Corporate bonds: How to automate trading for better returns

Published on 31 January 2019

Automation of corporate bond trading is small but growing, with investment grade instruments – particularly in the US – leading the way.

From auto-quoting to full auto-execution, buy-side firms are trying to take advantage of greater efficiency so they can better support investment objectives and ultimately greater returns for end investors.

In January’s show, Juan Landazabal, global head of trading at asset manager GAM, and Gareth Coltman, head of European product at MarketAxess, given their insights into where and how traders can automate dealing in what is often a highly illiquid market.

Dan Barnes Welcome to Trader TV Fixed Income – your insights into the trading climate for professional corporate bond investors. I’m Dan Barnes. Automation has long promised to reduce the cost of trading, but bond markets are a challenge. In January’s show, Juan Landazabel, global head of trading at asset manager GAM, and Gareth Coleman, head of European Product at MarketAxess, will be discussing when and where traders can use automation to optimize corporate trading on behalf of investors. Juan, welcom to Trader TV.

Juan Landazabel Thanks, Dan. Great to be here.

Dan Barnes So can you tell us why are the credit markets less automated than other instrument classes?

Juan Landazabel To me, it boils down to concentrated liquidity. What concentrated liquidity gives you is certainty of pricing and certainty of execution, or more often than not. And that is something that is not that readily available in the credit markets. So, I mean, if we compare to the US Treasury market fx, and just take a Treasury index; there’s about 8 trillion in terms of volume outstanding versus 5 trillion for the US. But we obviously have one issuer in the US Treasury market, versus around 800 odd issuers in the US IG market. And then in terms of number of bonds, we’re talking about 300 versus close to 8000. That will give you a sense of that lack of concentrated liquidity, and if I can expand that further, so the second aspect that that gives you, is that you have idiosyncratic risk. So, I’m now talking number of issuers, but also number of instruments as well. That means that hedging becomes much harder. And so from a market maker perspective, it’s much harder to provide liquidity because you can’t offset that risk.

Dan Barnes In which parts of the corporate bond markets spectrum do you see automation?

Juan Landazabel First of all, let’s make the distinction between electronic trading and automation, because there’s been a big uptick in electronic trading in credit markets, however, in terms of automation, that process has been very slow. How I define automation is ‘hands-free trading’; where we just send the trade, auto-route it to a venue and it comes back with the order filled. And having said that, we do have some automation and it is evolving. I would say that this started out mainly with the US IG market. If we take the sell-side, it’s only fairly recently that, I would say, a handful of banks have been fully automated from a price-making, market-making perspective on US IG, but it’s not the whole market. It’s a subset of that market and also in small sizes, so 500.000 to a million. And it started in the US because obviously trades provides a decent level of transparency and also the depth of that market is slowly moving to Europe. Some of them are rolling out those algos or similar algos adapting to the European market from the sell-side perspective. On the buy-side it’s been much, much slower and much harder with take-up. We’ve been dependent on trading platforms to develop certain tools that will help us automate good composite pricing and liquidity scores, which would give a trader a sense of, ‘OK, this has a high liquidity score, and therefore has a higher certainty of pricing, of execution and so on.’ Those are the type of tools which have slowly been rolled out by some platforms and which are being taken up by by the buy-side as well.

Dan Barnes So which parts of the trade life cycle can be automated?

Juan Landazabel There’s a lot of scope fx, to auto-routine from your OMS, staging directly onto platforms. You could go one further, as we’ve just discussed, and have those platforms do the auto-execution and therefore you have auto-origination pretty much to affirmation confirmation. I think the key for me is how to automate the high-touch trades, and that’s all about the data. Now, there’s many components which already exist within the firm’s infrastructure. Fx your hit ratios with particular counterparties, you might be getting feeds of axes directly into your auto-management system or onto some other platform. So, that’s one component. The second component really is the calibration of what the fair transfer pricing is, and that requires correlation matrices, volatility assumptions as well, and a sense of historical pricing, and for big trades as well.

Dan Barnes To bring that in then, what does a buy-side desk need to do?

Juan Landazabel To work either with certain platforms or more often than not, as is the case today, to bringing the different components and stringing them together. So ideally, and we’re not too far away, from a from a buy-side perspective, for a large, non-liquid trade, as it hits your blotter, you would have a tool, a pop-up, or an engine you can put your trade through, that would narrow down the suggestion of counterparties. Again it’s based on axes and historical experience as well, and also suggest a risk transfer price to me. 

Dan Barnes Does the automation play to the efficiency of the trading desk or does it actually have a direct impact on the returns the funds can generate?

