Ali Curi: [00:00:00] Hi everyone and welcome to ION Markets Quick Takes. I'm Ali Curi and every week along with my guests Amir Khwaja and Chris Barnes, we take a quick dive into the headlines on the Clarus blog. Let's get started. 

Hi Amir. 

Hi Chris.

Amir Khwaja: Hi Ali. 

Chris Barnes: Hi Ali. How are you doing?

Ali Curi: I'm doing great. Welcome back to Quick Takes.

Amir, let's start with you. What's your Quick Take for this week? Which headline from the ClarusFT blog would you like to discuss? 

Amir Khwaja: Sure, Ali, so I'm going to talk about "1Q24 CCP Volumes and Share in IRD." This is a blog I do every quarter and it follows the same format and I look at, the volume in the most recent quarter compared to a year earlier and the prior quarter.

I'll just say a brief few words about the content and I'll talk more about how I go about doing the blog. First quarter 2024 was a record quarter for many currencies for dollar, for euro, for yen. So it exceeded the COVID quarter; first [00:01:00] quarter 2020. So again, very high volumes in trade swaps that are cleared as, as we near a path of, "will rates start to drop or not," that uncertainty is being, higher for longer.

I covered that, but I think mostly today I really want to talk about how I go about doing, writing the blog. And I think this is one of those blogs, which is like, we all like habits, and which also, it's very comforting to do the same thing on a regular basis... update the numbers, and I do it in a way where it takes me time to do it.

So I log into CCPView, take a screenshot, I annotate the numbers and I take another screenshot and take numbers in terms of what's changed, look for trends. That takes a fair bit of time, right? Now I could automate it a lot more. So we have an API, we have spreadsheets, and I could highly automate that process that it pretty much created the visuals and the numbers that I highlight.

But I think in doing that, I think I would lose the time that I spend doing it. I would lack the understanding of it, right? I might publish something in a shorter timeframe. The process of writing will be quicker. But my understanding would be a bit less, [00:02:00] right? Because I would invariably get lazy and not read what it produced, right?

When you automate a lot, you tend to not read what it produces in a detailed fashion. When you have to write it, you spend time making sure every line is correct, and you read it again, etc. So that's something I find, worthwhile doing. Also, our customers at UCCPView get the exact same screenshots that I'm showing them, so they can log in and see the exact thing without having to set up APIs, and which can be customized, right?

It'll be a bit different, right? So I think that's important. I've thought about automating it, but I always think it's not worthwhile, given what I've spoken about. I did this time try and improve my summary, so I did ask Google Gemini to give me a hundred word prompt summary, and it did a reasonable job, which I have in the blog in italics.

It managed to pick up all it by itself from that content that Euro IOS volumes were now almost equal to U.S. dollar volumes. So that's something that found itself. And it found that LCH share was dominant, with some exceptions. So again, so that's interesting you picked those things [00:03:00] out, right? In terms of all the text that was in there.

Chris, I was going to just ask you whether you had any questions. 

Chris Barnes: Interesting on two points, because I must admit, when I'm presented with the formulaic type of blogs, I tend to skip the intro and just go straight to the charts, right? I miss the fact actually that Google Gemini had picked up on the combination of Euro OIS and Euribor is actually bigger than dollars.

So that was going to be like my first question on the podcast. Gemini is making me feel a little bit redundant. What you say about having to go through a process to write the blog is exactly right. When we were producing the RFR adoption indicator, we automated all of the process behind that, but every month we had to send the figures to ISDA.

And you can't just send a spreadsheet. You've got to say, look, I've "sanity checked" it. I've looked at it. And so doing that type of manual [00:04:00] reconciliation gives you a much better overview of what the data is showing. Having said that, of course, what we could start doing is sending Gemini those screenshots that you take of those charts and asking it to summarize the chart in words.

And given that it's picked up on the fact that euros is bigger than dollars, it may pick up on stuff that we've missed as well. So looking at using that as like an additional writing tool. However, from your blog, one thing that it can't pick up on is what's happened in yen swaps. I think. I don't think we've seen a big change in market share necessarily, but I think our next podcast is going to focus on yen swaps and a big part of that will be actually what's happened in the CCP basis in Japan, right?

We've seen the long end of yen CCP basis between JSCC and LCH go negative for the first time and go negative by a relatively large amount. [00:05:00] Do you think combining that element of like market color into this blog, is that achievable? Or is the purpose of the blog slightly different to only talk about the data?

Amir Khwaja: Yeah, it's a good point, Chris. I think our blog readership expects this sort of blog, which is purely a data, numbers. I do add a very brief line about commentary, but not too much. And I guess, the argument about, does changing basis influence market share is unproven so far?

Chris Barnes: Yep. 

Amir Khwaja: I think we can see there are reasons why it persists and it may or may not change market share. Clearly, if you did automate more, you could spend more time on, on commentary. Here, the commentary is really the numbers. And I expect people to think about numbers and form their own. But over time, yeah, I think it's a mix, right? I like your point about an AI, a large language model, given the images, could annotate language that would pick up things that we hadn't picked up. 

Chris Barnes: A lot of the art of writing the blog is deciding [00:06:00] which charts to talk about. There are a lot of people in the world who don't have the blog writing skill, but they might be able to pick out those charts.

It can be a very powerful tool for them, I think, to say, look, these are three charts. I want to structure a story around those three charts because what they're showing is inherently interesting. "AI, can you create me a blog?" 

Amir Khwaja: And I think that's interesting. And also, if you use prompts, it's very much a recursive process, so it has context and you can dig deeper and you can ignore some of the responses.

So it almost becomes like a dialogue with some expert chart reader, or sub chart image reader, then you can try and get into it. Of course it can lead down the wrong path, but I think it's interesting. So I think that's something we plan to explore more, right? So I'm interested in that.

Chris Barnes: It's a bit like that guy on the trading floor and and I hate to say it, it's always a guy, right? And you're like, I have this really interesting chart, but what does it mean? And you go and speak to him and he wants to talk to you about this chart for three hours. And you're like, no, no, no, [00:07:00] no, I just, I wanted like a three minute like download.

Amir Khwaja: Back to you, Ali. 

Ali Curi: Great. Thank you, Amir. And please share with us again, the title of your blog post. 

Amir Khwaja: So it's "1Q24 CCP Volumes and Share in IRD."

Ali Curi: Great. That works. 

Amir Khwaja, Chris Barnes, thank you both for sharing your Quick Takes.

Let's do it again next week. 

Amir Khwaja: Thanks, Ali. 

Chris Barnes: Always a pleasure. Thanks, Ali. 

Ali Curi: And that's our episode for today. You can read more about these topics on the Clarus blog, and you can follow ION Markets on X and on LinkedIn. Thank you for joining us.