Apama is a streaming analytics service operated by Software AG and a Trading Technologies partner since 2007. Founded in the late 1990s at Cambridge University, Apama was built on complex event processing (CEP) technology, which enables the trading community to execute low-latency trading strategies from the analysis of multiple, disparate streaming data sources.
Today, Apama has built a global client base of over 100 firms, including banks, brokers, prop groups and exchanges across all asset classes. Apama’s algorithmic trading, market surveillance and automated strategies, which are triggered by coded news announcements, leverage Trading Technologies’ connectivity and trading functionality.
Graeme: Why is TT and Apama’s partnership timely for futures traders?
Tony: In recent years, market surveillance has become a key focus area from the regulator, with most trading participants being required to demonstrate the ability to monitor for suspicious activity. The requirement has spread to include several groups that were previously exempt. Every trading firm, facilitator and individual local will have to be able to prove they can run trade surveillance over their trading activity, and that will, in most cases, involve an investment in a surveillance solution.
There are many vendors in this space, many of whom offer rigid, tick box solutions that charge on a crippling per-exchange or flow basis. Apama has market surveillance clients across all asset classes and client types. We build bespoke solutions that fit the exact client requirements. Last year, a major global prop group of more than 500 traders asked us to connect to the Trading Technologies infrastructure to offer real-time, cross-exchange trade surveillance and reporting. By connecting directly to Trading Technologies, we bypassed the per-exchange commercial model offered by our competitors and delivered an open-source, highly configurable solution to fit the client’s trading business.
Graeme: Which of TT’s client types benefit most from Apama?
Tony: Smaller TT customers can take advantage of a hosted solution, whilst larger, more sophisticated clients with an appetite to develop in house can take delivery of an open solution that can evolve and be configured independently of the vendor. Regulation will only increase, and our approach allows our clients to react to this changing landscape with the ability to build a new alert from scratch in only a couple of days.
Graeme: What are the market surveillance and monitoring challenges your clients face in today’s market environment?
Tony: Trade, order and reference data are the biggest challenges. There are too many vendors in this space who offer wonderfully rich GUIs which are quickly rendered useless because they can’t make sense of the multiple data sources. The first question that should be asked by someone evaluating a surveillance solution should be, “How do you handle the data?”
Outside of the technological challenges, there needs to be a drive in the prop space toward education with the purpose of protecting the trading participants, as the regulator will not accept ignorance as an excuse. With prop trading groups, many of these traders have never worked for banks or brokers or been exposed to formal training. For example, click trading involves the constant entry, amendment and cancellation of orders in the bid/offer range or deeper in the market depth, waiting for a sudden price move and the possible profitable correction.
Coupled with the sheer number of autospreading algos in the futures space, much of this behavior exhibits the same patterns as a layering or a spoofing alert, which would result in lots of false positives.
There is still a culture in the market where certain behavior, including layering, is still deemed a legitimate trading technique, with the argument that by showing the size, traders are opening themselves up to the risk of being traded with.
Graeme: With so much data to analyze, how does Apama find the middle ground between flagging false positives and missing valid warnings?
Tony: This is a huge challenge. Too many alerts can be as useless as none. Apama has rich back-testing functionality that has been utilized for years by our algo trading client base. With the ability to replay tens of thousands of trade-related events per second and the easy reconfiguration of Apama’s alert parameters in a test environment, Apama can quickly eliminate false positives. We also have the ability to analyze normal trade behavior patterns to alert for possible suspicious or dangerous trading behavior, which has strong benefits for the risk team.
Graeme: How does Apama keep customers, like our users, up to code with the latest regulations as the rules change constantly?
Tony: Apama has an in-house team that constantly keeps abreast of regulatory change. We also participate in interactive roundtable sessions with market participants and have membership to bodies such as the Futures Industry Association (FIA).
Graeme: What upcoming regulatory developments do you predict we’ll see in the near future?
Tony: This subject is vast. You only have to look at the armies of expensive third-party consultants advising regulatory experts already employed by trading firms. Every month, there seems to be another ominous set of initials that join ESMA, EMIR, REMIT, etc. Coupled with no real consensus amongst the different regulators, this makes the future seem bewildering. In the UK, one current bogeyman, MAR, will strengthen the existing market abuse framework by extending its scope to new markets, new platforms and new behaviors with provisions for insider dealing and market manipulation. It will apply from July 2016, which is now under a year away. The market has to be agile in this space; it is constantly changing and a trading firm has to be able to react rapidly.