I agree with you on AP and 70kWh not being included for $35,000. I think we only disagree on battery cost and its impact on Tesla's margin.
Simply stated, I think Tesla will sell the M3 at $35,000 with a reasonable gross automotive margin of 20-30%. Will they do that initially? Likely not, but it won't take long and will likely be in 2017 unless GF can't produce at scale yet.
In terms of content and features, I think I'm on the same page as you, largely because I don't think Tesla will deviate from its prior strategy with Model S/X.
My own guess for M3 is 55 kWh with the same features as a base Model S. That's higher content than a base 320i and more than enough to be attractive. AP will cost extra. Battery upgrade will cost extra. Usuals (leather, sunroof, sound, cold climate, etc) will cost extra. I would not be altogether surprised, though, if Tesla priced its base model slightly lower than $35,000.
Onto batteries...
I'm possibly being overly optimistic in my $150/kWh guess -- my guess is admittedly partly based on some math, some voodoo, a look at competitors and even an inference based on the way Elon and JB always carefully react on quarterly calls when analysts bring up competitors' cost/kWh.
Turning to Tesla Energy comparisons. If you assume that "other stuff" pack costs of Powerpack and Model S are the same and there's a 10% margin on Powerpack, then we know that current cost is ~$225/kWh including "other stuff". Of course, "other stuff" is a fixed cost while additional kWhs are mostly marginal so size of the pack matters when determining average $/kWh. But, it seems like Powerpack "other stuff" should be more expensive than automotive "other stuff."
Powerwalls are an unfair comparison since they are too significantly smaller and the "other stuff" costs make up such a big difference. Moreover, I think you used the dealer price for Powerwall, not the Tesla price -- I may have missed a bit of news, but I think Tesla only gets $3500 for a 10kWh Powerwall -- so in your math (20% transportation and margin), it's $225 kWh (pretty similar to Powerpack).
As an aside, the pricing differential among Powerwalls also gives a hint as to Tesla's true marginal kWh cost. It's only $500 to buy 3 kWh -- almost all other aspects of the 2 Powerwalls are identical. That's only $166/kWh before gross margin, or $150 kWh net of a 10% margin. Battery chemistry may explain some of the difference, but it can't explain that much since labor and manufacturing costs have to be the largest components of COGS -- that's why Tesla can assert that GF manufacturing efficiencies can lower costs by 30% even without cell chemistry changes. (This marginal cost view, though, tells a completely different story when you look at pricing differentials between 70D and 90D -- that marginal cost seems ridiculously high at $487/kWh assuming a constant 25% margin).
Finally, GM thinks they have a pack cost of $145/kWh this year. I can't believe LG/GM can make cells much cheaper, if at all, than Panasonic/Tesla.
All in all, I see a lot of hints that we are currently sitting in the $150/kWh range before feeling the impact of any chemistry changes or manufacturing efficiencies from a fully operational GF -- admittedly though, I can't prove it.
However, even if we are at $150/kWh for M3 (which would be more expensive than GM), then I think Tesla can sell a $35,000 50kWh car with a ~20% margin with higher content than a similarly-priced 320i -- even if the higher content is simply the center screen, etc.
If Tesla can get to $100/kWh, then look out....
Simply stated, I think Tesla will sell the M3 at $35,000 with a reasonable gross automotive margin of 20-30%. Will they do that initially? Likely not, but it won't take long and will likely be in 2017 unless GF can't produce at scale yet.
In terms of content and features, I think I'm on the same page as you, largely because I don't think Tesla will deviate from its prior strategy with Model S/X.
My own guess for M3 is 55 kWh with the same features as a base Model S. That's higher content than a base 320i and more than enough to be attractive. AP will cost extra. Battery upgrade will cost extra. Usuals (leather, sunroof, sound, cold climate, etc) will cost extra. I would not be altogether surprised, though, if Tesla priced its base model slightly lower than $35,000.
Onto batteries...
I'm possibly being overly optimistic in my $150/kWh guess -- my guess is admittedly partly based on some math, some voodoo, a look at competitors and even an inference based on the way Elon and JB always carefully react on quarterly calls when analysts bring up competitors' cost/kWh.
Turning to Tesla Energy comparisons. If you assume that "other stuff" pack costs of Powerpack and Model S are the same and there's a 10% margin on Powerpack, then we know that current cost is ~$225/kWh including "other stuff". Of course, "other stuff" is a fixed cost while additional kWhs are mostly marginal so size of the pack matters when determining average $/kWh. But, it seems like Powerpack "other stuff" should be more expensive than automotive "other stuff."
Powerwalls are an unfair comparison since they are too significantly smaller and the "other stuff" costs make up such a big difference. Moreover, I think you used the dealer price for Powerwall, not the Tesla price -- I may have missed a bit of news, but I think Tesla only gets $3500 for a 10kWh Powerwall -- so in your math (20% transportation and margin), it's $225 kWh (pretty similar to Powerpack).
As an aside, the pricing differential among Powerwalls also gives a hint as to Tesla's true marginal kWh cost. It's only $500 to buy 3 kWh -- almost all other aspects of the 2 Powerwalls are identical. That's only $166/kWh before gross margin, or $150 kWh net of a 10% margin. Battery chemistry may explain some of the difference, but it can't explain that much since labor and manufacturing costs have to be the largest components of COGS -- that's why Tesla can assert that GF manufacturing efficiencies can lower costs by 30% even without cell chemistry changes. (This marginal cost view, though, tells a completely different story when you look at pricing differentials between 70D and 90D -- that marginal cost seems ridiculously high at $487/kWh assuming a constant 25% margin).
Finally, GM thinks they have a pack cost of $145/kWh this year. I can't believe LG/GM can make cells much cheaper, if at all, than Panasonic/Tesla.
All in all, I see a lot of hints that we are currently sitting in the $150/kWh range before feeling the impact of any chemistry changes or manufacturing efficiencies from a fully operational GF -- admittedly though, I can't prove it.
However, even if we are at $150/kWh for M3 (which would be more expensive than GM), then I think Tesla can sell a $35,000 50kWh car with a ~20% margin with higher content than a similarly-priced 320i -- even if the higher content is simply the center screen, etc.
If Tesla can get to $100/kWh, then look out....
