Data becoming siloed across different parts of an investment management business is akin to an assembly line producing cars with mismatched wheels, and all levels of the organisation need to take responsibility for preventing this, experts say.

Data needs to be captured and catalogued properly at the source, traced to accountable parties and put through a robust quality control process to establish trust, said Zion Hilelly, Managing director, head of solutions operations & transformation at IHS Markit, in a panel discussion on data management at Investment Magazine’s Investment Operations Conference in late March.

An audience survey during the panel discussion found many institutional investors in Australia share his concerns, with only 9 per cent saying they “strongly agree” with the statement that they trust the data that critical decisions are based upon. The most popular answer from 46 per cent of respondents was to somewhat hesitantly “agree” with the statement.


Hilelly gave numerous examples of the potential risks of siloed data. If a portfolio manager is trading using a different dataset to the team measuring performance, measurements of risk and reward will be inaccurate. Mismatched data can lead to the reconciliation of a different cash balance to what is actually traded, leading to overdrafts or uninvested cash balances. It can also have serious implications for compliance and the prevention of violations.

But while high data standards often seem simple and obvious, they are hard to follow unless applied in practical ways, he said. This can mean hiring someone to focus on collecting data rather than expecting people to balance it with myriad other tasks, and educating the workplace on the sourcing and impacts of data collection upstream and downstream.

“Too often we see a disconnect between the governance function and the ability of data owners to actually adhere to their standards,” Hilelly said.

“An organisation should make sure the data owners are equipped and empowered with tools and business processes to follow the guiding principles set for data.”

Short-fallings in how data is managed and reconciled often come down to an organisation’s attitude toward its data stewards and data owners, said Dominic Dowd, Director, North America, at Shoreline Consulting. Data stewards in various parts of the organisation should be seen as data producers, he said, in what can often be a thankless job that comes with expectations and obligations that can be hard to fulfil.

“The stewards and the owners should really be integral members of any good governance committee, both at the executive level and then right through to the operational committees that are managing the day-to-day practical matters,” Dowd said.

This requires a cultural change so data owners genuinely understand the value of good data, and are empowered to produce it, he said.

“Data owners should be thinking about producing the best quality product and that means rapidly and continuously refining, updating and enhancing their product as a best practice,” Dowd said. “Let the producer manage the day-to-day…if you fix it at the source the entire organisation can benefit.”

Join the discussion