Big Banks, Big Data... Big Opportunities?
A UK government commissioned report has concluded that there is a broad market for exploitation of bank data. The report contains a strong call for pro-consumer open standards but without any real
A UK government commissioned report has concluded that there is a broad market for exploitation of bank data. The report contains a strong call for pro-consumer open standards but without any real economic analysis of the cost and consequences for the banking sector.
The HM Treasury report, written by the Open Data Institute and Fingleton Associates, promotes open standards in data exchange. Open standards would allow customers to easily use comparison websites, account aggregators and other banks' services. However, many banks see control of their own data as their lifeblood. Banks may well prove reluctant to share nicely if it affects their competitive advantage.
In supporting its findings, the Treasury report cites McKinsey research that making consumer finance data more accessible could generate $210-280bn of value globally. These staggering (if, McKinsey admits, tentative) figures are based on better financial data analytics aiding risk assessment and lending decisions. To achieve these huge benefits, a degree of data sharing is vital. However, McKinsey's figures assume a combination of open data and closed proprietary data, a point which the report ignores. A balance is necessary to allow consumer choice to grow; while encouraging continued bank investment in robust infrastructure and innovative services.
The report is confused on data protection issues, assuming all sharing will be consent based. The paper's call for financial personal data to be treated as sensitive data "as a matter of course" only creates barriers to data sharing. The position is regulated at European level and does not currently require financial data to be treated as sensitive. No doubt, any future government-led data sharing initiative will need rules of engagement compatible with data protection law. The ultimate model needs to be practical as well as compliant to allow the best use of shared data.
ODI and Fingleton reach tentative conclusions (based primarily on consultants views and re-use of current open standards) that implementing data sharing should cost less than £1m per bank and be achievable in under a year. This seems to assume a less complex landscape than in many banks with diverse data formats across their retail, mortgage, investment and pensions businesses. Perhaps the figures will be closer to those given by one of the banks themselves, that "it would probably take 'tens of millions of pounds'. Not only is it unclear who will foot the bill, but also whether banks will be able to charge for providing the platform for other providers to piggy back on.
Some other small set up costs are identified and ongoing costs are suggested to be "very low". The report (in assuming current open standards are sufficient) does not address detailed development of taxonomies, regulation and a robust framework. These initial costings are somewhat in contrast to the Health and Social Care Information Centre's £200m per year budget. Perhaps the creation of HSCIC to share government health data provides some insights about setting up data sharing schemes (and the teething problems involved), government leadership, data protection, timing and costs.
The UK's approach reflects a global drive to promote government data sharing. Access to detailed statistics on UK citizens should boost the economy. However, bank proprietary data is a different matter to public records. The Treasury Call for Evidence in the new year will no doubt raise some robust responses from banks, industry bodies and consumer associations.
As a first step towards exploring government-led open data standards in the banking sector, the report provides ample food for thought. Open standards will open up a world of consumer choice. However, the thinking needs development, both on the overall economic impact for the UK financial services sector; and the practicalities of a complex programme with ramifications far broader than just IT implementation.
The HM Treasury report, written by the Open Data Institute and Fingleton Associates, promotes open standards in data exchange. Open standards would allow customers to easily use comparison websites, account aggregators and other banks' services. However, many banks see control of their own data as their lifeblood. Banks may well prove reluctant to share nicely if it affects their competitive advantage.
In supporting its findings, the Treasury report cites McKinsey research that making consumer finance data more accessible could generate $210-280bn of value globally. These staggering (if, McKinsey admits, tentative) figures are based on better financial data analytics aiding risk assessment and lending decisions. To achieve these huge benefits, a degree of data sharing is vital. However, McKinsey's figures assume a combination of open data and closed proprietary data, a point which the report ignores. A balance is necessary to allow consumer choice to grow; while encouraging continued bank investment in robust infrastructure and innovative services.
The report is confused on data protection issues, assuming all sharing will be consent based. The paper's call for financial personal data to be treated as sensitive data "as a matter of course" only creates barriers to data sharing. The position is regulated at European level and does not currently require financial data to be treated as sensitive. No doubt, any future government-led data sharing initiative will need rules of engagement compatible with data protection law. The ultimate model needs to be practical as well as compliant to allow the best use of shared data.
ODI and Fingleton reach tentative conclusions (based primarily on consultants views and re-use of current open standards) that implementing data sharing should cost less than £1m per bank and be achievable in under a year. This seems to assume a less complex landscape than in many banks with diverse data formats across their retail, mortgage, investment and pensions businesses. Perhaps the figures will be closer to those given by one of the banks themselves, that "it would probably take 'tens of millions of pounds'. Not only is it unclear who will foot the bill, but also whether banks will be able to charge for providing the platform for other providers to piggy back on.
Some other small set up costs are identified and ongoing costs are suggested to be "very low". The report (in assuming current open standards are sufficient) does not address detailed development of taxonomies, regulation and a robust framework. These initial costings are somewhat in contrast to the Health and Social Care Information Centre's £200m per year budget. Perhaps the creation of HSCIC to share government health data provides some insights about setting up data sharing schemes (and the teething problems involved), government leadership, data protection, timing and costs.
The UK's approach reflects a global drive to promote government data sharing. Access to detailed statistics on UK citizens should boost the economy. However, bank proprietary data is a different matter to public records. The Treasury Call for Evidence in the new year will no doubt raise some robust responses from banks, industry bodies and consumer associations.
As a first step towards exploring government-led open data standards in the banking sector, the report provides ample food for thought. Open standards will open up a world of consumer choice. However, the thinking needs development, both on the overall economic impact for the UK financial services sector; and the practicalities of a complex programme with ramifications far broader than just IT implementation.