Data Management In Investment Management – Top Tips from TSAM

By HFI staff


Last week the Summit for Asset Management (TSAM) presented 6 co-located conferences in London, providing sessions around technology, operational strategy, data management, communications, client reporting, performance, risk, regulations and compliance. Hedge Fund Insight focused on the data management stream, and these are the highlights from the first presentation in that stream:

Creating the Foundations of Value in Data Management

Binary 2

From Andrew Barnett, Chief Data Officer of Legal & General Investment Management

Investment management is an immature sector for some of the data tools that are used elsewhere, making it a good sector for data scientists to enter.

Understand the culture of the organisation before changing any of the data management.

Appreciate that your data strategy once determined will not remain fixed – it will change every 6-9 months. It should be framed very simply, so that the CEO can understand it and it can be communicated to an 8-year old.

Binary 2 a

You should plan to execute against the widest capability, and look for quick wins. A Data Management Office should be able to add value easily in vendor management, for example.

There have been three attempts at starting-over in data management at LGIM, so there are three Enterprise-wide Data Management Systems. If that seems that “the cloak of legacy” has to be fought as it may take a million pounds to implement a new system, but it takes five million pounds to rip something out.

Consequently, when we set up a new fund we have to set it up on five systems – we can’t switch them off!

Binary 2 b

In his head, as a Chief Data Officer (CDO) Barnett has two-and-a-half years to show that his work has created value for the business – that is the average tenure for a CDO.

The business has to be involved frequently in large data management projects – feeding back on how the specifying has been interpreted by the data specialists. The business can’t say “we’ll see you in 12 months time at the project deadline to see if you have got it right.” Instead there should be showcases and reveals en route.

Build a Data Dictionary to share your resource through the business. This will be helped by using a range of visualization techniques like dashboards. It is a good idea to include a glossary of terms so that where a term like “duration” is used as a field or output the user knows that it means modified duration or macaulay duration, as the case may be. These kind of initiatives allow the users to be more self-serving.

Binary 2 c

LGIM’s CDO has these building blocks for the data management effort – Content, Governance, People, and Services. There are SLAs in place across the business, as users have got comfortable with the governed data lake they know they can rely on that data source.

The  data professionals at Legal & General IM can take placements within the business units – this exposes data team members to what goes on outside the department, and the capabilities of the team become more widely known.

Although it is tempting, do not waste effort trying to procure resource for an Innovation Lab. Although the business may benefit from such an initiative it will never have a Net Present Value, such that you can’t get a capital allocation for it.

CDO Barnett has been at L&GIM for 3 years. It has taken 18 months to get to the point of being able to innovate.