What does good data management mean for fund operators?

Guillermo Donadini, independent advisor and trustee, and Gayathri Pandurangan, Senior Director, Head of Innovation Engineering, KPMG, give their thoughts on capturing, retaining, and utilising data sets.

Fund Operator Editor POSTED ON 9/3/2024 8:00:00 AM

@Pixabay.

The debate around what does being a data-centric organisation mean in modern institutional investment is not new but has taken on new importance in the era of Artificial Intelligence, hybrid working and new legislation around data privacy and sharing.

To discuss these issues, Clear Path Analysis hosted a recent webinar “Utilising data volumes: driving profitability, efficiency and risk management”, which has now been published as a report in conjunction with Rimes Technologies.

In it, Gayathri Pandurangan, Senior Director, Head of Innovation Engineering, KPMG, and Guillermo Donadini, independent advisor and trustee and former Chief Investment Officer - General Insurance, AIG, discuss these issues and give their thoughts.

In the report, the panellists extrapolated on the idea posed that “Gone are the days when the focus was solely on acquiring and mastering data management, the real question now is what do we do with the data once we have it?”

See the excerpt below for their views on the matter from the report.

Andrew: Does good data management mean capturing, retaining, and utilising the widest set of data or focusing on the best and most meaningful quality? How do you discern it?

Gayathri Pandurangan: In the context of institutional investment, good data management is the balance between how you capture a wide array of data and then how you focus on the most meaningful, high-quality data.

On the one side, we are expected to get a lot of market data particularly when you work with the front office teams. In the capturing of the data, you need broad data collection, but you also have to do an exploratory phase to assess what you need to pinpoint. What is the priority and the outcome that you need are important questions when you capture the data.

"It is an iterative process but is one where we have to do the capturing, retaining, and implementing practical steps to discern the data."

When you retain the data, one of the techniques I would advocate is to first do a cost-benefit analysis. Yes, you may have started acquiring a lot of data but is it every few months, weeks, etc.? You want to see what the storage costs are versus the value add to see if you are benefiting from it.

So, for example, you might have 10 years’ worth of data but then it’s only applicable for one out of 100 use cases so you’re just paying for storage, capacity, and computing ability.

Practically, data governance is so important from start to end and advanced analytics is key as to how you analyse what the data does. It is not possible to check the data through manual means, you have to automate as much as possible to have quality.

Feedback loops are also something that is key as if you go back to your sources of data, whether it is the trading, risk, or compliance teams, they all produce data so having a constant feedback loop to be able to separate real data from the noise is critical. It is an iterative process but is one where we have to do the capturing, retaining, and implementing practical steps to discern the data.

Andrew: What about the opinion that you don’t need all the data in the world but instead you need more ways to sift through what you do have?

Guillermo Donadini: I believe the answer depends on who you are asking. In my experience, if you ask someone with a global responsibility, they will have a view but if you have to go into a global investment committee, aggregation is important. Maybe what happened in the last two years might be important but that is it.

However, if you are in 50 countries and you have different regulators worldwide and they all have a different approach to data retention as well as different boards within these countries, they will all come with different answers.

It comes down to governance and you want to make sure that everyone who has a say has one and is heard.

"It is a dynamic process, and technology does also come into consideration because the access of the information and processing of the information changes quite a lot."

It is hard to find a solution because it depends on whether you have one asset class or five asset classes or whether you operate in one country or 50 countries. It also depends on if you have one decision-maker or if the decision-making process is top-down or bottom-up.

I believe that the overall complexity and the number of stakeholders basically decides what data you need, how long you have to keep it and what is relevant for one might not be relevant for another.

If you also add the asset managers within your organisation, they will have a different perspective on this so it could be complicated.

I believe the best way to do it is to put in place a budget, to say what the money is that you have and to make sure that you prioritise what is the need to have, and you can have your own way of doing it and what is the like to have.

It is a never-ending story and a process of rediscovering and reviewing. It is a dynamic process, and technology does also come into consideration because the access of the information and processing of the information changes quite a lot.

You can read more of Pandurangan and Donadini’s thoughts, and the report in full, by clicking here

 

Please Sign In or Register to leave a Comment.