Peter Waters, Managing Editor, GlobalTrading, attended a groundbreaking summit looking to address the evolving needs of the alpha-seeking buy-side community.
The concepts of big data and social media are generating a lot of interest in the world of financial services. Just how to make use of the vast quantities of data being generated every day, and how to harness the power of new technologies to create new trading ideas is at the forefront of every buy-side firms’ agenda.
With that in mind, Franklin Templeton Investments put together an event between the 25th and 27th February 2014 with a single focus in mind. The “Alpha Innovation Required” summit pulled together more than 20 of the most cutting edge vendors in new digital technologies, with some of the world’s largest asset managers and top tier brokers to discuss just how to move forwards. Pulling back to view technology holistically and productively is just as much of a challenge as the eventual application of the technology.
The event was collated by Franklin Templeton’s Director of Global Trading Strategy Bill Stephenson who spent the preceding months reaching out to colleagues up and down the Street to gather the most exciting names, and to receive countless demos. Spread over two principal days, the first was a time for the innovators to present their ideas and products. Each of the vendors, split into five separate categories, spoke for 10 minutes and took questions on why their idea is worthy of further attention. The second day was a chance to drill into details with specific vendors that each delegate chose to sit down with. The extra-long sessions provided much more detailed demos and analysis of the precise trends that led to the tools being developed.
With masses of data being created 24 hours a day 7 days a week, just which sources of data a firm chooses from is a slew of decisions in itself. Big data is generally categorized into two streams; structured and unstructured. As one speaker said, structured data is the easiest to use, but the problem is that everyone else has the data. Unstructured is much more challenging, but infinitely more rewarding.
Twitter is one of the largest, and most obvious sources of trading inspiration, but the sheer quantity of data, and judging its quality, is seemingly an insurmountable task. Some of the presenters have developed different ways to manage the volume of data, and to separate out the valuable from the useless. The Icahn Apple tweet is valuable information, a retweet several days later is less so, and a tweet from a fake account can be deliberately misleading.
Another firm takes news from over 22,000 outlets, although not including Twitter, and feeds out to firms breaking news and summaries that tie together disparate strings of information into actionable content. The key is in choosing the right inputs, and constantly striving to ignore the fake signals. Through their system they broke potential investment news hours before it reached the mainstream newswires.
Once the data has been gathered it needs to be cleansed and analyzed in order to have value. And these systems need to be constantly getting better at their jobs. A process called machine learning comes into play, by which the algorithms that pore through the data teach themselves what is most valuable. For example with Twitter, each tweet gets read and re-read through careful examination by algorithms that are constantly learning and improving so that the input can be ranked and made useful – what content other uses respond to most, or key names and accounts associated with market moving events. This process is one of the major driving forces of constant evolution in this space, as the tools themselves become better at their jobs by design, and feed clean outputs to other developers, who translate that information into a form that can be used by human eyes, and straight into algorithmic trading engines.
If there is one tool which can be said to be at the heart of the entire financial services industry it is probably Excel. One theme of the event was how these latest innovations in big data management are trying to move beyond this reliance into more accessible and visual forms of interacting with the data.
With the exponential growth in research reports, trading information, data management, and a whole raft of other information that takes up both time and screen real estate, traders and PMs are always on the lookout for tools that can display information quickly and concisely. By simply using 3D modelling and selecting a color palate that is designed to be appealing to the human eye, huge amounts of data can be much more instinctively controlled.
The simple amalgamation of such data is also a vital changing process. With huge volume of research reports and analysts models moving across the buy-side desk, how can the PM layer their own perceptions on top of the data, and keep easy track of reports that may have arrived into an already crowded inbox weeks or months ago? Automated data management systems that keep tracks of keywords and linkages between records are all now making this possible.
The upshot is that the PMS and traders can enjoy immediate access to a wealth of data in a form that is far more flexible and can be modified to each individual user.