Printer Friendly
The two executives, appointed just about a year ago, both from the Nomura side, were Fumiki Kondo and Anthony Abenante.
As the duo approaches the one-year mark at Co-CEOs at Instinet, Advanced Trading sat down with Kondo and Abenante to hear how their roles are playing out, how the two firms are leveraging each other's assets, what new offerings clients can expect soon and how Instinet's subsidiary Chi-X is faring in Europe and beyond.
Kondo previously served as Executive Vice President of Nomura BlackRock Asset Management, while Abenante was Managing Director responsible for U.S. equity and global program sales and trading at Nomura Securities International.
The discussion comes on the heels of Instinet's first combined product offering. The recently released Instinet Execution Experts algorithmic trading suite leverages the best of both Instinet's and Nomura's algorithms.
Anthony Abenante (left) and Fumiki Kondo (right) co-CEOs of Instinet.
Photos by Stephen Aviano
Advanced Trading: - How do you break out your roles as Co-CEOs? Who is responsible for what?
Fumiki Kondo: Basically both of us run the whole business together but I'm in charge of Asia and Europe operations and Anthony oversees the U.S. operations and global technology along with the CTO. I also look after the more corporate issues.
Anthony Abenante: It's actually worked out very well because we have complimentary skills. I've been here in the U.S. learning more about the global markets and extending the agency model. Kondo was EVP at Nomura BlackRock and he knows the operating side very well. This setup has been helpful as we try to rationalize things here at Instinet. The firm is in a much better spot now.
Kondo: We have completely different backgrounds and I'm learning about the execution and technology business every day from Anthony. Part of my role is making sure the Instinet business model is not deteriorated under Nomura. We are very careful not to mix the brands of Nomura and Instinet.
AT: Where are you working to combine Nomura and Instinet's expertise and technology offerings?
Abenante: One of the first synergies we recognized was in the U.S. equities space. Clearly (Nomura didn't have) the scope of what Instinet had from a liquidity standpoint. But Nomura had a rich set of analytics, microstructure research and a broad component-based algorithmic system, which we called the Execution Experts. It was based on newer technology and built from scratch at Nomura starting at the end of 2003.
So we decided to take the two U.S. businesses and merge them together with Instinet remaining as the brand for U.S. equity and program trading. That began last fall and we finished up pretty quickly over the course of a few months.
The next step is to generalize the algorithmic suite for global and for multi-asset trading as well.
Instinet has a very broad swath of clients from hedge funds to traditional asset managers. And as we looked at Nomura's relationships, the complementary nature of them was amazing. In the six months or so since combining the businesses, we've increased our penetration and wallet share with some very large institutions here at Instinet because of the some of the relationships we had a Nomura.
At the end of the day, it's all about best execution, but how do you prove that? We need to traffic in being more intelligent by layering in more research and mathematics into our trading systems and combine that with the great distribution footprint and liquidity pool we have here.
Kondo: We also wanted to emphasize the Instinet agency model " the "unconflicted trading model," if you will. Nomura does have a proprietary desk so we separated very clearly what Nomura offers and what Instinet offers. Normura appreciates and understands Instinet's agency only model.
AT: When will the new algorithms for multi-asset trading be available and what will they cover?
Abenante: It's hard to say when exactly because the globalization side is job number one now. We are going through the globalization process so we can generalize at the same time and put them together. We'll be taking the infrastructure and having a practical application available in Europe and Asia first for cash equities. We hope to have all that done by the end of the year. So I would expect it would be a 2009 initiative that would look at multi-asset algorithms. I think it's a great growth opportunity for us.
The initial area of focus for us will be listed equity options and listed futures. I think its' a natural extension of our business. I posit that the listed equity options business looks a lot like the cash equity business 8 years ago with the same phenomenon around market dynamics and moving to algorithmic and direct access.
AT: Chi-X Europe, the pan-European multi-lateral trading facility and Instinet subsidiary launched last year, has been gaining momentum and volume. What do you see happening going forward in the European MTF/ATS space?
