Blog

 

Building the Right Investment Data Model

Monday, 15 August 2016

Earlier this month Bridge hosted roundtable discussions in Melbourne and Sydney on approaches taken by investment managers and superannuation funds to build the right data model.

The below blog summarises some key aspects of this discussion:

Investment managers’ data needs once were simple; based on simple investment processes, in-house systems and fewer regulatory requirements.  Investment data management wasn’t seen as something distinct, instead it was incorporated into middle and back office functions.   

Complexity: advent of Outsourcing & the GFC

Over time investment managers started to change their operating model and outsource their back office.  In doing so they lost control of their ‘book of record’ and their ability to extract investment data to support their own investment process.  Meanwhile, the Global Financial Crisis drove the need for increased volume, timeliness and transparency of investment data requiring complex, data-hungry systems.  The once-simple investment data model was suddenly complex, and it became difficult for investment managers to gather the data they required.

In response, the investment industry has sought to treat the symptoms, not the cause, by either:

  • deploying multiple single-point solutions;
  • not ensuring front office data requirements are tangibly tied back to the investment process; or
  • deploying a data warehouse as a panacea to their data needs without knowing their enterprise data needs or ensuring appropriate data governance. 

The common result has been an investment data model with multiple sources of truth, which also contains either inaccurate or stale data, and which is not aligned to the needs of the business.

Success Factors: the strategic importance of the Data Management Function

In order to address the current issues with data management for asset managers, asset owners and by extension custodians, Bridge has identified the following keys to success.

1.   Recognise the strategic importance of the data management function
Given its importance, data management should be recognised as a discrete business function.  The Data Management Function should be independent from other functions, have executive endorsement and a clear definition of its mandate.  This will help ensure proper ownership, standardisation, efficiency and prioritisation of data-related issues.

2.   Implement the ‘right-sized’ investment data governance
An Investment Data Governance Model should be put in place to ensure all relevant aspects of data management can be addressed in an agreed framework.  Recognising that excessive governance can be as detrimental as a lack of governance, the level of governance should be reflective of the complexity of organisations’ investment data model and culture.

3.   Define and enforce data structures, hierarchies and classifications
Investment classifications and structures should be defined and implemented in a logical manner.  This does not mean that there is only a single classification model that cannot be adapted to suit the needs of different users (such as investment teams).  Once developed, it is imperative that the data model is enforced and adapted through rigorous change control mechanisms.

4.   Develop a target state data architecture model that supports the target operating model  
A target state data architecture model should be developed that outlines the intended data flows and integration across the enterprise and is supportive of the target state functional model and business processes.  The level of granularity in the target state design depends on the level of uncertainty and time horizon of the target state view. A conceptual vision is more important and relevant than detailed future state planning which could be subject to change.

5.   Find right sized IT solutions to support the target architecture
In searching for supporting IT solutions it is probable that one system will not support the entire target architecture. Instead a modular approach to roll-out should be considered, where the end solution is split into components and prioritised (e.g. ETL/Data Quality Management, Data Warehouse, BI Tools etc.).   Custodians and OMS vendors may also be able to assist in supporting your requirements instead of maintaining all data internally.

Current trends – solutions from investment management partners arise

As the investment industry’s data needs and challenges increase, there are a number of developments occurring within the industry which promise to assist in dealing with the issues at hand.  These include:

  • Front Office vendors entering the data management space – order management system vendors are entering into new areas and increasing the data management capability they offer.  Many are beginning to offer combinations of ABOR, IBOR, managed data or business process outsourcing services where previously they may have only offered one (or none) of these solutions.
  • Custodians delivering IBOR related services – custodians are developing their investment data capability into services which are independent of their accounting related functions.  This includes the example of one global custodian who has developed and ‘hub and spoke’ model enabling their client to leverage the custodian’s infrastructure and support while enabling the client to undertake customised data management activities.
  • An increasing number of ‘Books of Records’ (BORs) – different business functions have increasingly different demands on data in terms of scope, timing and accuracy.  These include Compliance, Risk, Portfolio Management and Performance teams all requiring their own view of data and BOR in addition to the ABOR and IBOR.  Increasingly, a single source of truth does not mean there is only a single view of data.
  • Data Lakes replacing Data Warehouses – a Data Lake allows data to be stored in an unstructured manner, facilitating low-cost storage of significant amounts of data.  This ‘sand pit’ environment enables agile and ad-hoc analysis of data by different users (e.g. investment teams) and is supportive of the increasing demand for different uses of data.

Investment Data Management – a new hope

As the investment industry begins to address its current investment data management challenges Bridge has identified three key guiding points to assist clients in the design of their Investment Data Model:

  1. Treating investment data management as a strategic business priority is the first step needed to get on the right path;
  2. Business requirements and target state architecture should define the solution – not the other way around;
  3. New solutions are emerging from investment manager partners. These may help organisations to avoid the ‘build your own solution’ path.

How Bridge can help

Leveraging its capabilities in Operations and IT, Bridge is able to assist asset managers and asset owners with their data management needs in the following ways:

  • Data management function design
  • Data governance model design
  • Investment data definition and classification
  • Data architecture design
  • Vendor selection 

For further information about how Bridge can help your organisation in the area of investment data management, or to find out about future Bridge roundtable events please contact us.

Bruce Russell, Director

 

In order to address the current issues with data management for asset managers, asset owners and by extension custodians, Bridge has identified the following keys to success.

Recognise the strategic importance of the data management function

Given its importance, data management should be recognised as a discrete business function.  The Data Management Function should be independent from other functions, have executive endorsement and a clear definition of its mandate.  This will help ensure proper ownership, standardisation, efficiency and prioritisation of data-related issues.

 

Implement the ‘right-sized’ investment data governance

An Investment Data Governance Model should be put in place to ensure all relevant aspects of data management can be addressed in an agreed framework.  Recognising that excessive governance can be as detrimental as a lack of governance, the level of governance should be reflective of the complexity of organisations’ investment data model and culture.

Define and enforce data structures, hierarchies and classifications

Investment classifications and structures should be defined and implemented in a logical manner.  This does not mean that there is only a single classification model that cannot be adapted to suit the needs of different users (such as investment teams).  Once developed, it is imperative that the data model is enforced and adapted through rigorous change control mechanisms.

Develop a target state data architecture model that supports the target operating model  
 
A target state data architecture model should be developed that outlines the intended data flows and integration across the enterprise and is supportive of the target state functional model and business processes.  The level of granularity in the target state design depends on the level of uncertainty and time horizon of the target state view. A conceptual vision is more important and relevant that detailed future state planning which could be subject to change.

Find right sized IT solutions to support the target architecture

In searching for supporting IT solutions it is probable that one system will not support the entire target architecture. Instead a modular approach to roll-out should be considered, where the end solution is split into components and prioritised (e.g. ETL/Data Quality Management, Data Warehouse, BI Tools etc.).   Custodians and OMS vendors may also be able to assist in supporting your requirements instead of maintaining all data internally.