Software AG Enables Market Surveillance
For institutional market participants operating in today’s high-speed electronic markets, the risks of trades going wrong are arguably more acute than ever, and regulators are watching closely.
That’s a favorable business environment for Software AG, whose products and services are designed to cover the ground on trading, risk, compliance, and surveillance, and allow trading and investing firms to focus on investment selection and trade execution.
“Firms in the capital markets need, more than ever, to be able to get a handle on the opportunities and the threats to their business as quickly as possible, often with sub-millisecond latency,” said Dr. John Bates, CTO for Intelligent Business Operations and Big Data at Software AG. “It’s all about opportunities and threats, and moving on them before your competitors and before it’s too late.”
Founded in 1969, Germany-based Software AG is among the 25 biggest global software companies. Its U.S. business is based in Reston, Virginia, and aside from capital markets, it has customers in industries such as banking, energy and utilities, government, insurance, manufacturing, and retail.
In capital markets, “we’ve put together a platform that brings together the right technologies, the fastest in-memory technologies, the fastest analytics, combined with the fastest decision management to enable those applications in trading, risk, compliance, and market aggregation,” Bates told Markets Media.
Bates highlighted Software AG’s 2013 purchase of complex-event-processing platform Apama as a milestone. “That has really strengthened the capabilities of Software AG in capital markets, as Apama has a strong background in areas like algorithmic trading, high-frequency trading, foreign-exchange market aggregation and real-time pricing, and market surveillance and risk,” he said.
Software AG’s Intelligent Business Operations platform combines a number of capabilities and business lines, and capital markets is a key target area, Bates noted. He also highlighted the firm’s Universal Messaging and Terracotta products.
Universal Messaging “gives you high-performance scalable messaging, plus it is extremely secure and has the ability to go very flexibly through firewalls, so it’s great for collecting market data,” Bates explained. “It’s also great for publishing prices to customers and getting their responses back into a real-time pricing system.”
Terracotta “is all about in-memory computing — collecting data, storing it in memory, (and) being able to analyze it,” Bates said. “Bring those capabilities together and you’ve got in-memory, messaging, data management, and streaming analytics and complex-event processing.”
“Combine that with another piece we acquired from an acquisition called Jackbe, which is our Presto Visual Analytics Product, and you’ve got the ability to visually analyze alerts and decisions, so you can graph charts and heat maps, and then drill into it,” he continued. “Put all this together and you’ve got a pretty compelling architecture for capital markets.”
In the wake of the ‘flash crash’ of May 2010, and subsequent disruptions including Knight Capital’s haywire trading and near-death experience in August 2012, surveillance has been a front-burner order of business for regulators, and by extension it has gained importance for market participants.
Last year, the U.S. Securities and Exchange Commission launched its Market Information Data Analytics System, which for the first time provided the SEC with data about every displayed order posted on national exchanges. MIDAS collects about one billion records per day, time-stamped to the microsecond, according to the regulator; the information comes from the consolidated tapes and proprietary feeds of each exchange and includes posted orders and quotes, modifications and cancellations, and trade executions both on- and off-exchange.
Mary Jo White, who was sworn in as SEC Chair in April 2013, has indicated she means business, both in terms of sniffing out aberrant market activity and taking action on market abuse. “This broad focus demands that we use all available means to detect and pursue violations,” White said at an institutional-investor conference in September. “So we will be taking advantage of tips from whistleblowers, using quantitative data available to us, and conducting sweeps and other means of uncovering misconduct.”
On the most basic level, buy-side traders and sell-side trade handlers need to monitor their own market activity to ensure all is well, because if there is an issue, regulators are likely to find this out. “We’ve been in the market-surveillance space for a number of years, and we’ve seen that space evolve,” Bates said.
“We started off looking for patterns that would indicate market abuse, insider trading, market manipulation, and even algorithms that have gone out of control,” Bates said. “Rather than finding out about these things after the fact, if you look for them continuously and in real time, you can look for predefined patterns that show the relationships of events that indicate something is going on.”
One market where demand for surveillance is booming is foreign exchange, which has seen multiple instances of alleged manipulation and abuse over the past couple years. Foreign exchange market participants and market operators note that FX historically has been lightly regulated compared with stock and bond markets, but rule makers are starting to make their presence felt. Increasing regulation, plus the robust health of the market, begets strong demand for FX surveillance products.
“The Libor situation, which was around market manipulation and fixing, won us more new customers and brought in new types of tracking,” Bates said. “These include rogue-trader monitoring, human-resources data — like when the trader takes a holiday — and self-learning monitoring of behavior…The idea is that you can benchmark normality continuously so you can figure out what’s normal over a six-month period and then you can spot deviations from the norm.”
Bates drew the example of an FX trader who makes an unusually large trade in an instrument they don’t typically trade, and the transaction correlates with a market movement based on a news event. “There might be something there,” he said. “You might want to visualize the pattern’s relationship to the event, so compliance can drill down into it. Maybe it’s nothing, just a lucky individual, but on the other hand, maybe there’s a pattern.”
“Another problem now in foreign exchange is that you’ve got benchmark rates that are allegedly being manipulated by firing in orders during the time window when the benchmarks are being calculated,” Bates added. “We’ve built, and we are in production with customers on scenarios that let you detect that kind of foreign-exchange manipulation.”
Software AG has expanded via acquisitions, and Bates emphasized the company is also adding business on its own. “We have aggressive organic-growth targets for this year, and we will also look out for the right acquisitions,” he said. “These could range from large strategic acquisitions to smaller technology tuck-ins that can help us differentiate.”
Whatever Software AG’s strategic plans, the near-term future of its capital-markets business will be largely driven by underlying market trends.
“Foreign exchange surveillance is big. Surveillance in general is a hot topic,” Bates said. “Use of Software AG technology, and the Apama platform particularly, has now taken the box from regulators by saying this is a known solution to get a handle on the problem. So it’s very attractive to firms at the moment, we think that’s going to continue, and we’ll continue to enhance that with new types of monitoring.”
“Real-time risk is also a continuing trend, particularly around getting a handle on detecting an algorithm that is operating outside of its normal parameters,” Bates continued, citing the example of Knight, which “arguably was because they still had a market simulator baked into the algorithm, so it was firing off out-of-market orders.”
“That’s going on all the time, and it can be detected,” Bates added. “Just like a trader, if you know what kind of instruments an algorithm should be trading, what are its typical orders per second or orders-to-trade ratio, that can be monitored, and if it goes outside that, it can be locked out from the market. It’s kind of a new type of real-time risk firewall, which is very appealing for the market.”