There is a school of thought that adding a couple of extra indices to the Your Future, Your Super (YFYS) performance test will fix the bulk of its shortcomings. Sadly, this is incorrect.
The YFYS performance test only measures a component of performance rather than total performance. Because of this, and the benchmarking process, our research has shown that the YFYS performance test will statistically be ineffective. The performance test will create a range of unintended consequences with negative impacts on funds, consumers, and industry structure. Our research (here) has shown that the opportunity cost to consumers could exceed the estimated benefits of the entire YFYS reform package.
It is correct that the benchmarking approach is a major flaw of the YFYS performance test. A limited number of public market benchmarks creates benchmarking noise (variability against the fund’s benchmark of indices). Higher benchmarking noise increases the likelihood of failing the performance test at some point, just through random chance.
Unlisted infrastructure and unlisted property are asset classes which create significant noise under the benchmarking approach. It would be good for these asset classes to be benchmarked more appropriately. But there are many other sectors which face similar benchmarking issues. Consider the fixed interest and credit sector:
- Any attempt to achieve returns above cash through low duration bond strategies will be benchmarked against the composite bond index.
- Any investment in credit sectors (so investment grade through to bank debt, high yield, infrastructure debt and emerging markets debt) will be benchmarked against the composite bond index.
- Long duration bonds, which are used for return-generation and diversification, will be benchmarked against the shorter-duration composite index.
- Any attempts to manage inflation risk through investing in inflation-linked bonds will be benchmarked against the composite bond index.
Points one through four above incur benchmarking noise, in some cases sizable, under the benchmarking arrangements.
Over the long-term (1) would be expected to underperform the benchmark while (2) and (3) would generally be expected to outperform, in some cases to a large degree. I have no firm view on the relative performance of (4).
Each strategy is relevant to super funds focused on member outcomes. But each strategy will be treated differently under the YFYS performance test. It is easy to see the distortion to portfolio decision making that the performance test will create.
We estimate it would take at least twelve more indices to broadly (not precisely) benchmark the fixed interest and credit sector. To clean up all sectors could easily take more than 50 indices. I doubt that APRA couldn’t administer that at present.
Some may say that beyond unlisted infrastructure and property, the other sectors aren’t that important. That view is narrow and focused on the present. Market environments, hence appropriate investment strategies, change. There is also a reasonable (sensible, if well-designed) case for the performance test to be applied to decumulation strategies, an area where there may be greater use of some of these other assets.
We maintain that the YFYS performance test is well-intended, but the design requires improvement. A couple of extra indices isn’t the solution. Better designed, the intention of the test could be achieved without many of the adverse impacts generated by the present design. Some well-considered modifications would have a large positive impact and create minimal distortion to super funds focusing on best member outcomes.