Q&A with Joe Wald, Managing Director & Co-Head of Electronic Trading at BMO Capital Markets
The US equities market microstructure landscape is evolving. The introduction of Reg-NMS and the National Market System in 2005, designed to promote efficient and fair price formation across securities markets, has had many positive effects in terms of greater competition, tighter spreads and lower fees. But the unintended consequences have led to more fragmentation, complexity around order types, and data asymmetry, particularly between high-speed traders and institutional investors.
To address these issues, early in 2020, the SEC proposed a number of changes to Reg NMS, in an attempt to modernise and strengthen the structure of the US equities markets, which were adopted this past December. Last year also saw the launch of three new exchanges, while other exchanges introduced new order types designed specifically to provide greater transparency and improved execution quality for institutional investors.
In this Q&A, Joe Wald, Managing Director & Co-Head of Electronic Trading at BMO Capital Markets, shares his insights on the how the US equity market landscape is changing, and looks at how exchanges are innovating to create a more level playing field.
How has the US equities market landscape evolve over the last few years?
The big shift really is a move that was started primarily by ATSs, but has now been embraced by exchanges, to provide a more diverse set of order types for a more diverse set of market participants, based on the different objectives that they have. To give this some context: as the exchanges have become electronified, the focus of their technology has been around this ‘arms race’, building the fastest matching engines that were able to accept high volumes of trading at a very rapid clip, catering predominantly to the burgeoning class of high-speed traders. And part of how they incentivized order flow from those types of market participants was to offer them higher and higher rebates for posting liquidity to the marketplace.
That’s been the status quo of the exchange model since they’ve become electronified. But that has left out a really important part of the market, the institutional investors that represent retail long term investors. The needs of those long-term investors are much different in terms of executing large blocks of stock with minimal information leakage, and making sure that there’s a high degree of execution quality.
As a result, the market structure has evolved into a lit exchange model for high frequency traders and people who needed to take liquidity, and an ATS and dark pool model focusing on execution quality and information leakage for institutional investors.
What’s changing now?
Exchanges are innovating to be able to compete for order flow that they see as very valuable. They’ve always wanted to have that, but for the most part they haven’t made it a priority to attract that order flow, whereas now there’s some real innovation happening.
For example, IEX, an exchange focused on the institutional investor, has developed a lit order type called the D-Limit, which is unique in that its objective is to protect a market participant who has placed a bid or an offer on a lit exchange from high-speed traders potentially recognising when there’s a crumbling quote happening and then moving the price down. That’s a pretty powerful and unique order type.
Then you have the Nasdaq M-ELO order type (midpoint extended life order), which sits on a separate book for a minimum amount of time. That creates a disincentive for people to come in and cancel, replace, cancel, replace, at microsecond level. So it roots out order flow that potentially isn’t really there or isn’t natural liquidity but more high frequency trading interest.
BATS now have approval for their midpoint intraday auction, which is something that already exists in Europe, and they’ve seen some good success with it. They’re now bringing that to the US, where you’ll have intraday auctions happening throughout the day, at any period of time, which is basically a call to action for aggregation of large orders at a particular price.
All of these are examples of innovation to attract order flow from institutions with size, to execute in a way that is going to be efficient and effective. So we’re starting to see this really incredible trend towards innovation and new order types, catering to institutional market participants who, in many respects, have had to find different ways to execute.
IEX’s D-Limit order type uses machine learning to predict how the market is going to move in the next fraction of a second, so that it can move orders out of the way if necessary. Do you expect more exchanges will start to adopt AI and machine learning for specific purposes like this?
There are other ATSs like IntelligentCross that are leveraging machine learning to help the matching process and to have higher quality executions. From what we’ve seen based on our venue analysis, order types like D-Limit and venues like IntelligentCross have been delivering good execution quality and have been seeing a growing share of our order flow because of that. So there’s definitely real innovation and real technology being implemented.
Machine learning techniques are starting to be used more and more. We use machine learning techniques ourselves around some of the research for our strategies and our algorithms. So it’s becoming more used throughout the industry and there are definitely going to be benefits from it, just as there will be experiments with it that probably don’t yield what people are expecting. So it’s just going to be a matter of doing the work and the research to demonstrate where they actually add value.
In the US, we’ve seen three new exchanges launched recently to cater specifically to institutional investors, LTSE, MEMX and MIAX, offering fairly different operating, pricing and ownership models to some of the incumbent electronic exchanges. What are your thoughts on the need for new exchanges like these?
