Commodities, Infrastruture data

Hello all, I’m trying to gather some long-term data on commodities and infrastructure. I have the following monthly total returns data: 1991 – 2007: Dow Jones-AIG Commodity Index 1994 – 2007: Macquarie Global Infrastructure 100 Index 2002 – 2007: S&P Global Infrastructure Index I’m trying to backfill/append these data series back to 1973 (or close). A utility index could represent pre-1994 infrastructure as these asset classes are highly correlated. **So, does anyone know of a utility index or an index that could be used in lieu of a “real” utility index that goes back to 1973? I need the same for a broad based commodity index that could closely represent the GSCI or DJ-AIG above.** I need monthly total returns. If you can point me to a good source, that’d be great. If you can actually get me the data, you’ll be my hero. Thanks in advance. -XSellSide

Dow Jones Utilities Index is one of the original ones started in 1930’s (I think). GSCI has a correlation of 0.85 or better with crude oil (so much for broad-based) and crude oil data is really easy to get.

I thought the DJUI went way back, too, but I can’t track down the monthly returns data. I don’t really want to use the GSCI if I don’t have to for that very reason (oil). The DJ-AIG is more balanced. Last resort, as you stated, is simply using oil. What about the Monthly CRB Futures Index for “all commodities”? I’d have to make an adjustment for the return on the collateral, correct (t-bills)? Would I also have to figure in the roll return somehow?

It would be nice to be able to fix up the CRB index for rolls but without an absolute ton of work you can’t. The problem is that the CRB tries to keep its original price constant so they average futures prices across time. That means that they take, say, 5 contracts and average up their current pruces and get the commodity price. When one month goes into delivery, they drop it and add the next forward month. Because it’s averaged you don’t see that roll effect too much. It’s a silly solution because a) The index includes all kinds of odd convenience yields and holding costs. b) The index equally weights all months even if the most liquid month was an open interest of 1M contracts and the most forward month has an open interest of 2K contracts. c) Those roll “returns” don’t really happen on one day and it might look like you get 9 months of convenience yield in one day (probably divided by 3 or 4). Anyway, you can’t adjust for the roll. Adding in T-bill is really easy and if you don’t have T-bill use just about any short-rate that you have. Sufficiently pressed, I would use returns from some money market mutual fund. It’s not exactly clear that the right rate to use is T-Bill anyway (and there are roll issues in T-bills as well).

Really interesting color on the CRB, thanks.

Quick and dirty: I took the CRB Futures Price Index monthly returns and simply added 30 day t-bill monthly returns. Comparing this to the Dow Jones-AIG Commodity Total Return Index gives me the following total returns from 4/1991 to 12/2007: “Adjusted CRB” = +416% DJ-AIG = +374% The DJ-AIG is slightly more volatile, but if you squint a little at the chart, you can barely tell them apart. The is a 0.62 R^2 between the monthly returns of the two indices. Whatcha think, JDV?

Of course, there are lots of differences between the two indices so you shouldn’t expect them to be the same. In particular, DJ-AIG is partially weighted toward production numbers (but not as much as GSCI) and the contracts are weighted according to liquidity (somehow, I forget and too lazy to look it up). I think that your measure is as good as any other and better than some for something like a general price level of commodities on futures markets. I just never know what to do with such things.

I’m actually putting together portfolios and back testing them. Since neither of these asset classes represent more than 2-3% of the portfolio each, small differences in performance won’t matter very much.

Type your email adress Halifax