COO view: Why good data management is all about agility and using data dynamically

Davina Goodall-Smith, Chief Operating Officer – EMEA at Nikko Asset Management Europe, explores how to find the right balance when outsourcing and why it is fundamental to have a common understanding of data.

Sara Benwell POSTED ON 3/22/2021 4:33:36 PM

Davina Goodall-Smith, Chief Operating Officer – EMEA at Nikko Asset Management Europe

Sara Benwell: What does control and ownership mean to your business?

Davina Goodall-Smith: Ownership is about the ability to access the data and then being able to report, inform, or make decisions off the back of it. It does not necessarily mean physically holding onto the data as was seen historically.

A common theme throughout the industry is that regardless of whether or not you leverage data as a service or outsource, there is always a requirement to own your own data strategy.

“The way we think about data now, control is about agility, the ease of access”

You can’t outsource this piece of it, and this is where the intellectual property and the value add comes into it, as you need to ensure that the data you have is fit for purpose.

We also maintain an internal structure to support the data, so again, not around the nuts, bolts, and processes anymore but about understanding the data, where it is, and how we can use it.

The definition of control has also changed. The way we think about data now, control is about agility, the ease of access, the ease of taking this data and using it for what we need the data to do. It is no longer “data is king”, but rather what you do with the data that is most important.

Sara: Where do you draw the line between control and ownership? How integrated are your data management functions and how do you maintain data integrity?

Davina: Data is a massive topic, and I am not a data specialist. I have a responsibility across many different areas and so trying to grasp data as a subject is difficult, and we are actually going through a refresh of our global data strategy at the moment.

I am a visual person and so the easiest way for me to understand how we are approaching the data strategy is to think of data in a triangle. The nuts and bolts, the factory-level data, provides the foundation of that pyramid.

These are items that can have a standard process around them, standard attributes with pricing, corporate actions, income, and daily NAVs, etc. These would be the standard processes, and this is where you can put standardised controls, processes, and have a standard set of outputs.

When we talk about data as a service and being able to leverage an industrialised process, this is where it is easiest to outsource.

“Don’t go to a service provider wanting a standardised process at a good cost and then ask for all sorts of different things in it”

Taking this set of data, which has been applied to a standard set of processes and defined set of attributes, and getting a good cost ratio for this, compared to doing this inhouse is always a key factor as a Chief Operating Officer.

For this method, it is important to adopt a standardised process, don’t go to a service provider wanting a standardised process at a good cost and then ask for all sorts of different things in it.

It then becomes a bit trickier as to what you put in the middle and top layers.

The true top layer is the value add or the dynamic data. This is everything around the investment process, risk process, and even real specific requirements for your end clients i.e., your institutional investors, pension funds, etc.

“The true top layer is the value add or the dynamic data”

The reason I put this at the top of the pyramid, and that it is becoming value add, is because this is where you can take raw data.

You need to be dynamic with it, be able to run what if scenarios and analysis, your investment managers or other areas may need to do something different with the data on any given day or hour depending on what is happening with a political situation or market event.

When you ask the question about where you draw the line between ownership and control, there is a line somewhere there that from a data as a service perspective and from an outsourcing point of view, you’ve got your building blocks.

This is a very easy decision to make with our global partner to say that we can outsource it. It is this top of the pyramid that then becomes something of value and how we manage it.

Sara: If you don’t hold onto data physically any longer, what impact does this have on data lineage for reporting?

Davina: Having worked for a service provider, as well as now working with an asset manager, the absolute core of data is understanding it and mapping it with those boring items.

For example, creating and using data dictionaries so that you have a common understanding across whoever is holding the data, be it an outsource provider or internally.

In my first experience, 4-5 years ago, of being a relationship manager and asking for what my client wanted whilst working for a service provider, the conversations back then was that there was a cloud thing coming along.

“The absolute core of data is understanding it and mapping it with those boring items“

It would be great because we could chuck all our data in there and then we would be able to do absolutely everything we wanted with it and manipulate it.

If you never had any experience of data this sounds great, to put it all in and then you can do what you like with it, but you really need to understand what it is underneath.

Data as a service isn’t some kind of panacea that resolves all of your issues, but it does make you ask yourself all of these questions.

“Data as a service isn’t some kind of panacea that resolves all of your issues”

It is absolutely right that a service provider should understand this, but equally, you should know your data as an asset manager or asset owner if you are managing your own data.

This was the key for me in understanding data and how it was moving forward. It is about structure, data dictionaries, data mapping, and having a common understanding of the definition of all your data attributes.

Only when you have done this can you do everything else around manipulating and where you should hold your data.

 

Please Sign In or Register to leave a Comment.