The first problem was the inherent volatility of the SAA approach, the second was the assumption that all members were created equal, and the third was path dependency, which was the order in which returns were received and thus determined the final outcome, Cooper said.
Cooper called on the industry to examine its assumptions.
"The approach must change. The assumptions we make are wrong. There are a large number of constraints in SAA," he said.
"We can't hide in the averages - this is real people's real money. We have to be far more flexible in our approach."
He said the industry spent too much time thinking about asset class buckets and "slicing the pie vertically".
"But, we need to slice across the assets - not vertically, but horizontally - and get managers to focus on aggregated outcome," he said.
He contrasted a traditional allocation in a balanced fund with an objective-based real-return fund.
The traditional strategy (using the Schroder Balanced Fund as an example) was Australian equities 29 per cent, global equities 28 per cent, cash 13 per cent, Australian fixed income 13 per cent, objective-based 10 per cent and high-yielding credit 7 per cent.
In contrast, objective-based allocation (using the Schroder Real Return Fund Wholesale Class) was cash 29 per cent, high-yielding credit 20 per cent, global equities 15 per cent, Australian fixed income 10 per cent, Australian equities 9 per cent, inflation-linked bonds 8 per cent, mortgages and sub-debt 8 per cent and absolute return strategies 1 per cent.
"The problem with the SAA approach is that we don't take account of the fact that different members start at different times," Cooper said.