OK, let's amortize the costs over, say, five years--a typical time between model updates. We have a development cost of $1.2 billion, sales of about 15,000 annually (26,000 have been sold so far), and GM is currently paying about 6.5% on long term bonds needed to pay for this. Right? Right. Each year, to pay off R&D costs, GM's first $300 million in revenues from the Volt must be devoted to this, or we will say it is a subsidy from....let's be fair, the taxpayer.
OK, year one, about 12,000 sold, so the R&D cost per vehicle would be about $25,000. This year, maybe 20,000 will be sold, so the R&D cost per vehicle is about $15,000. Overall, the company would need to recover $600 million in R&D and interest costs, so for about 32,000 vehicles, the R&D loss would be around $18,000 per vehicle. Increase sales to the suggested 50,000 units per year, and the R&D cost lost is about $6000 per vehicle--count me skeptical on this one.
Next up is the BOM loss, and I don't think we'll be getting any precise information on this. Suffice it to say that the estimated cost to replace the batteries with a Prius is getting into four figures, and then you've got a motor for a few thousand dollars as well. I would have to guess that the BOM and assembly cost for the "Cobalt priced like a Bimmer" is about, or is somewhat higher than, the $40,000 sticker price.
Which is a long way of noting that my best guess about the loss