Defensive assets should be determined on an individual asset basis, rather than on an asset class level, delegates heard at the Conexus Financial Fiduciary Investors Symposium.

Paul Caskey, chief investment officer of the South Australian work injury insurance scheme ReturnToWorkSA, presented a new formula for helping to determine which funds could be included in the SuperRatings SR Balanced (60-76) survey.

After conversations with Sam Sicilia, chief investment officer of HOSTPLUS, Caskey proposes that defensive assets should be chosen because of their low volatility and for their low correlation to global equities.

The new model would also bring uniformity to the situation where some funds in the SuperRatings SR Balanced (60-76) survey use the strict definition of a defensive asset as cash and fixed income and where others count infrastructure and unlisted property assets.

Caskey said: “The real change that we are suggesting, is that stable assets do make sense as defensive assets.”

Sicilia said the advantage of this approach was that funds such as HOSTPLUS already carried out such analysis on all the assets they held.

“If we were to adopt this, all players can play this game,” he said, adding that funds should update their defensive growth split once a year when they set their strategic asset allocation.

The proposal received equal measures of approval and agreement from delegates.

The chief objection was that the new proposal could be gamed in the same way as the current defensive split.

It was pointed out too that the new proposal would gain more widespread uniformity of use if were not to be used in a competitive performance survey.

Another objection was that a three bucket approach would be more accurate, to which Caskey replied that this would create more complexity and that a two bucket approach allowed for continuity with the current SuperRatings’ survey.

The need for innovation was highlighted by Steven Jackman, director of investment consulting at PwC, whose analysis of funds in the SuperRatings’s survey showed that some funds had as much as 98 per cent in growth assets under the strict definition, while some had only 60-64 per cent in growth assets.

Furthermore, the majority of funds were grouped at around 75-76 per cent growth assets, which suggested they might be gaming the process.