To expand on this, the next two charts split the net long into longs and shorts. Here the new report shows a consistently higher level than the old report, in both longs and shorts - presumably this is because entities holding long and short positions do not need to use the futures market to offset them. Finally, the fourth chart below shows that the similarity between the reports in net long applies to most of the 12 individual agricommodities, although there are some differences, with sugar showing the largest.

The second source of information was based on what we know about the weightings of the two leading commodity indices, the S&P GSCI and the Dow- Jones UBS. We have this data for each commodity at the end of Q2 2009 (S&P GSCI data is taken from them directly; we have calculated it for the Dow Jones- UBS). We also know Dow-Jones UBS's estimate of the AUM of funds linked to their index, which is $36bn. While we don't know what the S&P GSCI AUM was on that date; at the end of August 2009 it was $60bn, and we will assume it was a little less, $58bn, at the end of Q2 2009. This makes $94bn in those two indices at the end of Q2 2009, which compares to a CFTC estimate of all indices of $118bn. This implies other indices had $24bn as of that date, or that we may have underestimated the main two indices.

The table below shows - for each commodity in the S&P GSCI or Dow-Jones UBS - its share of each index in percentage terms, and the amount in $bn if the AUM of the indices are $58bn and $36bn, and finally the combined amount. It then shows the CFTC's estimate as of end Q2 2009 for all commodity indices and then the CFTC's estimates scaled down so the total matches our $94bn, and finally gives the difference of our estimate, less the scaled down CFTC's.

The first thing to note, and as the chart shows, is that the estimates are quite similar, and support this method of calculation. But there are some obvious differences. The estimate of crude oil made by our method is too high - $29.2bn compared to $24.6bn. The natural gas estimate is far too low, $5.0bn compared to $7.7bn, as is sugar, at $2.6bn compared to $4.7bn. In metals, the gold estimate is too low, $4.3bn compared with $5.4bn, while silver is slightly too low, $1.3bn to $1.5bn, and copper (Comex) is too high, $3.5bn compared with $2.9bn.

These discrepancies could be for many reasons. For example, in crude oil there might be a greater preponderance of index shorts than one would expect from simply looking at the overall weightings, while in natural gas there is a large single-commodity fund, which would mean more investment than suggested by the weightings. This might also explain gold and silver, which have a large investment following. Furthermore, the other commodity indices tend to have different weightings than the S&P GSCI and Dow-Jones UBS, in particular often a lower oil weighting and higher gold weighting. Therefore it is no surprise that the overall CFTC figures show less oil and more gold than simply looking at the two major indices.

Other metals are not covered by the CFTC report, but our estimate for non-US index investment, $19.2bn, is close to the CFTC's $18.2bn, and therefore we assume that the individual figures given will reasonably approximate to the actual investment (in the two main indices) in those LME base metals - so $3.6bn in aluminum (equivalent to approximately 2.3 Mt), $1.7bn in copper (330,000t but there is also another $3.6bn in Comex copper), $1.6bn in nickel (95,000t) and $1.5bn in zinc (950,000t) and $0.3bn in lead (120,000t). This is just the S&P GSCI and Dow-Jones UBS - if we take into account the other commodity indices then we can probably add 15-20% to these amounts.

Useful - but inconclusive

Our conclusion therefore is that the new report is a useful addition to our understanding of index funds and their impact on commodity markets. It gives a more detailed and accurate picture of trends in these markets, one that is more comprehensive than the current CiT report, at the expense of timeliness and frequency.

But it does not radically change our understanding of these markets; indeed it supports our earlier findings, as expressed in our reports in the past few years across various commodities. It will probably fail to give sufficient ammunition to those who want to blame commodity indices for high or volatile price's not that we count ourselves as being in that camp. There is a certain irony, however, in that, as the CFTC provides more information on these markets, the extra regulation of futures markets that is part-and-parcel of the same political pressures behind the report is threatening to drive investors from US markets and the oversight of the CFTC. In September, for example, there were reports that S&P GSCI were looking into the possibility of commodity indices that used no US futures contracts. Such moves would completely undercut US political efforts to cast ever-brighter lights over the whole business of commodity investment.

Excerpts from Fortis Metals Monthly
Courtesy: Fortis/VM Group

Post your comment  (0)

Connect:

Post to Twitter

Post to Facebook