Jon Light Talks The Latest in CFDs Technology at FMLS

Jon Light, Devexperts VP of Trading Solutions, took part in the “Cutting Edge: The Latest in CFDs Technology” panel discussion at Finance Magnates London Summit 2021 and gave his perspective on the latest innovations in the retail brokerage industry, how brokers can use collected data, technology cost trends and more.

What are the biggest innovations that happened in 2020 and 2021?

Over the past two years, the number of beginner retail traders without any experience in capital markets has grown tremendously. With that, many brokers are demanding things like built-in educational tool tips and videos. We have a broker who built pure process flow, taking a retail client through the charts and order tickets to demonstrate that process. Many brokers also want to have simplified platforms. Certainly, with crypto platforms, people want to log in, buy, and sell. So, many of our brokers have split the platform into a professional-looking view and a simplified one.

Also, over the past years, more and more brokers have become multi-asset, and integrations have become the new space race. It’s vital to be able to integrate quickly, especially with crypto. You might be a broker that’s set up the top 10 cryptocurrencies you expect your clients to trade, but it just takes Elon Musk to tweet overnight, and the next morning your clients are demanding an old currency you didn’t have. Depending on your setup, you might need to integrate a new custodian wallet or liquidity provider. If you don’t do that quickly, your clients will go elsewhere to trade that crypto.

I first recognized a real change in the type of trader signing up while at a family barbecue. My older relatives asked me what platforms to use and when and how to buy crypto. It’s not something I’d have expected two or three years ago.

How do brokers utilize the data? Are they effectively utilizing it? Are there better ways to do that?

Collecting data is expensive, and the data’s useless if you don’t have anything to do with it. Surely, you can use the data for obvious things, such as transaction-cost analysis. If you’re a broker with multiple liquidity providers, you want to analyze the reject ratios, fills, and latency. Then, you’ll compare that to the trading cost with a particular liquidity provider. 

We’ve been working with multi-asset brokers who use many different platforms and want to use their data to cross-sell to their existing clients.

Using the data, we want to: 

• Bring the data into one place from different platforms
• Have an engine that can analyze the patterns
• Curate specific content for their clients
• Deliver that content over email, banners in apps, and through the chatbot

We’ve also spoken to some brokers about how they can utilize real-time brokers. And we discovered that they use the data in some very useful ways, like analyzing: 

• The markets
• Their clients’ positions
• Their positions
• Pending orders

They can also make real-time changes to trading configs for:

• Spreads
• Routing to different LPS
• A-book and B-book functionality

But, as I said, collecting data for the sake of it is very expensive.

How do technology needs change as the brokerage company grows its business and books?

It’s difficult for startups to look ahead and know where they’ll be in two to five years. You don’t know what’s coming, and the market has experienced many changes over the past two years with the move to crypto. Of course, startups want to start making money as quickly as possible. But you must consider things like offering white labels in the future or looking at new asset classes. Adding crypto CFDs to an FX/CFD platform is relatively easy. You’ll experience differences in financing, as more standard financing is required. But it’s difficult to add a crypto spot on an FX platform because it has multi-currency accounts.

I’ve also seen large retail brokers have become successful establishing large businesses and infrastructure. Then they want to move into the B2B space, offer fixed connections to their clients, and have an institutional UI. With a retail-focus platform, it’s challenging to add that. So they need a different platform with a separate team to monitor it, which can become very expensive.

What are the cost trends? Why is technology not getting cheaper?

While hardware costs may be decreasing, or are becoming more performant based on what you pay for, the overall cost to provide technology is increasing. Over the past few years, the demand for developers has gone through the roof. It’s hard to find these people, and their cost has risen. We’ve got nearly 700 engineers, and we’re continuing to grow. We have to look at imaginative places to open new offices to get talent pools that haven’t already been tapped. That also comes at a cost. 

Five to ten years ago, brokers were only using physical hardware. But now, this all moved to the cloud. Initially, it looks cheaper, but when you add the functionality you get with the cloud, e.g., disaster recovery and higher availability, it adds up. It likely isn’t cheaper, but you do get a better service.

We offer a crypto exchange, but ours is built on financial-grade technology used across the market. Many crypto exchange startups are run from someone’s basement and offer cheap software. We have clients that talk to these startups and then come to us – there’s obviously a massive delta there. So, this is an interesting conversation we have with many clients, explaining the differences between an established provider and some startups. 

Some built Bitcoin exchanges on fantasy playing-card exchanges. Mount Gox was originally made for transferring playing cards. It was adapted for Bitcoin, but it still had so many issues: downtime, hacks, etc. So, we often have conversations explaining the differences in software costs.

Many tech providers in the CFD industry try to build a captive ecosystem where they control the maker, taker, and everything in between. What do you think about that? 

These closed ecosystems have some advantages: time to market, quick to set up, and familiarity. End clients know what they’re getting and what’s on offer. Being locked in, though, is the biggest disadvantage. If you want a new LP that can’t join the ecosystem, doesn’t want the ecosystem, or you must pay to play (and you don’t want to), it’s challenging to get it in.