Blog article
See all stories »

Big Data, Magical Thinking and Investment Management

All information technology essentially offers an abstraction from an underlying reality, enabling that reality to be understood in simpler, more human centric and hopefully better ways. Some abstractions are good; the early GUI notion of “documents”, “windows” and the “desktop” allowed early PC users to interact with a strange new tool.  Some abstractions are simply bad and just confuse or get in the way (Microsoft Bob is a personal favourite)

Some abstractions such as Google’s search box are so good they are indistinguishable from magic. Users completely forget they are dealing with a tool and, in the case of Google and my Mum, attribute its abilities to some kind of mysterious intelligence which understands what she wants to look for on the Internet via some unknown mechanism (a lot of “magical thinking” relating to Google goes on at this point). Google achieved this abstraction by thinking about the needs of the end user and applying what has come to be known as “Big Data” techniques to analyse and present information.

Clearly transplanting Google’s approach directly into the Investment Management industry is inappropriate, and indeed would itself be a form of “magical thinking” (“big data will solve all my problems!”). However taking a step back, reviewing what data we have and what ways we could abstract that data for consumption by human beings will help us identify which new technology approaches may be effective.

In the Investment Management industry the key “reality” which needs to be abstracted and presented is:

  • A set of objective, factual data points about the state of an Investment at a given point in time (positions, holdings, valuations, NAV, VaR, greeks…..)
  • How this information has changed over time (performance analytics and history)
  • What constraints/rules are in place regarding this investment (mandates, applicable government regulation, best practice…)?
  • What decisions have been made in relation to the management of the investment?
  • What impact have these decisions had on the state of the investment?
  • What impact have outside influences had on the state of the investment?
  • Why where the decisions relating to the investment made at the point they were made?

The investment management industry generates stores and presents a huge amount of data relating to this reality. Specifically the first two questions are relatively well served by existing systems such as Fund Accounting platforms and Investment Performance systems. However as an industry we have not yet managed to address the later “What” and “Why” questions well at all.  Intelligent application of “Big Data” technologies, as well as leveraging existing IT infrastructure may give us the chance to create abstractions and tools which are as useful as Google’s search box (or at least not as bad as Microsoft Bob)


Comments: (0)