In some respects, you’ve got some innovation there and it will be interesting to see how the volume begins to aggregate at those three new exchanges. LTSE’s focus is primarily on issuers, and their rules can potentially benefit the way an issuer has visibility over their security being traded, as well as a host of things designed to give investors and issuers more transparency around the holders of their securities. To date, we’ve not seen a lot of traction in terms of equity trading on that exchange, but it’s too early to tell how that’s going to play out.
MEMX was formed by a consortium of broker dealers, to address some of the issues that the industry felt has fallen on deaf ears, regarding the way some exchanges charge for market data, and the cost of participating and becoming a member of those exchanges. Their volume has been growing.
MIAX, again, we’ll see what happens in terms of volumes, but they have new technology owned in part by the ownership of the options exchanges, so they’re potentially catering to the subset of the trading community that also trades options, for example options market makers that want to have a venue to lay off the equity legs of their option orders.
So each has a unique purpose. That’s a trend we’re starting to see more, bespoke exchanges looking to solve specific market structure issues that they feel have been underserved by some of the broader exchanges.
You mentioned the cost of market data, which is an ongoing challenge for the industry. How do you see that playing out, given that a large portion of exchanges’ revenues come from the sale of market data?
In the US, we’ve seen the SEC pass the new Reg-NMS rule in December. That’s been many years in the making, following a lot of advocacy from the industry to make market data more competitive. The rule still has to be implemented, and right now is being challenged in court by some of the exchanges. So we’ll have to wait and see how it plays out before we move forward.
But looking at what’s in that rule, a few things are critically important. First and foremost, it establishes a new definition of what is core market data. The data that was on the SIP before was insufficient for people to be able to use it effectively. It didn’t have odd lots, it didn’t have depth of book, it didn’t have auction data. So regardless of how fast or slow it was, the fact that it didn’t have the necessary data in there to be a viable alternative was really something that, on the face of it, appeared anti-competitive.
Another important aspect is who can deliver the SIP. The SIP today is controlled as a facility of the exchanges. Under the new rule, you’ll have a competing consolidator model where new entrants can have the opportunity to compete as a distributor of data. So that could put pressure on costs as well.
The other part of the rule is governance. At the moment the SIP Governance Committee is of course just the exchanges that are on there, but you’ll have governance from a broader segment of market participants, to be able to at least have a seat at the table when discussing market data.
So there are some very important, potentially dynamic changing aspects from this rule change, and now we just have to go through the process of getting it implemented. But we’re definitely very encouraged. We’ve been extremely vocal about the cost of market data, especially being a startup six or seven years ago when we started Clearpool, and recognising that market data was one of our most expensive operating expenses. And the way the structure was set up was that it made it very difficult for new innovation to enter the market, because you had this huge barrier of entry for people to be able to bring new products to bear.
Fortunately, we’ve been successful, but even though we’re now owned by one of the largest banks in the world, and have plenty of resources to be able to continue to develop, we haven’t forgotten where we came from. And we’re still looking out for new innovators to be able to enter the marketplace to bring meaningful differentiation to it. So this is really important. And we’ll see how it plays out.
What do you see as some of the biggest challenges that the investment management community faces today in terms of US equities trading, and how can those challenges best be addressed?
The problems have always been similar. Institutional investors are looking for liquidity at scale and at size, which is challenging when you have a market that predominantly trades in small 100 share lots. And probably even smaller than that, with the odd lot phenomenon that’s happening. So the challenges of trading in size are well known, and they’re challenges that we focus on trying to solve every day, whether that’s with liquidity sourcing algorithms, with different benchmark strategies that leverage all of the different venues that are out there, whether it’s the dozen-plus exchanges or the dozens of ATSs, and the hundreds of order types, our job is to figure out which of those venues and order types are the ones where we have an edge, where we minimise information leakage, where we can extract the best execution quality with the least amount of reversion and are able to execute and deliver solutions to our clients who are looking to trade at size, without disrupting the supply and demand or without signalling their intent.
The algorithmic landscape has to change and it has to change rapidly. An algorithmic platform and an algo management system has to be nimble, it has to be something that’s able to adapt quickly. One of the reasons that we’ve been so successful and one of the reasons why BMO made the acquisition of Clearpool is because we built our technology stack with that in mind, from the ground up, using a modern technology foundation. That’s critical, especially doing so with full transparency, because when you’ve got all of these different choices to make, you want to make sure that you are being transparent, and the client is working with you in a collaborative way to understand what you’re doing and what the potential outcomes could be.
So I think the key is being nimble, being able to anticipate that changes are going to happen, because they’re going to happen rapidly, being able able to iterate quickly. You’ve got to be able to run A/B tests and vet out new order types over time, you’ve got to be able to reconfigure strategies based on where liquidity is and where you’re seeing execution quality.