What the biggest challenges for your fixed income trading desk and where are the opportunities?
Lee Sanders: We’ve done a lot of work recently on our use of high- and low-touch execution channels. We’ve long recognised that traders do not need to take the same degree of ownership over certain orders to guarantee the best outcome for the client. We will start in the FX and listed markets and will follow with cash bonds. We have been exploring opportunities to route trades from our order management system (OMS) to the execution engines of major fixed income trading platforms and ECNs, using appropriate parameters. An important element is ensuring our fund managers are comfortable with the criteria we’re using to direct trades to different execution channels.
There are scale benefits to trade automation. Intermediation with banks via ECNs can allow our traders to put on more trades more quickly, but a more important driver for us is securing access to sell-side balance sheets when executing bigger trades.
In parallel, one of our biggest priorities is making more efficient use of data to determine the likely cost/benefit profile of a trade. We’ve been working with banks, ECNs, and other data providers to pull in more data, both to support our ability to assess the potential cost of a trade, and for use in our post-trade transaction cost analysis (TCA). These metrics are used in RFPs to demonstrate to clients our ability to deliver a good outcome when implementing investment ideas.
On a practical level, we feed these external data inputs into our proprietary pricing system, which generates an estimated aggregate cost for a portfolio which then determines whether we go into a single-name trading mode or start looking at a portfolio trade. We still rely heavily on the skills of the trader to get inside the bid-offer spread on single trades, or to negotiate with a bank to achieve the aggregate cost we’re looking for on a portfolio trade.
Over the years, we’ve built a substantial pricing universe which we use to assess over (or under) performance of individual trades and now have a 5-person business quality team attached to the trading desk. We continue to look for new efficiencies, for example automating the interaction between our pricing system and OMS.
The key aim is to use all possible relevant information to anticipate the execution cost and inform the trading decision. In the long run, it may be possible for artificial intelligence to interpret the data inputs effectively, but for now you still need an experienced trader to analyse and understand the market.
How are you evolving the skill set on your fixed income trading desk?
It’s increasingly important that traders are able to set up systems optimally and to adjust rapidly to changing circumstances. We don’t need coders on the trading desk necessarily, but we do need access to coders. Data helps us to predict trading costs. So especially for more passive mandates, the trader needs to be more agile to pull in a whole different set of wish lists, axes, and prices.
Even large asset management firms are running fairly lean trading desks, which means we need our traders to be equally comfortable doing an automated or a voice trade, giving market colour, filtering information, interacting with the market, and above all guaranteeing the best outcome.
We are keen to increase automation levels, but the priority remains the best outcome for our clients. Rather than being too rigid, we have to empower the trader to make the decision. If the trader thinks it’s not the right situation for an automated trade, they should be able to work it in a different channel. If you replace human traders with purely automated tools today, you won’t get the necessary flexibility. If the market keeps going wider, sometimes you have to stop and consider other options, perhaps by working an order agency rather than risk.
We will continue to automate, but only where we think it will benefit the client. Some of the markets in which we trade are still quite manual, such as the sterling credit market. You can automate pre-trade data inputs, but you’re not going to necessarily trade a larger sterling order of a long-dated sterling bond via an ECN. You’re still going to look for a single axe and make a judgement as to whether that’s the best outcome, given that asking more brokers could impact your market. For now, the benefits of automation lie in liquid asset classes such as rates and FX, and less in illiquid emerging markets, high yield, and investment grade markets, in something like US Treasuries the human trader many not be able to add too much value in normal market circumstances.
How are you responding to changes to liquidity provision across fixed income markets?
The liquidity conditions in the corporate bond market is one of the things that has kept me awake at night over the past seven or eight years, but today I wonder whether all that lost sleep was necessary. AXA IM’s corporate bond trading volumes in the UK have risen significantly over the last few years and our TCA is the same or slightly better. There are lots of different ways to access liquidity and I think we’re benefiting from greater automation and a wider range of connections and protocols across the market. Pre-2008, we would often be dealing with a handful of banks, approaching perhaps three to find the other side of a trade. But today, all-to-all protocols allow us to reach buy-side counterparties that we would not previously have been aware of. Importantly, these channels also keep sell-side balance sheet available when the need arises to transact in size, rather than being tied up with the recycling of small trades.
If you have time and flexibility, you can do business at the price you want. Sometimes, you know you’ve got a challenging week ahead, but we have the systems, skills, and experience to deliver and bring those spikes in turnover to a satisfactory conclusion. Equally important is the ability to demonstrate that we are getting the best outcome for the client on every trade. If the regulators come in tomorrow, we are able to show them through the processes we’ve put in place to deliver best price, independent of whether the trade goes to market via a machine or a trader.
Speaking of regulatory oversight, do you view MiFID II as having been positive on balance?
MiFID II was a lot of work for the industry, but it brought us more transparency, cleaner processes, and more protection for the end-investor. Has it made a big impact on liquidity? Not as much as perhaps expected. Trades are reported very quickly now, but risk is still taken down. The push for consolidated tape could bring a healthier outcome so that everybody has that information available equally, rather than delayed.
There is more governance over everyone working in financial services and that can only be good. I like the greater transparency, especially in regard to pricing. We are a large asset manager in Europe and we have set up trading to have a competitive edge, but I think we should all have access to the same data and information.
This article was originally published in the FILS in Philly Today magazine, produced by The TRADE, which was distributed to attendees of this year’s Fixed Income Leaders Summit USA conference.