Vertex’s visualizations: Advancing algos & spotting spoofers

April 16, 2016 11:00 AM

Since the inception of the global markets, compliance departments, regulators and traders have struggled to capture a real-time image of the markets. Trades, volume levels, indicators and more create oceans of individual data points — like a slow-motion frame — of the world’s markets, which can offer insight into what happened in the past. But Chicago-based Vertex Analytics has developed an advanced data set that offers billions of market instances that can display real-time events by the second, whether it occurred minutes ago, days ago or years ago.

For example, at the 90th Annual Mid-Winter meeting of the Security Traders Association in January, the firm’s COO Doug Duquette displayed the reaction in the currency markets of the 2015 decision by the Swiss National Bank to lift its peg to the euro. The real-time snapshots of what ensued that day are visually stunning, showing what risk looks like in the markets (see “SNB unpegs volatility,” page 33). Order and expectations of the status quo show the same breakdown and array of colors of what is visible when a meteor burns up in the earth’s atmosphere. In just a few seconds, algorithms sputtered: Slide by slide, shot by shot, the slow-motion replay of a market in chaos.

To showcase the advancement of data specialization in the fintech space, Modern Trader sat down with Duquette and Vertex CEO James Austin to discuss the dramatic advancement of data analysis of the global markets in the last decade and why analytics can help shape a common language on matters of spoofing, insider dealing and other significant market events.

Modern Trader: What market need were you trying to fill when you started Vertex?

Doug Duquette: Vertex — in a nutshell — [offers] the ability to pull a drop of water from the ocean with a click and without all of the capex usually needed to even begin to collect the data internally. We monitor the data quality, and it’s perfect every day.

MT: What do you mean by “a drop of water from the ocean?”

DD: First, even having a forensic record of events (the data) is huge, and, second, being able to walk the data using unique fetch and visualization tools. It became critical to provide a snapshot in time of what is happening and how; especially through visualization vs. rows of data in a spreadsheet. This platform is the ultimate slow-motion camera on market activity, all at message-level granularity. The microstructure detail of any market event can be made available instantly with a simple click in a browser. We answer the question, “What just happened?” With this platform, any question can be answered by the compliance person, trader or regulator. It truly is a slow motion camera on market behavior and events. You can see what happened one second ago, five seconds ago, even five years ago.

MT: How does this address compliance challenges?

DD: Electronic markets have necessitated a paradigm shift in the need for more sophisticated compliance tool sets. For example, if we’re talking about spoofing or other commonly used terms, you have a situation in which the definitions of such behaviors are largely absent; and where definitions do exist, it’s often without context. Intent? Patterns and practices tell the tale. Our platform is primed for that task — with the technology, any set of behaviors can be observed within context to help us all understand exactly what is happening. That allows any compliance person to walk into any regulator or other forum with absolute proof of events that can be shared, explained and understood with clean visuals.

MT: What was the moment when you realized that your data stream had discovered something of significance?

James Austin: What’s amazing is what can be seen when you are able to slow things down. A ton of activity can happen in milliseconds that traditional systems don’t make visible. We think we know what just happened in the markets, but it’s amazing to look at it frame-by-frame. You say to yourself, “I can’t believe that just happened.” These reactions happen on a daily basis. It’s extremely powerful, there isn’t a day that goes by that you don’t find something interesting.

MT: Vertex captures every message every day from more than 400,000 futures data products. Was that possible 10 years ago?

DD: It wasn’t really possible or practical 10 years ago. Exchange protocols were not as information rich. Our technology stack has massive potential to scale. We have a proprietary backend platform that stores and organizes data in a way that one can — in a single click — find and pull a metaphorical drop of water from the middle of the deep ocean. Forget how many data points there are in a day; that’s hard enough. 

The harder and more impressive trick is the billions of billions of data points representing five years of market trading, and the ability to reach into them and see how the 10-year Treasury future traded on a certain day at 7:30 a.m., three years ago or 100 milliseconds ago. That’s the amazing piece. We recently helped a client get crystal clarity on some soybean and corn spread trading from the fall of 2011 to know what happened on a single day to successfully challenge a regulatory action. That’s an example of a real world application of our technology with tangible monetary and reputational benefits.

