May 17, 2008

posted by William Birdthistle

Since overestimating the value of oil no longer seems possible ... the chances of making a profitable investment in the stuff appear to be drying up.

Isn't it the other way around? I.e., with rising prices, isn't it easier to make a profitable investment in oil? For example, expensive drilling operations all of a sudden become profitable.

Best Monty Python scene ever.

Bruce,

I meant that, although Exxon, et al., may have an easier time making an operating profit with high oil prices (as, indeed, they have been doing over the past few quarters), investors hoping to select an investment whose value is premised on the rising cost of oil will have a harder time once the market prices in Goldman's estimate of extraordinarily high price of oil. Indeed, despite Exxon's having posted record-breaking profits in recent quarters, it's stock price has dropped in 2008.

- William.

Orin,

I've always thought that peasant scene could be useful viewing in con law.

But "best ever" is a big title -- perhaps the RAF banter one is also in the running:

Maybe that's just my own penchant for WWII stuff.

- William.

Does it strike anyone else that there is something twisted about water as an investment vehicle? Shouldn't this notion give us pause, so that we can consider the distorting forces of the financial economy on human lives? We already see something analogous going on with food, in that hedge fund investments in food commodities are contributing to food shortages that are creating suffering around the globe.

Water and food are absolute necessities for human life. From this perspective, shouldn't access to them be part of human rights? If even thoughtful and decent people like the contributors to this blog can ask "So, how does one invest in the stuff?," shouldn't we be thinking long and hard about how our system of economics is flattening our human values, and about what can be done to change that?

BTW, the WSJ's statement "It takes 3,000 gallons to produce a single semiconductor before it’s placed into a computer" is incorrect. "Semiconductors" -- what WSJ means is microchips -- aren't manufactured individually, they're made on wafers with diameters typically of 200mm or 300mm. (Some older fabs, especially in the developing world, still use 150mm wafers). The number of usable chips (a/k/a "dice") that are cut from this wafer after processing varies according to the application; the dice that go into Pentiums are larger than those for memory chips, which may be larger than some of those used in chipsets for wireless communications, etc. (Since dice are rectangular and wafers round, the number that can fit onto a wafer isn't a linear function of surface area.) Also, both the size and the actual yield of usable chips will vary according to the processes used in fabrication.

As of around 2001 or so, it was estimated that Intel would get roughly 440 Pentium chips from a 300mm wafer. Also around that time, it was estimated that you needed about 1,500 gallons of water to make 1,000 gallons of ultrapure water, and about 3,000 gallons of UPW to process one 300mm wafer (see http://www.p2pays.org/ref%5C04/03271/). That works out to a bit more than 10 gallons per Pentium die, and considerably less for smaller chips. (However, a small amount more might be necessary when one includes the packaging and other back-end processing.) There have been a number of recent trends in semiconductor processing that might affect the numbers slightly, but WSJ's estimate is certainly wrong by more than 2 orders of magnitude.

Ah, I get it, thanks William.