Traders who code
===

Ali Curi: Markets ConversatION is an ION podcast where we discuss topics of importance to capital market participants with product owners, subject matter experts, and industry leaders. 

Tim Flinders: The actual nature of machine learning has been around for a long time, but the key thing was that required highly skilled individuals to work with this technology.

However, these new large language models that make this technology more accessible, a relatively inexperienced user can write code to extract or manipulate data. They can create a chat in a AI assistant and write code that will do the task. 

Ali Curi: Hi everyone, and welcome to Markets ConversatION, I'm Ali Curi.

On today's episode, Steven Strange and Tim Flinders from ION Markets will discuss "Traders who code." The new and exciting intersection of finance and technology. Traders who code are traders who combine their financial expertise with coding to build powerful tools like algorithm and AI driven strategies.

This shift is transforming how trades are made, allowing traders to backtest strategies, automate decisions, and respond to market changes faster than ever before. We'll dive into these innovations and explore how coding is reshaping the future of trading and what the industry can expect from this trend.

Let's get started. Steven Strange. Welcome back to the podcast. 

Steven Strange: Hi, thanks for having me back. 

Ali Curi: And joining us again is a friend of the podcast, Tim Flinders. Hi, Tim. 

Tim Flinders: Thanks Ali. Thanks for inviting me back. 

Ali Curi: Welcome back. 

Steven, you're a regular on the podcast, but for our new listeners, can you briefly share your background and your role at ION?

Steven Strange: Sure. I head product for the asset management side of ION Markets. So responsible globally for the buy side product management. That covers portfolio management tools, trading tools, execution, and investment compliance. I've been working in this industry, the financial industry for over 15 years. 

Ali Curi: Great, thank you. And Tim, please tell us a little bit about your background and your role at ION. 

Tim Flinders: Sure. I lead the technology and operations functions globally for our asset management group within ION Markets. I've got around about 25 years experience in technology and the majority of that is with asset managers covering everything from front office to back office.

The main focus is the Latent Zero product set at the moment, focusing on buy sides, but also cloud adoption within ION Markets. 

Ali Curi: Great. Well, let's dive in. Steven, can you give us an overview of this topic? This is a curious topic, "Traders who code." Why are traders increasingly turning to coding? Because coding isn't typically associated with traditional trading. So, what motivates them to learn coding? What's driving this shift? 

Steven Strange: Of course, so financial markets have become more data and technology driven than ever, and there's been an increasing trend to consider the technology that's available to you. But I wish to take this further and understand how can it improve your workflow and decision making beyond what is provided out of the box from software vendors.

What I mean by this is how can I use programming skills to analyze data, backtest strategies, or understand quant models, for example, without being reliant on tech teams or vendors. Essentially, how can I be empowered to do more with the technology available so I can create more positive trading outcomes? That's really the emphasis of it all. 

Ali Curi: Okay, large institutional traders use algorithmic trading, or algo trading. Is that what this is, or is it an extension of this? 

Steven Strange: Yeah, so, not quite. So, algo trading has been around for decades, however clearly has evolved with the advances in technology over the years.

It relies heavily on coding to build, test, and deploy algorithms that can fully automate trade execution based on a wide range of variables. It's used heavily by quant teams who are highly technical and very experienced in this area. Although the concept of traders now being able to code can be seen as an extension of algo trading, in the sense that more traders now have access to open source tools, predefined libraries, making this process much easier.

It is not the only reason these users are learning to code. I would see the algo trading as a subset of how coding can be applied to trading by the end traders themselves. There's a much broader set of use cases, as mentioned previously, whether it's analysis, backtesting, trend analysis. creating custom strategies, communicating information with your peers through data visualization.

There's many more use cases beyond the traditional algo trading that most of us are familiar with. 

Ali Curi: I see. You talked about these new tools and technologies that give traders an edge compared to traditional methods, also referred to as alpha skills, right? The intuition, the creativity in the marketplace. Do they help? Do they give these traders an edge compared to traditional methods? 

Steven Strange: Yes, they do. It allows traders to process and analyze data much faster than traditional methods. So you can identify trends, you can act quickly. Repetitive tasks can be eliminated almost entirely, so you can concentrate on deeper work.

There's less reliance on kind of off the shelf trading strategies. So you can optimize your own strategy, you know, test it to see if it works, what works best for you or your firm. I'm sure Tim will cover in more detail, but access to these tools and the knowledge is more accessible than ever. So it's important to be able to keep up to date with these.

Otherwise, competitors will gain an edge if you're not leveraging these tech skills. 

Ali Curi: Right. Yeah, that makes sense. Tim, let's come over to you for a minute. Can you tell us about the role of AI and machine learning influencing the landscape for traders who code? 

