Little Update on Commodity Re-Thunk
Reading some of the comments people have posted in various places, I thought it would make sense to spend the time to re-create the projected-index chart in the prior article, but using something approximating GSCI weights. The GSCI is production-weighted, which means it is very heavily energy-linked (I used 2008 weights just because they were the first ones I found: 78% energy, 10% agriculture, 6% industrial metals, 3% meats, 3% precious metals), yet I was comparing its 10-year real return to an equal-weighted “prediction.” How much does this change the picture?
As it happens, quite a bit. Here is the new chart (sourced: Enduring Investments).
Categories: Commodities, Quick One
GSCI, precious metals
Leave a Reply
that one makes a nice fit with observable data, but it also has a pretty crazy outlier in, I would think. The GSCI hit a pretty crazy high 875 in 2008, which would be, let’s call it, $1000 in 2014 dollars. But -20% annualized for 10 years is brutal! It suggests that the GSCI should be $100 in 2014 $s by 2018. That’s crazy enough, but alongside the claim that its fair valued at the present price of 388 its just impossible. Even the 2011 projection of -13% is pretty hard to swallow, since 13% annualized for 10 years is a 75% decline.
I’m not unpersuaded that there is some sound theory to this (mostly based on that one gold chart,) but there is frustratingly little useful data against which to get the correct reading for present day projected value. At the very least, it seems to exaggerate the impact of overvaluation.
Gosh, yes. I can’t imagine that one can take that prediction at face value. These are all fit logarithmic curves, after all, and that projection was made mainly on the back of the oil projection (which projection was WAY beyond the range of the data the curve was fit on. But let’s suppose there was a miss similar to the 5% miss at the left-hand side of the chart. Then oil, from a high real price of about 151, would (losing 15% compounded per year) end up at $30 (real). That’s not terribly far-fetched given that it recently hit $45! So maybe it’s not as crazy as it seems.
By the way, who said commodities were fair at 388 on the GSCI? the chart above shows it’s lower than the prediction would have had them. I see commodities as significantly cheap. Just maybe not as cheap as I had thought them before. (Actually, let me revise that because I don’t know what “fair” necessarily means here…I guess a modestly negative real return projection, say -3% or so, in line with productivity gains in production, would be fair? But I don’t know what that number is. What my claim SHOULD be is that the expected real return of this asset class is quite generous compared with other real asset classes, once you add all the sources of return. And I’m comfortable with that.
p.s. thanks for these thoughts. I love the give and take!