Bubbles are mass psychology phenomena and this one is no different. They will not hit an inflection point where cap-ex scales down. That's the nature of the fast moving auto business.
If you look at auto manufacturers and other capex intensive industrial companies you will see that enterprise value/invested capital is generally below one because returns on capital are so poor. Tesla's will also revert to sub-1x (e.g. 90% downside) as their business model is so flawed.
I'll bite... "They will not hit an inflection point where cap-ex scales down". True, but cap-ex as a percentage of revenue will certainly start to drop as some point because revenues will start to go way off the charts.
Just putting it simply, Revenues were 4.05B for 2015, CapEx was 1.63B for the same year. Assuming Tesla continues to grow revenues by 50% each year, I would expect about 6B in revenues this year, 9.1B in 2017, 13.7B in 2018, 20.5B in 2019, and 30.8B in 2020. If you honestly think that in 2020 they are going to be spending an equivalent 50% Y/Y growth which would be required to hold the same high percentage they currently sit at, then you are seeing them spending 12.3B in 2020, no? I mean, I would be extremely happy to see them still expanding that fast in 2020, but it seems a bit extreme...
To put that in perspective, 12.3B is enough CapEx for 2 full Gigafactories to be built in 2020 alone. Or more realistically, if it takes 5 years to build a Gigafactory, and let's just assume they spend the same amount each year on capex (1B a year for 5 years). That would be enough capital to be building 6 gigafactories at the same time with another 6B left over for whatever else they are doing. If you actually think Tesla is going to continue to be this aggressive with their spending, then it would be in preparation not just for "a couple million by 2025" as Elon stated, but to be essentially planning to take over as the top manufacturer of cars over Toyota, GM or VW!
So either you are expecting some serious growth that no bull is even expecting (which is greater growth than anyone is anticipating, and that type of spending in 2020 would theoretically grow revenues by much more than 50% at that point), or, more likely, their CapEx spending as a
percentage of revenues would drop off in the coming years.
So, now that we have that established. GM as a percentage of revenue is expected to hold firm in the 20s, because even with the Model 3 they are expecting some pretty aggressive GMs which is likely to translate into even more aggressive GMs on the high end (major costs to be cut in the core cost components of the car like the motors and the battery pack will translate into savings across all platforms). This means that as CapEx falls and GM holds steady, you will get cash flows that will end up with, as Elon put it, more money than they know how to spend it fast enough.
At that point (previously stated by Elon as "by 2020") they will have no choice but to show GAAP profits and crazy positive cashflows because they just can't spend money fast enough. And that is also the basis for the AAPL comparison as AAPL essentially has the same issue. They have held super high GMs for so long that they are making money hand over fist that they just can't spend fast enough.