Where do your priorities lie in terms of optimising AllianceBernstein’s fixed income trading operations?
Jim Switzer: Data aggregation is probably our number one priority right now. We’ve been amassing historical and point-in-time data through our ALFA market liquidity surveillance tool over the last five years. This has given us a fairly robust picture of the market, which is now informing our approach to direct connectivity, both to ATSs and the sell-side.
Internally, our focus is on systemisation of our investment process. In today’s challenging liquidity conditions, systemisation allows us to use different liquidity strategies. One example is the concept of portfolio trading. The aim is to look at baskets of risk in a similar way and be able to trade multiple portfolios at the same time. This requires not only electronic process change but also behavioural process change from traders and the investment team in how they approach risk management and portfolio optimisation. It also requires support from the top: senior management has made it very clear we should strive for innovation and efficiency.
In the three months to the end of May, we executed around US$10 billion of portfolio trades, 25% of our high-touch volumes year to date. Recognising the growing role of passive investment vehicles in the fixed income markets, the core idea is to transfer diversified risk baskets, as opposed to large blocks of idiosyncratic risk, which the sell-side do not want to buy because they can’t hedge it. Instead, we can sell a US$500m diversified basket, for example, if we can identify opportunities from total return funds or exchange-traded funds (ETFs). The process is automated through our Abbie ‘bot’, a virtual portfolio assistant, which can quickly optimise these portfolios, helping traders to price them and transfer the risk.
How is the rise of passive investment vehicles changing the fixed income market?
JS: The growing size and presence of ETFs has created an ‘arb-able’ event. Ever since the crisis, there has been speculation about the potential role of alternative liquidity providers. Compared to the pre-crisis environment, today we no longer have sell-side prop desks, dealer inventory is substantially reduced, and hedge funds use less leverage. But now we have an arb-able opportunity, with dealers using their balance sheet to support diversified risk transfers, which offer us an important way of sourcing beta.
Because they have insight into the ETFs’ create/redeem process and can hedge their risk, the sell-side can facilitate riskless transfers, acting as intermediary between active and passive managers. This market was still fairly sketchy only a couple of years ago, but we’re now able to transact in it very consistently today. In parallel with this change from a principal-based market-making model to a riskless transfer model, the bigger change is how the sell-side engage with us as clients. Rather than telling us what they’re trying to get done, they ask us what we’re trying to get done.
What support do you need from third-party providers to augment your own initiatives?
JS: One of the most important things we need is a utility to connect fixed income market participants and trading platforms. I don’t think any asset manager wants to maintain 10-20 APIs to connect to their main sell-side counterparts, nor do dealers want to maintain different APIs to their 100 best clients. There needs to be a utility to connect us all together, allowing direct connectivity to and from our execution management systems. We’re digesting lots of data sources, we have multiple tools and analytics running, and we want to be able to engage the market efficiently at the point of risk transfer. We can’t always do that right now given the existing infrastructure and often have to resort to voice trading.
Another thing we’re focused upon is bot-to-bot trading. Using our bot, Abbie, we’re in beta testing with a number of sell-side firms to enable bot-to-bot smart order routing, where the trade is sent to the sell-side firm and is automatically routed either to an algo or for risk pricing based in the information Abbie sends with the trade.
Automation in primary issuance has lagged the secondary markets. How should the challenge be addressed?
JS: Without a doubt, the single biggest waste of time for a trader on the fixed income desk is the new issuance process. It could and should have been handled two years ago. I believe there is a willingness on the sell-side to come up with a digitised solution. It’s not a matter of changing the allocation process, rather it’s about the securities set-up process and how we communicate orders back and forth. It’s a problem for the sell-side too and it’s impacting the secondary market. We’re agnostic to the solution and are willing to listen to the syndicate desks. But make no mistake about it: we need a solution.
How are you augmenting the skill set on your fixed income trading desk to ensure you trade effectively in an increasingly automated?
JS: We still need the people with 20-30 years’ trading experience, but we’re interspersing them with traders that have smart order routing skills, or the ability to write code. There are also similar developments in our analyst and portfolio management teams.
Even though the equity markets are ahead of fixed income from an automation perspective, they still have high-touch traders driving performance. We too need experienced traders that are skilled in risk taking and alpha generation, but a large amount of our day-to-day activity can be smart order routed and traded low-touch. We need a different skill set there. Using technology to engage the market quicker and more efficiently really does allow us to generate alpha; we can quantify the benefit to our clients via transaction cost analysis.
We recently hit four million messages a day, in terms of data coming into the fixed income desk, all of which helps traders make faster decisions. Most people think about best execution at the point of trade, but it as a more multidimensional process with many inputs. For us, the Abbie bot highlights opportunities based on our quant, fundamental, portfolio positioning and liquidity views, then builds approved trades for review by the trading desk where ALFA flags all the available liquidity options to the trader, with reference to various momentum indicators and liquidity scores. Based on all this, we consider the possible strategies to engage liquidity, remembering of course that the best way to generate alpha is to be a liquidity provider.
It’s important to have diverse experience on the trading desk. We have traders that have traded on all three of emerging markets, investment grade and municipals, while others have moved around between desks geographically. An ideal candidate is someone with equities experience as it is clear we in fixed income are now following the same path. When I first talked to our equity guys, I learned about dark pools, all-to-all trading and direct connectivity, but I also heard about illiquidity and fragmentation, which are not words I expected to hear from an equities trader. They are a lot further along with the electronic process change, mostly because of regulatory drivers. In fixed income, we’ve had relatively little direct regulatory intervention, but we all recognise that to compete going forward we have to be able to do things in a much more efficient manner.
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.