Fireside Friday with… RBC’s James Hilton

The TRADE sits down with European head of multi-asset agency solutions at RBC Capital Markets, James Hilton, to explore shifting buy-side demand for algo solutions, multi-asset, and artificial intelligence.

How are you seeing buy-side demand for algorithmic solutions change?

Algo trading is very well established and we’ve had this catalyst of Mifid II, which effectively means that more and more flow is now directed based on performance, particularly via performance weighted algo wheels. The biggest ask that we get when we’re pitching with clients is to present a differentiated approach. The idea of selecting a pool of homogeneous algorithms just isn’t going to result in performance improvements over time.

The challenge is coming up with new ideas which are potentially going to result in outperformance. You’ve got a lot of brokers delivering very similar VWAP strategies, but if they’re broadly built in the same way, you’re not going to learn from that. RBC built an AI Research Institute called Borealis AI, nearly a decade ago. It serves the entire RBC group. We’ve been a key beneficiary of that and we’ve used that expertise to build a segregated algo platform called Aiden. We’ve been running that in North America for over four years and we launched in Europe at the back end of last year.

We go through periods of different types of algos being more or less popular depending on the different challenges or the different trading environments. Broadly, we find VWAP strategies and close strategies being commonly used by the index or quant firms who often will be trading baskets of orders. There’s a huge focus on the closing auction, giving the amount of volume that’s now trading there. From a single stock trading perspective, there’s an enormous amount of fragmentation in the market now. Not just in terms of number of venues, but also the different types of liquidity. Whether it’s SIs, periodic auctions, and lots of OTC liquidity now.

Clients are starting to use liquidity seeking algos a lot more than traditional POV algos because they deem them to be more intelligent. The quant hedge funds will either be using DMA and running their own strategies or basket based algorithms. But single stock investors, where they’re literally trading a single stock at a time, they’ll be looking for something much more liquidity focused.

How is buy-side demand for multi-asset capabilities changing and what is driving this?

More than anything clients are looking for choice. In the equity world, we’ve traditionally had different execution desks – high touch, low touch, portfolio trading – clients are starting to look at that and want to see those same choices across the different asset classes. We’re seeing more and more of our clients working on multi-asset desks.

There’s an element of resource constraints. Firms are looking to do more with less. There’s an element of wanting to learn from the different asset classes and optimise what you’re doing versus best in class across these different asset classes. There’s lots of different examples of that across FX and futures. The latest is probably the ETF marketplace.

ETFs have traditionally been a big RFQ market and it hasn’t really been disrupted for a long time, but more recently, we’re starting to see lots of new ideas around how to change that marketplace. From the buy-side perspective, there’s more and more firms looking to launch ETFs and not just passive but active fund managers as well. There’s a drive to build a more transparent marketplace where they believe that asset class can thrive. Specifically, we’re partnering with a number of firms to build algos that encourage more liquidity onto lit markets, whereas traditionally it’s been RFQ or OTC.

How do you expect buy-side demand for artificial intelligence to develop in the coming years?

Lots of clients are learning about how AI can help them in their own businesses. The evolution of generative AI is incredibly exciting and the scope to adopt these technologies to drive efficiencies and improve service across all sorts of things within financial services is huge. Anything which is repetitive and not really adding too much value, AI can be massively helpful. At the same time, and it has been discussed many times at conferences, we have to make sure that we’re delivering these things in a responsible and ethical way and that you’ve got really strong governance around implementation.

I suspect that AI is very well adopted amongst a relatively small number of our hedge fund clients already. I suspect there’s plenty of clients out there who are already reaping the benefits of this type of technology. But there’s an enormous number of clients that are still trying to understand what the potential is and where exactly they can use it, how they can use it, and what the governance structures are going to have to look like internally etc. Over the next five years, there’s no question that people are going to be forced to take note and figure out how they can use this technology in order to stay competitive.