When it comes to risk parity strategies, all things are not equal. While bonds and equities conform nicely to the basic risk parity premise that assets should be equally weighted according to their risk contribution, commodities are a wild card.
They have had historically low correlations with equities and bonds and as a result can come out well in any optimisation approach based on diversification. But as Eddie Qian, CIO at Panagora Asset Management, points out, commodities are not a conventional asset class in that they do not produce income in the form of dividends or interest.
Investors holding physical commodities can only expect returns in the form of capital appreciation and so, while commodities may have low correlations to bonds and equities and can provide a useful hedge against inflation, their expected returns rely on the economic forces of supply and demand.
A slowdown in the Chinese economy has led to significantly reduced commodity prices in the past few years and investors have suffered as a result. Moreover, Qian argues that the impact of more central banks adopting inflation targeting across the world means that investors have limited access to the upside of inflation shocks. Qian therefore says that multi-asset risk parity strategies should not apply a simplistic equal-risk-weighted approach to commodities.
Alexandre Deruaz, head of systematic equities and alternatives at Lombard Odier Investment Managers, agrees. His firm’s multi-asset strategy has a 20% risk exposure to commodities and 40% each for equities and bonds. Deruaz says this is a pragmatic response to the view that commodities are useful to hedge inflation risks but offer little long-term value.
Panagora also does not stick to a rigid allocation. Commodities form part of a strategic risk allocation to inflation hedging assets that is lower than stocks and bonds because their risk-adjusted returns are lower, according to Qian. “Our approach takes it one step further and we can introduce modest changes in our commodity exposure as the risk-adjusted returns between asset classes change in the short term.”
Commodity investing raises several practical questions. Institutional investors are invariably focused on gaining exposure through derivatives, as few could envisage holding pork bellies or gold bars. Holding futures contracts introduces another source of volatility through the requirement to roll over futures contracts on expiry. This introduces the issue of roll yield.
At a glance
• Commodities are not a conventional asset class, so adopting a pure equal-risk weighting for them in a multi-asset portfolio makes no sense.
• Investing via rolling over futures contracts introduces practical challenges arising from contangos and backwardations.
• Traditional commodity indices are flawed and can overweight energy.
• Adopting risk parity methodology to weight individual commodities within the commodity portfolio is attractive.
As Qian points out, a roll yield is not a yield in the conventional sense. It exists because the price of a futures contract due to expire in, say, three months’ time can be significantly different from the current spot price, although it has to converge to the spot price on the expiry date. If a commodity had no storage costs and could be easily borrowed with no transaction costs, the futures price would always be equal to the current spot price plus interest. A purchase of a future is economically equivalent to placing the cash value of the spot-priced contract on deposit for the period of the contract. In this case, rolling over futures contracts would incur a loss each time, which would be offset by the notional interest earned on the cash value.
In reality, the futures price can swing wildly relative to the spot price and can be lower, giving rise to so-called backwardation, as opposed to the normal contango position where the commodity futures price is above the expected future spot price. Backwardation happens because commodities generally cannot be borrowed and require storage costs. Futures prices also reflect the supply and demand in the market, with producers generally hedging risks by selling futures contracts to risk-taking investors. Rolling futures in conditions of backwardation means investors generate a positive roll yield which, proponents of commodity futures have argued, transfers risk from producers to investors.
Investors also have to grapple with the problem of how best to construct a diversified portfolio of commodities. It is widely accepted that the S&P GSCI and the Bloomberg Commodity index (formerly known as the DJ-UBS Commodity index) give the best exposure. But index construction is underpinned by the fact that the weights are proportional to the global production of the commodities, similar to equity indices weighted by market capitalisations. But as Deruaz points out, the cap-weighted approach will invariably give a disproportionately large exposure to the energy sector, which makes up about 70% of the S&P GSCI.
Instead of an equal-weighted approach, institutional investors running pure commodity strategies can gain more attractive results by adopting a risk parity framework, according to Deruaz. He explains that Lombard Odier divides commodities into three groups – agriculture, energy and metals – giving each equal risk weight in the firm’s commodity strategy. Individual commodities in each sector are also equally risk weighted.
As a result, diversification is improved and the exposure to energy risk reduced. Deruaz adds: “Each commodity is characterised by a specific risk. This risk is evolving with prevailing market conditions and capital allocation is sized dynamically to each commodity, such as to achieve the targeted risk contribution.”
Lombard Odier’s risk parity index with monthly rebalancing limits the exposure to boom-bust scenarios and the main objective is to provide investors with a broad set of commodities that conform to UCITS IV rules, according to Deruaz. The index includes 21 commodity futures, which have been selected based on their liquidity, representativeness, risk characteristics and tradability. It excludes several commodities including electricity (too volatile), soybean oil (tradability) and orange juice (poor liquidity).
Lombard Odier has a flexible monthly futures-contract rolling strategy which includes screening for maturities and retaining those with the optimal yield. This means the firm could roll out contracts onto further-out, less-liquid futures contracts to minimise the cost associated with negative roll yields in contango and maximise the cost linked to positive roll yield in backwardation.
The use of further-out contracts could introduce liquidity problems, however, and Panagora prefers to stick with nearby contracts, although it will occasionally use the second nearest, according to Qian.
Risk parity applied to commodities does makes sense if an investor wants to have commodity exposure. The question, however, is whether commodity investment is viable at all.