Juan Landazabel No, I do believe that it helps the returns of the fund. The more you can give time back to traders to focus on the harder trades, or support traders with the harder trades in terms of pre-trade, due diligence tools that would lead to better execution, and hence would help the investment process. And if the trader has half a day left not doing anything, because everything is so efficient, hopefully that time is invested again. So, it’s talking to the PM’s, understanding what they need, what are they thinking, what to go out to the market to look for, and engaging with counterparties and so, to become more of an expert in that sector and contribute to that investment process.

Dan Barnes That’s great. Juan, thank you very much.

Juan Landazabel Thanks very much, Dan. It’s been a pleasure being here.

Dan Barnes Now, Gareth Coltman, head of European Product at MarketAxess, is going to give us his insight into how traders can deliver better results for investors. Gareth, welcome to Trader TV.

Gareth Coltman Thanks, Dan. Pleasure to be here.

Dan Barnes What are asset managers trying to achieve by automating the bond trading process?

Gareth Coltman Fundamentally asset managers are looking to reduce cost and increase efficiency on the trading desk. It’s very challenging market conditions at the moment, a very, very competitive environment for them, so they’re looking for ways to automate manual processes and free up their traders to focus on the higher value trades.

Dan Barnes And how involved is automation in the trading space?

Gareth Coltman So automation, particularly inside credit trading is very emergent at the moment. We’ve seen automation in a number of other fixed income areas for a number of years, fx government bond trading. But one of the challenges in terms of automating credit trading particularly, has been a lack of an available reference price. So, in other markets we’ve seen that this is available fx in equities where there’s a best bid and offer in the order book. But the lack of a kind of centralized order book inside fixed income means that the clients have to find other ways to fill this gap.

Dan Barnes Whereabouts in trading is automation most effective right now?

Gareth Coltman So, we’re seeing clients experiment with automation in a number of different product areas. So we have clients using automation fx inside our US high grades, our European credits, but actually also in emerging markets and some clients in high yield as well. Typically clients are more likely to use automation on the more investment grade end of the credit spectrum, and typically in the more liquid products and in smaller size today.

Dan Barnes How is regulation either hindered or helped the automation of trading?

Gareth Coltman So recent regulation, like MiFID II, has really accelerated the electronification of credit markets, and I think automation is really an inevitable consequence of that electronification. The increasing transparency has obviously created a lot more availability of data, and that’s really helped to drive the demand for the automation side of credit trading at the moment.

Dan Barnes What tools do trading desks need in order to evolve automation within their firm?

Gareth Coltman So, what we’re seeing with the clients that we’re working with, is they’ll initially set up their automation capability, embed their ruleset and best execution policy into that automation, and start off with a relatively low bar in terms of what they’re trying to automate, so fx very liquid or small size. And then as they move forward, what they’re looking for is feedback in the form of a best execution analysis, which they can then use to look at the performance of automation and to increase those levels to larger size tickets, more liquid credit. And that’s my expectations as we move forward over the next 12 months, is we’re going to see clients trying to push the boundaries in terms of what they’re doing around automation.

Dan Barnes Do you see any particular types of funds automating their trading process more than others?

Gareth Coltman Automation is probably most suitable for those funds who generate a large amount of smaller sized orders. Fx firms which have passive strategies or firms which look after a lot of private wealth individuals, fx, but we are actually seeing take up automation across the board in all types of asset manager, where they’re finding that even if they’re not using automation every single day, where they have a significant increase in the flow that’s going through the desk, fx inflow or rebalance, then automation is available to help them at those times.

Dan Barnes Finally, when clients are looking to a corporate bond trading platform, what sort of questions should they be asking of that platform?

Gareth Coltman When I talk to clients, they’re really focused on three things. And the first one of those is really about access to liquidity; it’s very important that through automation they get the same access that they could get on a manual basis. So, fx if their best execution policy allows them to look at all-to-all trading opportunities on a manual basis, they’d want to find a provider that can allow them to embed that on an automated basis as well. And the second thing is really about data; they’re going to be looking for providers who can provide them with reference pricing, liquidity score information fx. And the third thing is really tools to allow them to monitor and manage their automation in real time as it’s going on. So, fx to be able to see in-flight inquiries, see what the automation system is doing, look inside the machine, so to speak.

Dan Barnes That’s been great. Gareth, thank you very much.

Gareth Coltman Thanks very much. Dan.

Dan Barnes I’d like to thank Juan Landazabel of GAM and Gareth Coltman of MarketAxess for their insights into automating corporate bond markets. And of course you for watching. To catch our monthly reports on other markets or to subscribe to our newsletter, go to TraderTV.NET.