Bloomberg has developed pre-trade analytics to support transaction cost analysis and execution quality analysis, which can be embedded within existing trading tools and workflows. Not only are these an assist to high touch traders, explains Ravi Sawney, global head of trade automation and analytics at Bloomberg, but they can support the transition to greater trade automation for bond traders.
Speaking at the Fixed Income Leaders Summit in London, Ravi talks through the latest evolutions of Bloomberg’s analytics, their integration into BTCA for post-trade support and the latest research on bond trading automation, ‘Taking Trading Automation to the Next Level’ from Bloomberg and WBR which you can access for free here.
Dan Barnes: Welcome to Trader TV – your insight into trading for professional investors. I’m Dan Barnes here at the Fixed Income Leaders Summit in Europe, and joining me is Ravi Sawney, Global Head of Trade Automation and Analytics at Bloomberg.
Ravi, welcome to the show.
Ravi Sawney: Dan, thank you for having me.
Dan Barnes: So you guys have had a very big year for new product launches. Can you tell us what are you finding exciting at FILS this year?
Ravi Sawney: Yeah, you’re right. We’ve had a really exceptional year. We started the year launching our Bloomberg Rule Builder (RBLD) automation product for FX, for ETFs. So that now compliments our fixed income and equities offering. So we’re really excited about, first of just meeting people face to face at FILS, of course, like everyone else.
We are launching our conditional release offering on our Rule Builder product which has been a long requested feature from our customers. So this will allow clients to trigger automation, based on conditions of time. So fx, if you are trading in Asia and you’re trying to reach liquidity out of hours in the US, or they’re trying to reach a particular market close.
And we’re also, in parallel, launching the ability to release orders against limit price. Our machine will basically continuously evaluate your limit and see if the market is close to your limit, and only then will it actually send it out to the market.
Dan Barnes: Traders are typically comfortable in automating trades up to a certain size. Are you going through any processes of expanding their view of that?
Ravi Sawney: Yes, we’re trying to. I think, you know, we are at the stage now where people want to embrace automation on the trading desk. They are looking for ways to increase the percentage of orders on their blotter that can be eligible for automation. And we’re looking to find ways to support our traders to do that.
One of the ways we’re trying to achieve that is through the use of TCA and pre-trade TCA. So fx, today you might have a rule that says, ‘automate everything up to five million in euro credit.’ Quite simple. But actually with TCA, you can actually ask more nuanced questions. You can actually look, fx, at the pre-trade market impact cost of that particular bond. You can look at the probability of execution. So these are new data points that have only been recently made available to our market participants, and we’re going to be integrating into our Rule Builder product.
Dan Barnes: Well, that’s really interesting because best execution in fixed income is not just about price. It can be about, you know, liquidity available in the market and whether you can execute the whole order at one time or the speed at which you can execute an order. So you’re adding those analytics into pre-trade so that traders can actually get that concept and then make a decision about where to execute?
Ravi Sawney: Yeah, exactly. These analytics, as you’d imagine, on Bloomberg we provide it on a screen and the terminal, people get comfortable with the analytics. They understand the data points, and at that point, we can start to integrate it into the trading ecosystem. So the fixed income TCA project was a big launch for us this year. We’re just starting to see the benefits we launch integrated into our BTCA product for post-trade TCA, and then Rule Builder will be next to help drive trade automation.
Dan Barnes: And which instruments and asset classes does that cover?
Ravi Sawney: So this covers fixed income, cash and corporate bonds.
Dan Barnes: When we think about automation, data is absolutely key, but often traders and asset managers want to use it in their own particular way. How are you catering for that?
Ravi Sawney: Yes, good question, Dan. For us, it’s about giving our clients choice. If they want to use the data point on the terminal or if they want to take it outside of the terminal and use it externally, fx in a third-party BI tool, if they want to have one of their quant traders evaluated in a Jupyter Notebook in Python. You know, we’re offering greater and greater choice in terms of how clients can use data through the Bloomberg platform.
Dan Barnes: That’s very interesting because we’ve just had somebody on the panel here at the FILS event talking about the way that older traders think about bringing data into Excel spreadsheets. Younger traders think about bringing data into Python. Do you think the flexibility you’re providing is perhaps catering for this wider generational shift that we’re seeing?
Ravi Sawney: Yeah, I think that’s an interesting way of looking at it. Definitely, you’re seeing the makeup of the trading desk, especially the buy-side change, right? So you’re seeing kind of more people coming in with a computer science background and they traditionally are more comfortable in Python than, say Excel.
So again, it goes back to just giving them choice. They are absolutely nothing wrong with Excel. I use it a lot, as do most people. But you know, if they want the flexibility of looking at their data in Python, where you know, it’s easier to bring in other datasets on top of the ones that we provide, then that’s an option.
Dan Barnes: And are there any other unexpected benefits to automation other than, say, just reducing cost?
Ravi Sawney: If you look beyond the time saved, as you would expect, you know, increasingly clients are looking to industrialize their trading workflows. Especially, you know, clients operating across multiple trading desks, but they have different processes in place. Automation really is presenting an opportunity for them to re-engineer the process first before automating it.
If you take a bad process and just automat it, it’s just a bad process being automated. Is this re-engineering AI industrialization? I think that’s what a lot of, especially global desks, they’re asking questions like, ‘why are we doing this differently? How do we consistently give clients best execution?’
Dan Barnes: So automation is creating the context for them to think about how to re-engineer their processes?
Ravi Sawney: Absolutely, absolutely, that’s definitely happening right now with our clients.
Dan Barnes: Do you think that the general angle of that is about efficiency or is it perhaps completely rethinking how they go to different trading partners? What sort of shape does that take?
Ravi Sawney: I think, you know, there’s an efficiency play. There’s a consistency play in terms of how they are delivering best execution, rationalizing their cost base as well. Does it make sense to have people doing various different things around the globe? So there’s definitely a multitude of benefits. We recently completed a survey in conjunction with WB, and one of the questions we asked was around the benefits that people are getting from automation. A lot of respondents, one of the biggest strengths they saw in automation was that it acts as a great tool for retention in hiring.
Dan Barnes: Oh, that’s really interesting. So is that because they’ve got those younger traders coming on who actually want to engage in the process of engineering that?
Ravi Sawney: I think that’s definitely a key factor, right? Nobody wants to join a trading desk now and be left to automate small clips in US Treasuries and government bonds, right? That’s not rewarding for anyone when there is technology available out there to automate it.
That’s a big shift in my mind from many years back when we’re starting out the conversation around automation, and there was a kind of a fear element to it. Now it’s actually expected.
Dan Barnes: That’s great. Ravi, thank you so much. Thank you.