Intelligent is the New Smart

It has only been a few years since brokers and third-party providers on both sides of the Atlantic started marketing their latest execution offerings-smart order routing (SOR) tools-developed initially to address the challenges posed by fragmented liquidity. But that was three years ago, which, from a technological standpoint, is an eternity in this industry. Now, we are seeing the emergence of second-generation tools, designed to predict the probability of advantageous executions based on existing parameters and previous execution data.

As a result, the leaders in this segment are constantly forced to exert more effort to quantify data and make upgrades to infrastructure and technology, says Thomas Chippas, managing director and head of North American Autobahn Equity at Deutsche Bank. Instead of looking for quarterly analysis of SOR strategies, this is now done on a daily or weekly basis, with changes being made on a weekly or monthly cycle, he says.

"After the implementation of Regulation NMS in the US, smart order routing has been in a perpetual state of update," Chippas says. "To use a simple analogy, there is no longer a single event that says we are now moving from coal-fired stoves to natural gas furnaces."

This evolution has meant that smart order routers have become more adaptive to changing strategies and regulations. They have also become more predictive in the sense that they are now better able to understand where liquidity is being hidden. But in this rapidly changing environment, it is likely that the tools on the market today will be vastly different from the ones available a year from now.

Routing around

A year ago, traders were content to use smart order routers to find the national best bid and offer (NBBO) without having to route around exchanges themselves, Chippas says. However, that one-size-fits-all approach is no longer conducive to an industry where liquidity is becoming increasingly fragmented. Client firms now want more leeway in how they access displayed and hidden liquidity, while not leaving large footprints themselves, says Ruth Colagiuri, head of electronic products, Americas, at Bank of America Merrill Lynch (BAML). This has led to the need for predictive and adaptive technology.

"When we talk about 'predictive' and 'adaptive,' we are referring to the ability to modify your behavior in real time based on the results you see from the venues you access, the fields you get back, and the quality of the executions that you are receiving," Colagiuri says. "Those are definitely all things that have evolved over the last 12 months or so."

Harry Gozlan, CEO at liquidity management firm SmartTrade, says his clients are now asking for flexibility in trading techniques. Whereas SOR went from being static to more of a code-smart routing system, the technology today requires that firms be able to take incoming data to evaluate the quality of those various liquidity sources.

"For me, the main evolution is in terms of flexibility and integration that characterizes what smart order routing is," Gozlan says. "I don't know what generation you can call that, but that is where we are."

This evolutionary drive has also created a new challenge for the smart order routing industry, says Jamil Nazarali, managing director and head of the electronic trading group at Knight Capital. He says he estimates the current crop of smart order routers to be, roughly, generation 2.3. He says firms still need to be sophisticated and not rely solely on the intelligent behaviors of the technology.

For example, if a firm places a large block order on say, Microsoft or Starbucks, and lets it sit all day, an SOR will look at it and think there is a greater likelihood of getting a fill in one alternative trading system (ATS) as opposed to another, which chopped the parent block into smaller, child orders. That is how most dark pools work, says Nazarali.

Not all those child orders get filled, but that does necessarily mean that the other ATS has a better chance of getting a fill. "You do have to be adaptive, but you have to be adaptive in a smart way-if you are adaptive in a dumb way, it is a lot better not to be adaptive at all," Nazarali says. "I have seen clients who think they are clever with these adaptive technologies, but they are missing out on huge orders of liquidity because they don't understand that that's how that particular ATS works."

Looking for liquidity

It should be kept in mind, however, that this is a fairly new space. Even so, the technology is constantly being refined due to an increase in the depth and quality of liquidity. Over the past year there has been a proliferation of liquidity sources, as the number of ATSes and dark pools has increased.

The smart order routers available today need to adapt to this increased fragmentation, says Dhiren Rawal, managing director at third-party SOR provider Quod Financial. As liquidity becomes more dynamic, smart order routers need to create new execution strategies to take advantage of hidden orders without going through hoops to find that new liquidity.

And regulation will also play a role in how SOR technology evolves over the course of the next year. Restrictions on short sales and the US Securities and Exchange Commission (SEC) considering the possibility of treating indications of interest (IOIs) as quotes, will continue to change how CIOs upgrade their smart order routing tools.

"As liquidity fragments-and even more so as these regulations come into play-you will see greater fragmentation of liquidity," Rawal says. "With all of that going on, you need technology that can adapt to the changing nature of that liquidity and the changing source of that liquidity."

BAML's Colagiuri says her firm is aggressively investing in ensuring that its technology can continue to upgrade its market data feeds, enhance the speed at which the platform can operate, and continue to move as close as possible to various emerging venues. And, of course, enhancing the system's intelligence is a must.

"We have focused a lot of attention in the past, and we are continuing to focus on anti-gaming logic and building intelligence into the smart router to ensure that we are able to protect our clients' order flow, and provide high-quality execution," Colagiuri says.

Evolution

Deutsche Bank's Chippas says he believes that the industry will continue to evolve over the next year as clients look past the "smart" moniker to routers designed specifically to help them create different trading strategies. Essentially, he says, clients will express a preferred tactic and need, and the broker will decide how best to address it.

We are likely to witness something of a shake-out in the industry, Chippas says, as smaller SOR providers with limited budgets get squeezed out of what has become a very competitive market. Those that play in smart order routing will have to be able to spread their costs across a large base of users and transactions, he says.

Effectively, the SOR market will become a numbers game, where continued success will be contingent upon economies of scale where, for example, the larger the number of executed trades means more data can be collected, which in turn will allow the leading SOR providers to gain a competitive advantage because their tools will have a wider, deeper information base on which to make decisions.

"If you are at a smaller shop, you are likely to see fewer orders and you are interacting with the market less often," Chippas says. "So the size and sample set that your quantitative research team has to work with is going to be smaller."