By Carl James, Global Head of Fixed Income Trading, Pictet
Quant-driven trading has been around for years, so it’s nothing new, but the pace of change is accelerating as more and more trading desks are adopting this type of trading.
Why is this the case? Regulation is forcing traders to be much more focused on delivering evidence of best execution. The phrase that best encapsulates this is, “show your work”. The standard for evidence is higher and it now has to be predicated on data. At the same time, developments in technology mean data processing is cheaper and data capture is easier. Finally, the people element is changing. The next generation of traders entering our industry is data savvy and knows how to use the data to inform their trading decisions.
The challenge facing many buy-side trading desks is how to effectively integrate these new tools and methodologies while retaining the accrued expertise and time-built relationships already present on the desk.
MANAGING THE TRANSITION
While there are clear benefits to quantitative methods, no one would reasonably suggest a head of trading make personnel decisions for an entire desk based exclusively on technological proficiency. The need for quantitative thinking depends on the asset class, geography, business mix and other factors. As a head of trading, you need people on the desk that have relationships and market knowledge, but you also need the right blend of people who understand datadriven tools or can write code themselves, in order for the entire (continued from previous page) desk to trade more effectively.
At Pictet Asset Management, we are making this transition across the front office. We are offering Python programming courses and at a recent offsite, almost everyone in attendance had done this course, including senior people. Rather than just segmenting the dealers to be the hub of data, it has become a cultural element of the whole front office. We work closely on this transition with the portfolio managers and analysts, who are also increasingly data savvy.
It is important to have people on the desk that are able to communicate this new type of additional information to the rest of the team. To do that, we must first ensure they have the tools to pull out that data as well as the data, itself, to analyse and make conclusions from. When you have people who are and are not data savvy, people from each group will come at a challenge from different angles, and the sum of the parts is more than the whole. The person who knows markets can bring their historical experience and the person with the data brings quantitative analysis, and in combination, it is quite powerful.
Cultural change is always difficult, but you can make it less difficult based on how you approach it. As an industry and as a firm, we had to make changes based on MIFID II, but we challenged ourselves to ask how we could turn that into an advantage. Suddenly, this adoption of quantitative tools required for best execution reporting became a soft message. The firm as a whole took a much broader approach to encourage quantitative and digital thinking. We had townhalls looking at digitisation and printing less, not just for environmental reasons, but to encourage the usage of digital tools. Anecdotally, we can see the impact. When we recently moved offices, we were given physical storage space that’s still empty because the desk now stores materials digitally.
We are also looking for new ways to replace traditional spreadsheets with a more robust solution and remove spreadsheets entirely from our organization. We now use a commercial platform to look at data, which enabled us to completely revamp our broker review process as well as how we do internal presentations. I no longer send presentations, I just present with live data because I can manipulate it and veer off as the audience needs with live examples.
A key challenge with any cultural change is how you approach reluctant individuals. Some may cross their arms and think, why should I change, but we respond that there is no need to fight it. The skills and experience of people that have been in markets for 10-20 years are valid and important. The key is showing our experienced people that they can adapt and evolve. When they started trading, it was different from what they are doing now, so they have adapted already. The evolutionary curve has become steeper, but we can work with them through the transition phase.
If it is positioned as additional value, for some people it takes longer, but generally, they have come along. Part of this is ensuring there is a clear plan for how behaviour changes right away. In the push forward to strive for excellence, some ideas you try will not work. But as long as it is accepted to test something that does not ultimately work, people will be encouraged to learn, improve and build a base of knowledge, as well as the courage to keep trying new ideas.
CASE STUDY: BEST EXECUTION
As mentioned earlier, a fantastic example of the benefits of this transition is our best execution reporting. Internally, my mantra has been, if you want them to, the dealers can provide alpha. Some firms view their dealing desk as a processing unit, but as for me, I believe our dealers can add value because the markets we trade, the strategies we use and the technology we deploy are all complex. Culturally, we recognise that dealers add value and we’ve done largely on the back of our quantitative approach.
It took a while to get the data right and analyse it properly, but we can now demonstrate that a standard dealer produces a certain average trading outcome, but we can demonstrate how our people and technology consistently outperform the market average. This has now given us a way to turn around and show the business we have added basis points to a client’s portfolio. We are still making market calls, but our job is to make sure we are executing as efficiently as possible and we can now provide that evidence.
On the qualitative side, our use of data has improved our broker management process. We know now the hit ratio and flows that come through each broker, their hotspots and the areas they are less good at. This allows us to assess claims from our brokers based on historical data. We are not going to bash brokers, but we need to help our brokers to see where they can improve to meet our needs. This has helped increase the thoroughness of our broker reviews with our top brokers, and the feedback has been quite positive as it drives better engagement with our brokers.
As a result of all this, our dealers now go to every single morning strategy meeting. People want to hear what the dealers are saying because they can communicate market action and where the flows are. We have pushed that by looking forward at the information people do not have right now so we can help our portfolio managers work their way through the markets. As is natural in any large organisation, hiring is one part of this transition. When there is an opening on the desk, it can become all too easy to look for people based on years of experience, but this approach risks viewing people as cogs to be replaced. To stretch the analogy, the machine can never evolve. We are not quite at the stage where data is a given for every candidate, but it is asked about every time. A candidate would have to be very compelling if they did not have those data skills, and that corridor for acceptance is getting narrower and narrower.
THE ROAD AHEAD
This transition toward a more quantitative approach will accelerate across the trading industry in the next few years. On the sell-side, some people that have been hired to drive fixed income tend to have an equity background, and my suspicion is that some of the equity electronic trading protocols will come across in the years to come. I also expect a blending between portfolio managers and dealers, where portfolio managers share the exposure they want or do not want, and the dealers share how they can best get that risk on, whether that be synthetically or through a mix of instruments.
An apt analogy for where we are headed with quantitative trading in fixed income is car-making – people used to make cars, but now computers do. Now, the people tell the computers what to do and help the drivers have