of that and contribute to it felt like a once-in-a-career opportunity. I’m learning, problem-solving, and contributing to the firm’s operating vision, and that excites me each day. I’ve learned a great deal about how we do what we do across an incredibly broad range of assets and strategies. I hope I’ve contributed to how we analyze our business and make decisions, and there are still many things I’m interested in exploring and doing here.
What does effective risk management look like in practice at a firm like Millennium?
It starts with making the scale of our platform work for us. I believe our analytics team is very strong, and our technology teams are amazing partners. Jointly, we create risk visualization and assessment techniques that enable us to deeply understand our positions and how they are evolving in real time. Doing that well across the scope of what we trade is no small task, and I like to think we hold ourselves to very high standards in doing it.
A big part of that is creating clarity around risk. I want our Portfolio Managers to feel that their risk capital is clearly defined, so they understand our shared goals.
How is AI shaping the way you think about risk management today?
The pace at which AI model advancements are unlocking new capabilities is very impressive. Parsing overwhelming amounts of market content into insights tailored to our interests is a big productivity boost. Interrogating market and risk data to measure and understand changes is much easier. One can build skills for daily tasks, and agents which combine skills in dynamic ways. Lastly, testing models in code has become much faster and easier – we can try many methods to solve a problem without worrying that we will have wasted time if they don’t bear fruit. For people who like working with data and models in finance, AI is very exciting.
What is the best piece of advice you have ever received?
I’ll give you the best piece of advice I never received, and had to work very hard to learn: I spent a long time falling in love with getting answers right, but you also have to learn how to handle being wrong and respond with perspective.