Tim Flinders: Yeah, I think when we talk about AI machine learning nowadays, the majority of us are thinking about the tools like ChatGPT or perhaps Microsoft Copilot.

These are chat interfaces that are appearing everywhere. And like Steve says, that the actual nature of machine learning has been around for a long time, where you've had sophisticated quant funds might use it. But the key thing was that required highly skilled individuals to work with this technology.

However, these new large language models that make this technology more accessible. So given the right infrastructure, such as open systems, open data platforms, a relatively inexperienced user can write code to extract or manipulate data. They can create a chat in an AI assistant and write code that will do the task, which might extract the data or even create a graph or some other level of analysis.

Ali Curi: Tim, let's continue with you for a minute. Let's talk about the business side. Because coding has enabled entrepreneurship among traders, what can you tell us about these types of ventures? 

Tim Flinders: From a business side of things, it allows traders and fund managers to research their own theories and hypotheses without needing outside assistance.

Increasingly, we've seen software developers working on the trading floor. But they often will need help from traders to actually understand the nature of trading. And in some cases, the requests from the traders have had to go to separate IT teams, and that causes delays and discussions, which sort of holds back the entrepreneurship.

It might be that data generates more questions and queries. Allowing direct access to the information can reduce the time delays and inherently unlocks the ability to see things differently and quicker. 

Ali Curi: Okay, I'm curious about the specific programming languages. Do those make a difference? 

Tim Flinders: Well, typically we've seen that scripting languages are used more than compiled languages because they're sort of easier to get started with.

Most common is probably Python, and it tends to be the preferred option. Typically, the tool is also used by data scientists, so using the same tools means they can achieve amazing results. It's relatively easy to get data out of it and analyze it and tabulate it and perhaps chart it using Python and Jupyter notebooks, and do that in a repeatable way.

But honestly, the language doesn't really make a difference, but the easier the language and the quicker the ability to get results is key. The code is part of organized operational processing. The choice might be important, but typically firms aren't allowing traders to write code that is part of the order processing workflow.

It's typically analysis at this point due to fund compliance requirements. 

Ali Curi: Okay, thank you. I appreciate you explaining the technical part a little bit better. 

ION Ad: This episode is brought to you by ION. At ION, our clear derivative solutions automate your complete trade lifecycle and deliver actionable insights whenever and wherever you need them.

We offer execution and order management, post trade processing, and a complete front to back business solution. To learn more, visit us at iongroup.com/markets, or email us at markets@iongroup.com. 

Ali Curi: Steven, as more traders use AI and automation, some concerns are being raised about transparency, and how algorithms make decisions.

How do you both see this issue being addressed? Let's start with you, Steven. 

Steven Strange: So whenever AI or automation is discussed, the biggest concern is always transparency. How was the decision made? And what were the steps to making such a determination? This isn't a new concept for financial markets, which are highly regulated.

There's internal and external auditors always asking this question on transparency. Therefore, before implementing any automation or AI tool, you have to be able to demonstrate the results, and the transparency needs to be part of the solution. If you cannot clearly explain and show detailed audit information, then the tool should not be in production, regardless of who created it.

Tim Flinders: Right. Tim, what are your thoughts? I agree with Steve. A detailed audit of the information is definitely necessary and, you know, how the technology is used to extract the data and audited is necessary and not locked up in a single trader's head. And obviously we've mostly talked about codes to, to extract data, but it's also possible with these tools to, to analyze the data and make sense of it.

But I think with these new large language models. Techniques around security are needed to know that, that you're not letting data outside of your firm and information leakage, come into play. Funds do need to make sure that there's clear policies on that. So, so we don't allow information to get into the public domain that shouldn't do.

Ali Curi: Great, thanks Tim for that explanation. Steven, what are some challenges that traders face or could face when they start coding applications? Are they outliers or is their boutique approach more of an advantage? 

Steven Strange: Yeah, sure. There's certainly challenges that can be faced. You know, some kind of obvious ones, lack of training or technical skills that end up producing inefficient tools that are actually counterproductive as you start to build out these applications.

Data will always remain critical. If the data available to you is poor or incomplete, this immediately causes problems right away. So that is a challenge. That's why it's always important to have data governance in place. This isn't simply just empowering your traders by providing programming courses and training.

It's more about having a firm-wide backed approach that includes data governance and being able to ensure that the whole firm really understand the process. So understanding what can and can't be done from a legal and reg perspective. So yes, it's great to analyze data, extract information, potentially build out some custom trading strategies.

But you need oversight from the firm to understand what is actually acceptable and finding the balance between this productivity and having an edge versus creating inefficiencies and adding distractions to the trading process really is key. And I see that's the main challenge. 

