Which index?

Iain Morse takes a look at the index choices available to investors taking a passive approach to commodities

Whatever your views on the current state of play in the commodities markets, it has been difficult to argue against those advocating exposure to commodities as an asset class. Portfolio theory endorses the inclusion of asset classes that have low or ideally no correlation to each other as a means of reducing overall portfolio volatility relative to liabilities. On the optimistic side, commodities certainly fit that bill. They are potent diversifiers when held with equities and bonds.

Better still, commodity prices just keep rising. And there may be long term cyclical relationships between commodity values with both inflation and with equity values, which make commodities even more exploitable as an asset class. With all this in mind, it is easy to see why so many larger pension funds have sought exposure to commodities.

"This leaves open some very important questions about implementing a programme of commodity investment," cautions Richard Cooper, (pictured right) a principal at Mercer Investment Consulting. First there is the vexed issue of index or benchmark selection. For instance among mainstream equity index providers there is less and less to choose in terms of fundamental design and constituents. After all, the equities listed in the indices are separately exchange traded and independently priced.

By contrast, commodity indices vary widely in terms of their design. There is no exchange traded set of price and volume data on which they can all be commonly based. In fact, there is a profusion of indices, all purporting to do something the others don't. Next comes the vexed question of whether to choose passive index replication or active management and if the latter, the consequent necessity to choose a relevant performance benchmark.

A glance at the two most widely used commodity indices, the Dow Jones AIG (DJ AIG) and Standard & Poor's Goldman Sachs Commodity Index (S&P GSCI), makes the point about the complexity of this subject. The S&P GSCI is based on world
production of 24 liquid commod-ities, with the weighting of each commodity in the index determined by the average quantity of production over the previous five years. Returns are based on the average arithmetic average of the returns on the underlying
future contracts.

By contrast, the DJ AIG Commodity Index uses 19 commodities with the weightings based on liquidity but also on world production dollar values, both averaged over the previous five years. Most importantly, commodity sector weightings in the index are capped at 33% when the index is re-weighted annually and allocations under 2% are excluded.

The very different designs for each index have large consequences for their composition and relative behaviour. For a start, the S&P GSCI tends to be very heavily weighted to energy; as at 1 January 2008 crude oil alone accounted for 35.3% on the index and all energy for 69.7% of the index. The DJ AIG allocation to crude oil is far lower at 13.2%, with the total for energy capped at 33%.

There are other noteworthy variations between the indices. Grains, currently rising in value to historically high values, account for 9.3% of the S&P GSCI and 18% of the DJ AIG. Each index's allocation to precious metals, gold, silver and platinum, also varies widely; the S&P GSCI allocates a mere 2.3%, DJ AIG no less than 10.1%.

"The relative performance characteristics of these indices varies significantly," adds Cooper. Comparative real data from 1 January 1991 and the inception of the DJ AIG index shows that the DJ AIG has achieved an annualised return of 8.4%, while the total return of the S&P GSCI for the same period was 7.3% a fairly significant difference of 1.1 percentage points.

Volatility, measured as an annual series of past returns, varied even more. The DJ AIG index volatility was 12.6% but the S&P GSCI no less than 18.8%. Even more significantly, the risk/return ratio for each index was respectively 0.7 and 0.4 against 0.7-0.8 for equities and 1.1 for global bonds.

Other indices extend the duration of the future contracts or allocate equally to all commodity classes. In other words, choice of index is a form of active risk even if the index in question is passively replicated.

"This is one reason why we have moved to a position where we think active management is preferable to passive management," argues Cooper. One major consequence of this lies in choice of performance benchmark. For any active mandate this is typically set on an absolute return/cash plus basis rather than index relative basis. Mandates on LIBOR/cash plus 3-5% are now fairly commonplace for actively traded commodity mandates. When this happens, choice of index becomes a secondary issue and none may be identified when a mandate is set up.

Most money allocated to commodities remains passively invested relative to an index. There are three broad routes to gain this exposure. The first is for the investor to ask a bank with the relevant capacity to run a programme of rolling futures to the required exposure. This approach is best seen as analogous to running a currency overlay. The bank will credit-rate the investor and give them a risk rating. The investor pays or receives cash flows depending on whether the futures are in or out of the money.

This, of course, does not require much up-front capital, which is why the bank credit-rates the investor. The exact relationship between the bank and investor will vary, as can the means by which exposure to commodities is implemented, but the basic relationship is always generically very similar. The alternatives are far more capital intensive.

The investor can invest into a collective unitised fund, which may or may not leverage, but where redemptions are via the fund provider. As an alternative, the investor can buy an exchange traded fund (ETF) or exchange traded commodity (ETC). The difference between these approaches is crucial; ETF/ETC pricing is made via the open market.

Some pension fund managers, such as PGGM in the Netherlands, or Hermes in the UK, have the capacity to run future programmes in-house or do so via their own asset management divisions. But even if this facility is available, the governance standards to be met by fiduciaries and boards are becoming more, not less, demanding in all areas of portfolio management. The number of funds with any real expertise in the direct use of futures to gain commodity exposure is small and concentrated among the largest funds. This makes it far easier for pension funds to use one or more of the unitised commodity funds or ETFs to gain the relevant exposure.

The commodity funds are usually constructed as private partnerships that invest in commodity and financial contracts. Access to them is limited usually to ‘expert investors'. Normally an investor redeeming their units or shares in a fund will receive a cash value equal to net asset value on a pro-rata basis.

From a governance point of view this can be an opaque pricing method compared to the pricing of exchange traded securities. Charges on these funds range from 0.25% to 0.5% plus internal transaction costs. Hard data on the amounts invested in these funds is scanty, and there are other related means of gaining exposure to commodities, via multi-strategy hedge funds and TAA fund products. By contrast, price data on ETFs and ETCs is readily available.

Partly as a result of this transparent pricing structure, ETF/ETC growth has been rapid in recent years despite the relatively high charges and costs. The total expense ratios (TERs) on some of the largest US commodity ETFs range from 0.4% to no less than 0.75%. Examples include the streetTRACKS Gold Trust from State Street with a TER of 0.4% and AUM in excess of US$17bn (€11bn) via Victoria Bay Asset Management's United States Oil Fund ETF with a TER of 0.5% with AUM in excess of $1bn, to BGI's IShares GSCI Commodity Indexed Trust which charges 0.75% with AUM of $130m.

European commodity ETFs and ETCs have surprisingly low TER's compared to the US ones. Most have TERs under 0.5 % but exchange listing can hinder access. Examples here include the AXA/BNP Paribas EasyETF range which are listed in Germany and Switzerland. Some of these have TERs as low as 0.35% and the Merrill Lynch Commodity Index Extra fund has a TER of only 0.10%, cheaper than many comparable commodity funds.

Access to ETCs is far wider as these tend to be cross listed. ETF Securities dominates European ETC issuance and list most of its funds in Germany, France , the UK, Italy and the Netherlands. One again, TERs for these funds are typically under 0.5%. As a result they are more expensive than the commodity funds, but their advocates argue that this additional expense is offset by liquidity and lack of minimum investment limits.

As European pension funds develop a greater propensity to invest in commodities, their advisers, the investment consultants, are turning from advocating passive to more active exposure. Meanwhile, index investing is becoming cheaper and easier.

As long as commodity prices continue to rise, current trends will continue. A commodity bear market will change minds both on whether exposure is desirable and how best to obtain it.