Giant step into a new world
Last year, PGGM, the Netherlands’ second largest pension fund, took the revolutionary decision to invest in the commodities markets with a commitment of between 3% and 5% of its total portfolio value of about e50bn. This was no sudden move, but came after an extensive asset and liability management (ALM) study, which brought other changes in its wake, including a further increase in its private equity allocation, as well as the implementation of a 100% hedge of nearly all currency risks.
Remarkably, the increased allocation to private equity is closely connected to the decision to start investing in commodities. As senior investment manager Jelle Beenen says: “Private equity is interesting for its returns, but the risks attached did preclude an increase in allocation. If the ALM study was done without taking commodities into account, by having a portfolio of equities, private equity, fixed income and real estate, to increase the private equity allocation within the existing allocations, increased the risk to an unacceptable level. But when commodities were added then we found we could have a higher private equity allocation, now with an acceptable level of risk.”
Two characteristics of commodities as an asset class are at work here, he points out. “Firstly, there is the negative correlation to other asset classes. This is due to the fact that commodity prices are primarily based on the current supply and demand picture, whereas the values of other asset classes are based on expectations about the future.” For instance, when approaching the top of the economic cycle, equity prices decrease. At the same time commodity price levels are at a maximum because of the high demand due to the economic activity. “These high commodity prices in turn slow down the economy, putting pressure on equity prices.”
Secondly, commodities are slightly positively correlated to the liabilities of the pension fund. The fund’s obligations are indexed to wage levels and therefore positively correlated to inflation, as are commodities.
The ALM exercise showed the negative correlation to other asset classes to be of greater importance to the fund than the positive correlation to inflation, says Beenen.
“At the same time if you just add commodities within the existing assets, but without increasing private equity, commodities did not have much added value.” He adds that it is important to realise that the study was designed specifically for PGGM’s assets and liabilities and might not be fully applicable to other funds. “Another fund might have different risk and other preferences and come to a different conclusion.”
The ALM study showed a double figure percentage of the total portfolio as the preferred allocation to commodities, though that would be completely impractical, he says. “Our impression is that other funds undertaking ALM studies taking commodities into account would also come up with a double figure percentage allocation to the class.”
The allocation range of 3–5% implies an exposure of around e2bn, a figure that makes PGGM one of the world’s largest institutional investors in commodities. The fund decided to invest in commodity futures rather than physical commodities. “In physicals, there are the obvious difficulties of storage and rentals, which cannot be implemented quickly.” But there are more advantages to futures. As futures approach their expiration, they have to be rolled to a future contract further out. In commodity markets nearby futures often trade at a higher level than futures with a later expiration, which is known as ‘backwardation’. Particularly in energy markets during periods of low supply and high demand, and therefore high spot prices, backwardation is a common phenomenon. “In this way, the roll adds to the dynamics too! Backwardation during the rolling of futures helps to achieve more attractive returns than buy and hold investments in physical commodities.”
Since it is the role commodities play in the mix of assets that is of interest rather than the returns they are expected to bring, PGGM regards the investment as a strategic allocation. Therefore the performance should be measured against a customised but transparent benchmark to meet the Dutch regulator’s requirements. “So the choice of an appropriate benchmark had to be taken with considerable care. Key phrases here are transparency, public accessibility and investment possibilities,” says Beenen. Liquidity and ease of investment are important issues as well.
These considerations led PGGM to use the Goldman Sachs Commodity Index (GSCI) for benchmarking commodities investments. The GSCI is calculated in a fixed formulaic manner from commodity future prices, which makes it transparent. Moreover there is an accessible market in futures, total return swaps, certificates and structured notes on the GSCI and related benchmarks. Other counterparties besides Goldman Sachs are providing products on the GSCI as well. “This is not the case with some other commodity indices, which are more or less ‘owned’ by just one party,” he says. To capture the result of the rolling of commodity futures in the benchmark as well, PGGM’s benchmark for commodities investment is specifically linked to the GSCI Total Return (GSCI TR) index. This index replicates the return of a fully collateralised portfolio of commodity futures, using the weights of the GSCI.
One feature of the index is the heavy weight of energy-related futures in the index ,which together constitute about two thirds. “The energies futures market is quite liquid, while with others like livestock and cattle, the volumes traded can be pretty low.” Altogether there are 26 commodity contracts included in the index, including agricultural commodities (17%), livestock (9%), precious metals (2%), while industrial metals make up 7%.
The heavy energy bias in the index is considered by many as a drawback of the GSCI as movements in grains, livestock or metals do not influence it as much as they influence other commodities indices. Arguably for PGGM this is not a really important disadvantage. The GSCI is calculated according to a formula where the higher the value of the world production of a certain commodity, the higher its weight. As soon as the value of the world production in energy commodities reduces, its weight in the GSCI will automatically decrease. One can argue that the weights in the GSCI represent the economic importance of the commodities. The higher the economic importance, the more pronounced its interaction with other economic assets like equity and bonds. “The correlations that arise from these interactions are what make commodities interesting as a strategic investment, says Beenen. “These correlations are strengthened by the higher weightings. What is seen as a shortcoming by other investors, we actually like.”
