ベンダー
English
search
  2. //
  3. Reports & insights
  4. //
  5. The Computer Trading Debacle and Data Analytics Lessons for Banks

The Computer Trading Debacle and Data Analytics Lessons for Banks

Create a vendor selection project
Click to express your interest in this report
Indication of coverage against your requirements
A subscription is required to activate this feature. Contact us for more info.
Celent have reviewed this profile and believe it to be accurate.
We are waiting for the vendor to publish their solution profile. Contact us or request the RFX.
Projects allow you to export Registered Vendor details and survey responses for analysis outside of Marsh CND. Please refer to the Marsh CND User Guide for detailed instructions.
Download Registered Vendor Survey responses as PDF
Contact vendor directly with specific questions (ie. pricing, capacity, etc)
2013/04/26
Bob Meara
Last Tuesday, 23 April, a hacked Twitter message from the Associated Press about an attack on the White House that injured President Barack Obama, turned out to be bogus. But the tweet sparked a brief 145 point market selloff that dramatized the power of algorithmic trading - sophisticated computer software that analyzes language using algorithms to allow high-speed trading in financial markets. More on the event is available here. I’m not qualified to comment on algorithmic trading (more than I already have…), but do see a teachable moment in this event for banks exploring the use of data analytics for less controversial applications such as marketing and risk management. Two points come to mind.
  • Algorithms are stupid. For all their elegance, algorithms have no wisdom. This is not to say that data analytics is not highly useful (it is!), but the underlying algorithms are powerless to perceive their impact. That step requires perceptive humans willing to undergo some rigor and discipline.
  • Models are fiction. Data analytics is based on the construction of simplified models of things that have relevant business interest. Their simplicity is important, because it allows modelers to isolate variables that are relevant and available. But, it is all fiction – an approximation of reality. For this reason, models must be validated and results treated with care. Next best product prompts will be wrong some of the time. Virtual agent prompts will be nonsense some of the time.
These two aspects create both challenge and opportunity for banks seeking to leverage data analytics for competitive advantage. Culturally, doing so requires a deep commitment to a “test and learn” way of doing things. One is never finished in this new world. There is always opportunity for improvement. Models get stale and must be continually revisited. Celent observes a useful data analytics process in place in a number of banks (below). Data Analytics Process It’s hard work. That may be one reason why so few banks are broadly deploying these new technologies. But, it’s rigor that can’t be avoided if you want the good without the bad.

コメント