How to clone commodity indices for the better

Posted on 01. Apr, 2014 by Jean Jacques Ohana in Weekly Focus | Comments Off

Commodities’ performance has been lackluster since 2006 (Figure 1). On table 1, we note that, even though the spot performance was very good, it was eaten up by the negative roll return. This is due to the steep contango of commodities’ forward curves from 2006 to 2012.

However, Figure 2 shows a revival of the commodities’ theme since the end of 2013, as commodities curves have exhibited a clear trend towards backwardation (defined by spot prices being more elevated than one-year-out forward prices). A series of climatic and geopolitical problems have indeed disrupted the supply of a significant number of commodities (impact of El Nino on soft commodities, potential consequences of the Ukrainian crisis on the production/transportation of wheat, natural gas, nickel…). In spite of this positive signal for the commodities indices’ roll return (Figure 2), the spot return performance remains depressed by the chronic lack of interest of index investors in this asset class (Figures 3 and 4).

Due to their low upside potential and the looming threat of regulatory backlash, commodity indices carry few promising long-term prospects for the investor. Therefore, we have developed a framework to replicate the performance of commodity indices by using a basket of commodities producers’ stocks and currencies (Table 2). The clone has a 60% correlation to commodities indices and its Sharpe ratio is markedly higher (Tables 4 and 5). Indeed, stocks and high-yield currencies do not have rolling costs and offer instead long-term risk premiums (dividends or carry returns). This commodities’ clone does not come without costs, though: its beta to equities is higher than the one of commodities (Table 5) and it correlates a bit less to CPI changes than commodities indices (Table 6), hence offering a less efficient hedge against inflation.

We may enhance the risk-adjusted performance of the clone by divesting away from the basket when its trend gets negative (see Figure 7). This increases the Sharpe Ratio by 20% and reduces the drawdown by half (Table 7).

Figure 1: Total returns of two main commodity indices since 2006

Table 1: Statistics (March 2000 – March 2014)

Figure 2: average one-year curve of commodities across a basket of 20 commodities (the curve is defined as the ratio of thirteen-months-out to prompt-month forward price -1) compared with roll return of DJ UBS over the next year (green, left scale)

Figure 3: Roll return (YoY) and Spot return (YoY) for the DJ UBS Index

Figure 4: Cumulative flows to commodity indices in billion dollars (initialized to 0 in 2006). Source: CFTC

Table 2: Basket of commodities producers’ stocks and currencies composing the clone

Table 3: Components of the basket ranked by decreasing correlation to the DJ UBS

Table 4: Risk and performance statistics (March 2000 – March 2014)

Figure 5: Ratio Commodities Producers & Commodity Currencies / DJ UBS Total Return

Figure 6: Total return performance of commodities and commodities’ clone

Table 5: Correlations with commodities indices

Table 6: Protection against inflation rise: correlation between year on year US CPI variation and year on year assets classes evolution

Table 7: Passive vs Trend-driven allocation to the basket of commodities producers’ equities and currencies (2000-2014)

Figure 8: Average Riskelia’s trend of commodity related markets

Tags: , , , , , , , , , , , , , , , ,

Comments are closed.