Measurement Deficit Disorder: Which Data to Measure and How to Measure It

A plot point in the latest season of the HBO comedy series Silicon Valley has the founders and backers of startup tech company Pied Piper glorying in the high numbers of downloads of their app after its launch. They are slow to realize that the more relevant measurement of their success is the number of daily active users.

This story arc from Silicon Valley illustrates a point that Ron Lee, chief information and operations officer at MUFG Canada Branch, made at the Toronto Financial Information and Technology Summit last month.

Lee cited a statistic from recent research conducted by McKinsey and the Harvard Business School that 60 to 90 percent of strategies fail to execute. “Executives want to understand why projects that were so cut and dried at the beginning get so delayed, get canceled, or get beaten by their competitors,” he said. “If you use traditional methods such as scorecards and APIs, they usually point to the same handful of culprits: budget, people, systems and culture. But the underlying cause is that the strategy and planning never considered these things in the first place.”

And, as Lee’s colleague, Tarun Dhot, director of advanced analytics at CIBC, also pointed out in the conference, “actions without metrics are even more useless.” The Silicon Valley characters were not using the right metrics to measure their performance.

The importance of using the right metric is also becoming apparent with the assignment of legal entity identifiers (LEIs), within the context of supporting compliance with Europe’s Markets in Financial Information Regulation (Mifir). Chris Johnson, head of product management, market data services at HSBC—who is a leader in LEI adoption efforts—stated in Inside Reference Data’s recent Mifir webcast that the 450,000 LEIs now registered are not enough to ensure Mifir compliance.

The inability to produce the right data is similar to consulting the wrong metric or measurement.

This might indeed be the right metric to watch, although there may be others that should have also been monitored and considered at the same time. In the same webcast, David Nowell of the London Stock Exchange Group pointed to the ISIN securities identifier. It appears that working with both LEIs and ISINs will be essential for data management and getting the greatest accuracy and quality possible for Mifir compliance reporting when that data is being gathered from multiple sources.

Choosing More Wisely

A similar issue concerning tracking and managing data necessary for compliance surfaced recently, this time with Solvency II, the European directive for capital adequacy in the insurance industry that took effect in January this year, with data reporting begun in April. As is likely the case with many other major firms, Henderson Global Investors, with £92 billion ($123 billion) under management, disclosed that it was still working on adjustments to improve the accuracy of data that goes into its Solvency II compliance templates.

The answer for these compliance issues—both for Solvency II and Mifir—may have been out there since at least the end of last year. In December, Stephan Malrait of ING, spoke about getting value for compliance with multiple regulations from firms’ overall data stores—also addressing BCBS 239, Basel III and Emir. At the same time, Gareth Mee, of consultancy EY, said the changes that Solvency II would drive may create pitfalls for asset managers that cannot provide “look-throughs” on data in a timely way. 

The inability to produce the right data is similar to consulting the wrong metric or measurement. Either way, you don’t have the right data or you aren’t working with the correct information.

In Silicon Valley, the company’s founder barges in on a focus group to ask users what they don’t like about his platform, ending up with more information—that might help raise that important daily active users number. In data management, firms probably already have the information they need for compliance. However, choosing and using the right measures of data—and the right measures for the performance of data governance plans—proves to be more challenging.