By Michael Mayhew, chairman and director of research, Integrity Research Associates
Last week, at A-Team Group’s Intelligent Trading Summit held in New York City, I moderated an insightful panel on the growing use of big data and data driven analytics in the buy-side research process. The panel of experts discussed a number of interesting topics including what they see as the future for ‘big data’ on Wall Street.
Big Data Panellists
Members of the panel included Harry Blount, CEO of DISCERN; Brian Lichtenberger, principal and co-founder of 7Park Data; and David Kedmey, president and co-founder of EidoSearch.
DISCERN is an institutional investment research firm which conducts statistical analysis of huge amounts of structured and unstructured data to identify unique investment signals. The firm then overlays this data analysis with contextual insights generated by experienced sell-side analysts to identify potential investable opportunities. DISCERN provides a wide range of unbundled services, including allowing buy-side clients to license its data, the predictive signals it has developed, or the extensive research it generates.
7Park Data is a boutique data and analysis firm that sources exclusive structured and unstructured data from third-party providers, cleanses and stores it utilising big data technologies, and develops predictive analytics based on this data to generate investment insights. 7Park markets these data driven research and information products on a subscription basis to institutional investors worldwide.
EidoSearch is a financial technology firm that applies pattern search technology to an extensive historical database of market data to generate predictive analytics and return projections for stocks, futures, currencies and other market indices. EidoSearch quantifies the way investors are likely to respond today by studying their behaviour when similar price patterns and market environments occurred in the past, providing a historical backdrop for every trade and investment decision.
Shouldn’t it be Called ‘Big Analytics’?
Warehousing huge volumes of data is really irrelevant to analysts or investors. Instead, the real key to driving value from big data initiatives is the development of predictive analytics at scale. When asked how their various solutions address this issue, Harry, David and Brian explained the following:
Harry explained that DISCERN sells signals as a service. What this means is they try to increase the signal to noise ratio, thereby enabling them to discover weak signals from the structured and unstructured data they collect. Consequently, the team at DISCERN is all about developing analytics that can identify the patterns or signals that could identify possible investment opportunities for clients.
David noted that EidoSearch’s clients are under tremendous pressure to sort through vast amounts of time series data in order to make better investment decisions. He explained that EidoSearch harnesses the power of pattern recognition technology, applies this to decades of market data, and generates a series of possible forecasts for the future of a security price based on how investors have reacted to similar securities in the past. This enables investors to make better informed trading and investment decisions.
Brian commented that his real business at 7Park Data is delivering insights to clients which result from the analytics the company develops. However, he warned that you can’t generate investment ideas from all data sources. That’s why 7Park spends considerable time and effort finding unique and exclusive datasets. Once that is accomplished they spend an equal amount of effort developing quantitative analytics which produce the predictive insights their clients are really looking for.
Big Data and the Human vs Computer Debate
Most traders and analysts have traditionally relied on human judgment to make investment or trading decisions. Big data is pushing investors to rely more on computer generated signals. When asked how they see users integrating these two approaches, our panellists replied in this manner:
Brian explained that the job at 7Park Data is not to replace human input to the research process, but rather to take on many of the mundane and time consuming tasks many buy-side analysts perform today. He contends that this will give investors more time to make thoughtful investment decisions. Ultimately, Brian noted that 7Park’s team works with buy-side clients to conduct the analysis they are interested in. Brian acknowledged that the only person they might be replacing in the investment research process is the sell-side analyst.
David commented that some clients could use EidoSearch to generate computer trading strategies. However, he saw most clients using a tool like EidoSearch to conduct an analysis that would require human decisions and insights to make actual trading or investment decisions. For example, David explained that clients might decide to eliminate certain time frames or securities from the historical analysis based on their own experience or insight.
Harry noted that the DISCERN platform is not just about generating investment ideas, but it is also about helping a client get answers to all the follow-on questions they are likely to ask. Harry felt the relationship between the DISCERN platform and the client was symbiotic where their system learned what matters to that client over time. In other words, Harry felt that the platform doesn’t replace human input or insight, but requires it to work most effectively.
The Future for Big Data on Wall Street
As is the case with most panel discussions, we ended the session by asking the panellists to share their visions of the future of the big data landscape for investors in three to five years.
Brian commented that in the next few years, a small number of the largest mutual funds and hedge funds will have built their own in-house big data programmes which would serve internal users such as analysts and portfolio managers. He expects these groups will act much like quantitative investors, consuming unique data sources and generating their own proprietary analytics and signals. However, he expects the rest of the market will look to external providers like 7Park or DISCERN to help them conduct analysis on big datasets to generate unique investment ideas.
David felt that the big data trend will likely have a significant impact on a number of major vendors to the buy-side, but the one player he thought would be impacted most would be market data providers. He explained that these vendors would need to switch from providing just data, to providing a platform where clients could generate real unique insights. As a result, David thought that market data vendors would either have to develop a big data strategy or else they might lose relevance in the marketplace.
Harry’s vision was probably the most expansive. He explained that in the future, corporations or institutional investors who did not have a big data strategy in place would be at a severe competitive disadvantage. He expects the firms with a working big data programme will be able to better identify new trends quicker and take advantage of these trends, whereas firms without such a programme will be forced to react to developments – a fact that could lead some of them to go out of business altogether.