Ali Curi: Okay, let's dig a little deeper because AI is currently used in backtesting and in risk management strategies.

Do you see these tailor made apps contributing to these areas as well, or are they more niche? Are they more trade dedicated? What are your thoughts? 

Steven Strange: Yeah, no, they certainly can be used for backtesting. They will use AI capabilities to enable that, and there's a lot of open source backtesting solutions available that, again, provides accessibility.

But it all goes back to the data part, as I mentioned, that if you have incomplete or poor data, that's going to be a challenge. You can add all these new apps and do back testing, but it's going to produce results that are potentially inaccurate. And clearly that's not good. So again, you do need to ensure that the data governance is, is in place. It's essential. Otherwise, you won't really get much advantage by, by looking at backtesting and using these tailor made apps for that. 

Ali Curi: Great. And now this next question is for both of you. What do you think is the future outlook for traders who code? How might this trend shape the finance industry in the coming years?

Steven, let's start with you. 

Steven Strange: Sure. The trend of financial institutions becoming technology firms has been on the way for many, many years. You know, most, Asset managers have very large technology teams or are very technical resources. However, with an increased number of tools now accessible to end users, it really does open up the opportunity for firms to carve out a competitive edge.

We've touched upon a few examples of what use cases could be solved, but in a similar way to AI, the possibilities are endless. The most important point is to ensure there is a governance in place, adequate time spent defining what are you trying to achieve with this new tech. Adoption of these new tools for investment and trading teams will only continue to grow, but you do need to pause and figure out, well, what is the value that you're trying to add to your, your workflows?

Ali Curi: Right, right. Tim, please share your thoughts. 

Tim Flinders: I think we're just at the stage of dipping our toe in the water for the industry. Business value is key. And if traders can be, get improved value and better returns for the fund, perhaps lower costs, this will certainly accelerate. And whether this is just hype, the effect could be similar to when in-person trading moved to electronic trading.

Some of the possibilities could even include complete automation of the trading process. But business outcomes will drive this, and I think full automation in the short term will be very challenging for the majority. That the knowledge and the relationships of the trader are key, and that's not something you can get straight from technology.

But if we can capture some of the essences of that, we can get more automation than we have today, and perhaps better value. 

Ali Curi: I think, that is a great take. Thank you. Again, a question for both of you. What's the one big thing that you hope listeners will take away from the episode? And Steven, we'll start with you. 

Steven Strange: Sure, I think the environment we're in is changing and providing more opportunity than ever. So it really motivates you to consider how can you maximize efficiency, productivity. Find that competitive edge with this fast growing technology set that's available. As I mentioned before, I think the key takeaway is pause, reflect, consider how to utilize this rather than being swept up in the buzz of it all.

Ali Curi: And Tim, what do you hope listeners would take away from the episode? What's the one big thing? 

Tim Flinders: I think information is key and accessing that as quickly, as efficiently as possible is really the message. Using these new AI tools can unlock individuals' ability to do tasks that they could never do before.

That said, this topic is quite early on for most fund managers and traders, and the capability to implement these techniques is just starting or in the planning phase. Though I must say, it's amazing what you can see from early adopters have achieved with the technology. 

Ali Curi: Yes, certainly. I think it's something that might merit a follow up episode in the future.

Steven, I've asked you about career advice in previous episodes. My question for you today is, what's one lesson your job has taught you that you think everyone should learn at some point in their life? 

Steven Strange: Sure, yeah, being adaptable is essential, as things often don't go as you envisioned. It's important to remain calm, find a path forward.

And whether that's through upskilling, kind of problem solving, building a mindset that helps you keep resilient and more open to growth. I think that's really the key lesson there. It's not about going with the flow or reacting to changes as you see different obstacles. It's more looking at an opportunity to learn and grow as we are in this ever changing environment.

Ali Curi: Okay, great. And Tim, same question, what's the one lesson your job has taught you that you think everyone should learn at some point in their life? 

Tim Flinders: Yeah, that's an interesting one. I think I've learned over the years that nothing is constant and everything is evolving in technology. I think I would probably advise that you embrace continuous learning and training, regardless of your level, and stay well informed and relevant.

If you aren't doing that, somebody else will be doing that. So, and you become less relevant over time. Treat every day as a, as a school day and you'll keep sharp. 

Ali Curi: I think that is some great advice from both of you. Thank you. Steve Strange, Tim Flinders, thank you both for joining us again on the episode. I look forward to our next time.

Steven Strange: Thanks, Ali. 

Tim Flinders: Thanks, Ali. 

Ali Curi: And that's our episode for today. You can follow ION Markets on X and on LinkedIn. Thank you for joining us.