Abenante: I think the backdrop in Europe with MiFID gave our firm a unique opportunity. There have been other attempts by ATS' to gain traction but we clearly have an operating expertise in the space. There are plenty of people moving into that space now and we never expected to be the only one in the space, but we benefited from being the first mover so our hope is there will be some stickiness to liquidity. With that said we're not resting on our laurels. We continue to work on the underlying technology, the matching engine, creating more efficiencies and continuing to reduce latency.
We've also launched Chi-X Global and Chi-X Canada is in a soft launch period and we're starting to see some decent market share growth. We think that the Canadian market has great growth opportunity given the market inefficiencies there that make a facility like Chi-X attractive. It will allow some of the high frequency traders there to run models there that they haven't been able to in the past.
We've also filed paperwork to launch Chi-X ATS in Australia and we're waiting for comment. We think that Australia will be the first market in Asia to set precedence from a regulatory standpoint and we're hoping some of the other Asian markets will take that lead.
AT:Do the ATS' leverage the same technology?
Abenante: Yes the majority of the technology base will be the same for all of the Chi-X platforms, with some additional code written to accommodate microstructure differences in each of those markets. We developed the platform from the ground up with the idea it would be a global matching engine and we think that's an operating advantage.
AT: How do you work to bring down latency?
Abenante: Our software engineers are constantly working to optimize the code and achieve performance gains. We've seen continued improvement but we'll also continue to re-optimize the model we have.
Beyond latency, capacity is a huge thing people are focusing on. It's one of Chi-X's key value positions and latency and capacity definitely go hand in hand. Ultimately the whole idea is to continue to increase throughput. As we want to add other features into the platform we're careful not to do anything to the core-matching engine. It has been designed to provide a very efficient processing environment.so when we do other things such as add-ons; we have to be cognizant of what the performance is going to be.
Instinet's co-CEOs on the trading floor
AT:Are there any new features you're working on for Chi-X Europe?
Abenante: Adding smart routing will be very important as we don't currently offer that functionality in Europe. We're in the process of merging that code into what we've done in Canada. That will be a huge new feature. Part of how successful you are depends on the having the makers and takers. How do you get people to make? One way is to ensure they won't have stock trade around them. So having a router that allows a client to post flow on a new fledgling exchange is a very powerful tool. It allows them to say "If I see a better price in a different market ,I can go out there and do it." That will be a very significant add-on to Chi-X Europe.
AT: Do you think the volume in dark pools will continue to grow?
Abenante: Ultimately I don't think the market is going to go 30 or 40 percent dark unless you have a substantial amount of market making being done by the bulge bracket guys by putting prop flow in the dark pool. I think what's going to happen is that people get smarter and smarter in understanding the cost of trade. We're focusing on TCA tools and we'll continue to evolve in that space.
AT: How exactly do you see TCA evolving?
Abenante: The important thing about TCA is maintaining a consistent approach. Now certainly one could argue that any of the market impact estimators that are incorporated into TCA has some flaw like, how do you model once over 100 percent of a mid cap or small cap name? " but maintaining a consistent approach to measurement is key.
Also, by brining in more risk modeling techniques into TCA tools will be important. That's something we're focusing on— giving our clients more real time feedback as to what their trades are doing. When we give pre-trade feedback obviously its based on historical data and we don't know what the market is going to do that day so it's static. What we're building now—we're calling it Instinet Insight— is a web-based full analysis suite of pre- and post- and intra-trade analytics.
When you look at portfolio trading typically its the tails - those two or three sigma trades - that kill you. Our thought is, let's give people information earlier in the day, not at 4:15. We'll give you the tool kit to see what's happening and our sales traders will have it as well to provide better color. Part of Instinet Insight will be a Web-based product where users can look at how trades are doing within each of their algos, giving them real feedback and perhaps the opportunity to change tactics. I think people will focus on intra trade as the next cycle of TCA evolution.
Instinet Insight is used internally right now and we hope towards the end of the summer to begin pushing it to select beta clients. We currently have some integration into our Newport EMS and soon hope to have a richer set of features.
