AustralianSuper’s head of finance and operations, Peter Curtis said the country’s largest superannuation fund was weighing whether to bring more investment operations in-house including private assets.
Curtis said the $180 billion fund was spending more time analysing whether an investment process should be outsourced to their custodian, JPMorgan. He said risk reporting and analysis, particularly around private equity, were areas that could be internalised.
“Three or four years ago we would see something new and throw it over the fence to the custodian,” Curtis told delegates at the Investment Magazine‘s Investment Operation Conference in Sydney this week. “We are now spending much more time thinking: Is it their skillset or is it something they we can actually do? This is especially as we are putting more money in larger amounts into unlisted assets.”
Curtis, who was appointed to his current role in October, said the fund was in the midst of a big project with JPMorgan that had identified a “large number” of investment processes which were managed by the custodian but were also unique to AustralianSuper. He said it was those processes which could be brought in-house.
“It’s a continuing evolution as we get bigger and start to do new things,” he said. “It gets back to where is the custodian in their product road map and is that an area that they want to play in.”
Speaking on a panel with delegates from Australian Custodial Services Association (ACSA), CLS Group and NAB Asset Servicing, Curtis said the need to access granular levels of data to manage counterparty risk and exposure from its securities lending program, had led to the fund’s decision to internalise risk reporting.
It is a “competency that we want in-house,” he told delegates. “It is something where we are going to have to bring the data in and put it into our own risk systems, because we can’t get the view from an individual investment option in the way that we think about the portfolio.”
Curtis said while AustralianSuper could “sort of get” a view of counterparty exposures at a whole of fund level from their custodian, it became more problematic when the fund wanted to drill down into a specific investment option to understand the risk exposures.
Margaret Law, head of client management at CLS, said risk analysis was becoming increasingly challenging for assets owners because of the complexity of the investments. She added that there were many funds, not just in Australia, that “may not have a good grip” of the growing level of risk in their portfolio.
A question of data
NAB general manager Allyson Bradnam said custodians had always provided data as a service, but it was the “granularity of the data elements” that clients were now looking for. ACSA’s chief executive Rob Brown said while data as a service was “embedded in the old world” of custodians, it worked in a very “linear way”.
“In an unstructured data-driven space, where you don’t even know what question you want to ask yet but you know you need the data to answer it, I think that is a different pathway,” Brown said. “We’ve got a very interesting competitive landscape that will continue to evolve.
“Will there be demand and will technology align to provide solutions for large and sophisticated funds for data? Yes. Will those solutions necessarily come from the custody community? Maybe not.”
As for private assets, Curtis said it was another area of operations that he may bring in-house because there was no one model from the custody perspective that worked well across the whole unlisted space.
“I’m actually wondering whether this is one of those areas where we actually might need to do it ourselves,” he said. “Our fund has been investing in those assets for yonks, the custodian industry hasn’t come up with a solution, so maybe that is just not there strength and we need to take more responsibility for that internally.”
Bradnam said while there were multiple stakeholders involved, private equity remained an underlying issue that the whole custodian industry needed to “tackle.” Brown said from a back-office perspective, finding the right model could be done, it just needed “a lot of people to be convinced that it should be done.”