Implementing the Hybrid Model
Options and considerations for getting out of the business of loading and feeding data
The second RIMES 2nd C–Level Round-Table, hosted late 2014 in London, developed viewpoints raised at the first meeting: all participants expressed a strong desire to get out of the business of loading and feeding data. It was agreed that incremental improvements in data management were insufficient to deliver strategic benefits. A radically different approach is required to deliver transformation in a time of change.
Data governance is an ongoing challenge that requires standards, policies and perseverance. Moreover, data must be owned by the business to ensure that it is always fit for purpose, so a department-by-department approach is essential. Several participants reported that the data management team is often blamed when data is found to be unfit for purpose; however, with users and uses constantly changing, it is unrealistic to expect the data management function to own and control all of the data within the firm.
Data quality is crucial to business success, but it cannot be managed at the corporate level as it goes through numerous changes and there are always multiple layers. There are also many aspects to data quality, including timeliness, accuracy and increasingly, transparency.
The meeting considered the potential benefits of the hybrid data operating model.
Meeting tactical needs and strategic ambition
Most firms face many similar data management challenges and participants noted some universal bad practices, such as manipulating and storing data in spreadsheets. All firms need a data architecture that can respond quickly to new business opportunities and deliver strategic benefits.
RIMES proposes a hybrid data operating model, which combines traditional in-house data management with managed data services. This enables firms to manage proprietary data in house while delegating the management of third party and specialist data to an expert service partner. One of the participants has adopted the hybrid approach and reported increased data transparency and improved business flexibility.
The hybrid model seeks to ensure that data sourced meets exact business needs, with sufficient controls. As well as being good practice, controls at the point of usage are increasingly necessary to meet emerging legislative requirements, such as Solvency II.
By implementing the hybrid model firms can adopt a proactive approach to data management, based on future requirements rather than making changes retrospectively, often in response to complaints or inaccuracies.
Participants agreed that many firms do not have enough in-house expertise spanning data and business functions; data quality must be measured, maintained and controlled in the context of its business use.
Many firms start out with the same data but it is transformed in-house to meet different needs. In practice, all firms need to maintain ‘multiple versions of the truth’ to meet the various needs of heterogeneous users. The hybrid model enables this and allows firms to make informed choices about which data to manage in house and which to delegate to a third party.
Supporting multiple views and data uses
It was agreed that it can be difficult to understand who is accessing which data and for what purposes. With increasing requests for more granularity of views, for example to ascertain direct and indirect exposure, much data is transformed and augmented in house. In most firms, there is a need to manage both granular and enterprise data. Participants supported the idea of a federated data model to accommodate different standards in diverse regions.
The real problems arise when data is distributed externally to clients or third parties, because that is when reputational risk is at its highest. There is a universal need for comprehensive data lineage. However, a ‘golden source’ may be good for some purposes but not for others – it depends on specific data and systems. The problem is exacerbated as data complexity increases.
Benefits of the hybrid model
Participants agreed that working with an expert partner could ease the burden of data management and help firms source the exact data necessary to run the business. The hybrid model accommodates change and supports a strategic approach to data management.
The second RIMES C-Level Round-Table demonstrates an appetite for change at a senior level within many investment management firms. Although the discussion is ongoing, several consistent themes emerge:
- Time is scarce and resources are stretched. Asset management organizations have to manage BAU (business as usual), strategic change and the impact of existing and emerging regulations! Multi-year monolithic change management programs are no longer a viable solution.
- There is a clear demand from all stakeholders, internal and external, for improved control and visibility of data; in other words, good data governance. However, this has to be achieved without impeding business flexibility and agility.
- Many organizations would like to get out of the business of loading and feeding data. Some are already embracing this path by using managed data services in a hybrid model to enhance and supplement traditional data management operations.
We are very grateful for the contributions of C-Level participants to the data management debate and we are committed to developing these themes further during 2015.
- RIMES partners with AWS to offer its ETF data to the AWS Data Exchange’s millions of users
- Meeting the Ethical Obligations of Data Governance
- Constant Vigilance and Action are Crucial to Deliver on Diversity and Inclusion
- The ETF Market Calls for a Customized Approach to Data Management
- RIMES brings its ETF Data Management solution to Snowflake Data Marketplace amidst global ETF boom