Often when an investor starts to explore ‘smart beta’ – alternative weighting systems for equity portfolios and benchmarks – it signals some dissatisfaction with traditional active management, traditional market cap-weighted benchmarks, or both. Not so in the case of the Rabobank Pensioenfonds.
For a start, the fund believes in active management. When it overhauled its equity portfolios in 2010, it split its allocation into developed and emerging markets and started working on an enhanced-index strategy that seeks to exploit value and momentum effects for both portfolios, and a fundamental indexation approach with minimum-variance elements for the developed markets. The developed-market strategies have been in place for two years, the emerging markets strategies for about 18 months. But alongside these two smart-beta strategies it has retained traditional bottom-up stockpicking in the mix.
And now consider what it has done within smart beta. Its enhanced-index strategy combines two approaches – momentum and value – that tend to diversify against one another; and the fundamental-indexation approach complements both the bottom-up active portfolio and the value element of the enhanced index.
This is all about balance: the fund wants to exploit fundamental mispricing of companies and, to some extent, short-term systematic anomalies, while not straying too far from the market portfolio. Indeed, both the active management and the smart beta are benchmarked against market cap-weighted indices. In the developed market portfolios, the stockpicking strategy has an expected tracking error of 4%, the enhanced index an expected tracking error of 4%, and the fundamental index-low volatility mix has a tracking error of 6%. Those are relatively high – but the expected tracking error for the three combined is just 3% – and the realised tracking error has so far been more like 2%.
“We feel comfortable that the market-cap index is a good approximation of the overall stock market, and we like to keep things simple: our objective is clear – to realise an excess return above the stock market,” says CIO Bernard Walschots. “The background to our decision to use these different strategies was the idea that using them in combination gives us a higher probability of a high information ratio.”
The fund has covered a lot of the smart-beta bases – fundamentals, factors, minimum variance – but certainly not all. There is no attempt to isolate the size effect, for example. Walschots says that “the academic literature is a little more hesitant about the size effect than it is about the others” and diversification-maximising strategies like equal weighting, risk parity and maximum diversification are also conspicuous by their absence.
“Clearly, any strategy should have a rational explanation as to how it works, but also show that it adds style diversification,” Walschots explains. “With renewed interest in these smart-beta possibilities has come new insights, and one of the latest is maximum diversification, which may turn out to be another robust anomaly – we will see in time.”
While risk parity in the equity-only context does not play a part at Rabobank Pensioenfonds, it has allocated €500m to two unnamed external managers to apply the strategy on the more common multi-asset basis.
“You could see it as a replacement for part of the equity portfolio because we target equity-market volatility but with more balanced risks,” says Walschots. “In practice, however, we will be funding it from our cash balance, but also from our long position in bonds that we would like to reduce.”
Walschots concedes that one logical end-point of a belief in the risk-parity concept is to overhaul the entire portfolio along such lines, and indicates that the fund does indeed intend to grow the allocation over time. For now, however, the fund will monitor the performance of the two allocations before expanding them significantly.
Walschot is extending the alternative thinking into the broader portfolio in other ways. If the size anomaly is not so interesting in equities, it is much more interesting in the credit part of the fund’s liability-matching portfolio – where the market cap-weighted approach forces investors to allocate more money to those borrowers with the most debt.
“We are in the process of an ALM and portfolio construction study which we hope to finish by the middle of 2013, and at that point we hope also to decide on the inclusion of smart-beta into the credit portfolio,” says Walschots. “We started with the minimum-variance effect, and we accept that a low-vol credit portfolio can add value, but we also discovered that there is a size effect and are looking into that.”