We’re not just focused on what happened yesterday, we have the capacity to focus on things happening right now. We call it “T-Zero,” and it offers the same depth of visualization instantaneously on the current trading session. All messages from the current session with a click. Right now, if a flash crash happened and we were in a meeting, we can reconstruct the event from seven minutes ago. This is a monumental piece of the puzzle for those who want to know [what] happened in the markets — whether it was years ago or just minutes ago.”

MT: Bloomberg recently wrote: “Vertex Analytics may have devised a way to make high-frequency trading more transparent and spoofing easier to detect.” What are they getting at? 

DD: You need technology like Vertex to say this is a spoof or not. We are just on the cusp of everyone having the necessary technology to be able to say that this is the line for market behavior. Everything has a context. It’s very difficult to say if is there a spoofing case that is recognizable. Assumptions are made in a vacuum [and are often] misinformed. There are claims being made, and you can look at data and say, “Yes, that shouldn’t be allowed.” But we can do better. The lack of specific understanding of the technologies that underlie electronic markets and the trading processes that occur within markets by people who do not trade for a living is expected, but concerning. Too much at stake to get it wrong. Platforms like Vertex allow people to get on the same page, using the same language. We can all go frame-by-frame, and we can walk through the process and reach consensus on what happened and how to regulate it.

JA: Sometimes, I’m not sure that people really want to know what is actually happening in the market. We still do not have [complete] answers about how to define spoofing or other crimes in the market.

MT: What benefits does Vertex offer the professional trader?

DD: Just look at the markets every day in this new context. Learn how they behave at the microstructure level. Additionally, the platform allows you to import your trades from the day (or any day) and navigate to the context of the auction at those moments. A full 360° view of everything. Unprecedented insight. Every person who has engaged with the platform has had an evolving experience. Something they saw today or two weeks ago or last year on that Treasury 
melt-up day. Traders walk away saying that they had no idea what was really happening out there. This makes crystal clear what is happening. With traditional systems things are largely left up to traders’ imaginations to fill in the blanks. Why does any given chart pattern occur? Now, that cerebral sorting of information is backed up by actual fact that reveals what happened, and then what happened, and what happened next. 

MT: Many folks blame high-frequency trading for market volatility. What do your tools reveal?

DD: When it comes to the markets, anxiety levels go through the roof. The markets are going to do what they are going to do. If markets can, they will. It doesn’t matter if it’s guys in a pit or algorithms driving the markets. The same event or result is going to happen. Electronic trading is making things different in sheer speed. It scares people. It’s gotten into sub-human speed that makes things invisible. Vertex gives people a counter balance. We now see exactly what traders are doing, and how they’re doing it. We can’t regulate our way out of volatility. What we’re talking about is risk and what does it look like? Well, now we can see what risk really looks like. At least we can now see it and understand it better. 

Contrary to some recent claims, the ability to do real-time surveillance is not only possible, the basic pieces needed are operational today. Our near-term plans include the streaming integration of execution messages from other platforms to give traders, compliance [personal] and risk [officers] a real-time microstructure view of events. Internal and private of course. The platform can also be coupled to the exchange back ends in the same manner giving exchange and government regulators a similar capability. 

MT: What is next for Vertex?

DD: Our next focus centers on increasing usability features, which includes work on more intelligent analytics and making the data easier to work with in more complex ways including customized visuals. We are meeting with different folks throughout the ecosystem. Everyone looking to find out how to proceed with cases and next gen analytics. Our goal is to help provide clarity and easy to use tool sets. Create reference cases with a forensic detail, knowing when algorithms go awry and when they are operating as intended. Develop deeper and clearer understanding of market dynamics. That is the big push: to take the power of this technology and apply it to everyday solutions. We want to continue to help increase transparency, clarity and productivity for every role in the electronic trading ecosystem and hopefully add some value as we go.

Page 1 of 2
About the Author

Garrett Baldwin is the Managing Editor of the Alpha Pages and the Features Editor of Modern Trader. An author and Baltimore native, he earned a BS in journalism from the Medill School at Northwestern University, an MA in Economic Policy (Security Studies) from The Johns Hopkins University, an MS in Agricultural Economics from Purdue University.