While the GSCI is less diversified than other indices, its volatility is higher. But it is this high volatility that makes it work as a diversifier in a portfolio with other assets, while the fund’s rebalancing takes place at the level of asset classes.
For a number of reasons, the decision was to have the commodity exposure in place before year-end. “This was a considerable challenge, as we had no experience in the commodities markets,” he says. The first thing to decide was what particular form the commodity exposure was to take. There were several possibilities: directly in individual commodity futures, GSCI futures, structured notes, GSCI total return swaps, or to use asset managers. Keeping the strategic nature of the commitment and PGGM’s novice status in mind, it was decided to try and keep close to the chosen benchmark while building the commodity exposure. The instrument best suited to this turned out to be the GSCI total return swap. In this swap , the fund receives the total return of the index in exchange for fees and an interest rate payment (usually three-month T-bill). These fees have to cover the costs and risks associated with the rolling and holding the futures.
Next problem was how to inject e2bn in a relatively small market without disturbing it. The GSCI-linked market before June last year, when PGGM started investing, was approximately $6bn. The market of the underlying commodity futures is substantially bigger, more than 15 times for energy-related futures. “The problem is, of course, that there is one-sidedness in the market, as only investors want the GSCI, whereas GSCI-sellers are in the underlying commodities,” Beenen says. But it turned out that sizes up to $100m–200m in one particular day posed no problem in normal circumstances. “Some market participants may have been waiting for our entrance, which is another reason why we staggered the implementation and we did not give any information about our schedule.”
During the pre-implementation phase as well as the early implementation, PGGM says it obtained valuable advice from Goldman Sachs. In later stages other, professional counterparties were involved as well. The availability of a choice of several counterparties is one of the attractive features of the GSCI as a benchmark. These other parties showed their professionalism on the product level by providing alternative strategies and fee-structures.
“We did look at employing asset managers and while they could show outperformance, they had quite a wide tracking error. So there was a risk of deviating from the benchmark to too great an extent, particularly in the initial stages. But it is something we will keep an eye on.” The cost-structure and the stage of development of asset management in commodities at this stage are not considered favourable. The fact of the matter is that most asset managers providing commodities mandates only manage relatively small amounts. “So if a big pension fund enters the market, they could easily double or triple the amounts the managers are handling. Many commodity managers do not have someone full-time on the commodities side.” However, PGGM acknowledges there are a few notable exceptions and it is exploring this route as a possibility for the future.
By early December, just five months since its first commodity investment, the fund reached its desired allocation to commodities, all in GCSI total return swaps.
It is too early to tell at this point whether or not such quick implementation was the best choice. “But without doubt, it has worked well until now,” he says. “By moving quickly we were able to enjoy part of the 50% rally of the GSCI TR last year, and we captured good double digit returns.”
The correlation structure predicted by the ALM study for the long term manifested itself already in the first couple of months, says PGGM. In a year where it was faced with lower equity returns, commodities delivered the fund already an attractive return during the five-month build-up phase.
It illustrates that commodities tend to do exceptionally well in periods of global overheating. Over the long term equity-like returns of around 12% can be expected. Market perception is that the return will be considerably lower in the coming year.
Already, PGGM is assessing what more it can do. It has identified ‘backwardation’, the effect where nearby futures trade at a higher level than futures with a later expiration, as an important source of return on its commodity investment. It says that the GSCI TR is partly designed to capture the possible extra return from backwardation during the rolling of futures. However, the formulaic way the rolling of the GSCI TR is defined (one of the requirements of a benchmark) is not necessarily the most effective way to profit from the backwardation at any given moment. This is just one example where a more active management style than just GSCI total return swaps can result in higher commodities returns, without endangering the strategic nature of the commodities investment. Being fully invested in commodities means that PGGM will now be able to investigate, and implement, those possibilities for more active management. For the implementation of these opportunities the total return swap alone will probably not suffice. Futures, be it on the GSCI or on the individual commodities, will be necessary. External asset managers will be able to be of help here as well. PGGM expects asset managers will develop their own clearly identifiable styles. This will help to identify sources of benchmark outperformance (alphas) and the alpha-correlation structure. An appropriately chosen combination of these sources might enable the fund to employ sources of outperformance with relative higher tracking errors, as the total tracking error of this combination becomes smaller than those of the individual sources.
Futures as well as the combination of asset managers are considered for the coming year.
If there is an increase in interest by institutions it could encourage more companies and producers sell their risk in the futures market. “The institutional investors can play a part here, because we are prepared to take these risks on long term. Even if there is a period of negative returns, these can be offset with the returns of other assets,” he says. “With more big investors, companies and commodity suppliers will see more possibilities to hedge their risks in the futures markets. This could lead to bigger and more liquid markets.” But he does see that smaller pension funds might have their own particularly problems to become active in the commodity arena. “For a start it is hard to find people with experience in commodities investments. PGGM is a big fund and can afford to employ people specifically to manage the area.”