I would probably bet my entire TSLA stake that Tesla will NOT sell 4M cars by 2020.
Even 1 million cars is stretching it, IMO.
Ok. I appreciate the frank response.
With which part of my forecast do you not agree? The following are the logical pieces in order:
1. Precedent. By the end of 2017, Model S and X will have achieved more than 30% US market share in their respective niche segments. Model S already achieved this in 4 years and Model X is on track to achieve it in less than 2 years, and that's including the six-month long ramp-up issues, which an optimist may reasonably assume out for Model 3 timeline. This acceleration from 4 years for Model S to 2 years for Model X should be noted, because it screams growing brand value of Tesla, but my argument does not rely on this assumption. In fact, I assume the opposite: a slowdown from less than 2 years for Model X to 3 years for Model 3. This is part of my "margin of safety."
2. Demand. Arguably Model 3 should achieve a
larger market share for three primary reasons: (i) With Model S and X, Tesla and Elon Musk have already proven themselves as reliable innovators building consumer goodwill, which translates to higher brand value, which Model 3 will leverage in 2017/18; but more importantly, (ii) the lower ticket price of Model 3 vs. Model S/X will translate to higher relative global adoption due to the differences in purchasing power between the US and ex-US countries. In other words, if Tesla achieved more than 30% US market share with Models S/X, but the global market share achieved was say 5%, a reasonable assumption would be if Model 3 achieves the same 30% US market share, that should translate to ~10% global market share; and most importantly (iii) I expect Model 3 to be marketed as fully autonomous from the get-go, which would make the phrase "game-changer" a significant underestimation. I don't think even the bulls truly comprehend what this really means, but my perspective is that a fully autonomous car approximates about $15,000 lower effective price ($20/hr avg earnings x 5hr/w saved not driving x 50 working weeks x 3y avg holding period, which are all conservative assumptions individually, which makes the compound calculation stupidly conservative), all else equal,
excluding the cool factor.
Having said all this, I projected only
5% US market share
3% global market share for Model 3 in
three years. For reference, take a deep breath and make sure you're seated, a base-case projection that incorporates everything I discussed in this paragraph would add up to at least 30% US market share and 10% global market share in two years, or ~6 million Model 3's in 2019. An optimistic case, could assume 35-40% US market share and 15% global market share in one year,
and a halo effect to Model S and X, or 9+ million Model 3's in 2018 with incremental demand on Model S and Model X. TLDR: demand for 4m all-electric fully autonomous Model S/X/3/Semi/others in 2020 is there.
3. Supply. When Tesla first announced the Gigafactory in 3Q14, it projected 500,000 cars in 2020. Shorts said impossible. Two years later in 2016, Elon pulled the 500,000 projection to 2018. Shorts said impossible. Elon last year doubled his original Giga1 2020 capacity to 1.0m, then increased it further to 1.5m but said they would use the third half million for Tesla Energy. Nobody cares what shorts say at this point. Then Tesla bought Grohmann solely for the purpose of creating "alien dreadnoughts." As far as I can tell from this timeline, 1.5m capacity forecast for Giga 1 does not include any improvements that can be reasonably expected from Grohmann acquisition. Arguably, the 1.5m capacity projection incorporates limitations that would reasonably come with designing something three years ago in 2014 and being constrained to make only incremental changes as you move through time, as opposed to designing Giga's 3/4/5 with the accumulated knowledge, innovation from Grohmann, easier access to capital, and a more reliable demand forecast. Keep in mind that Elon has repeatedly noted that he expects significantly more improvement in manufacturing of the product than the improvement Tesla has achieved with the product, which has been at least an order of magnitude as can be concluded just by looking at the fast and vast consumer adoption of a completely new product in a slow-moving automative industry with slow replacement rates.
Having said all this, let alone further improvement, my projection of 4m cars in 2020 with four Gigafactories assumes a
deprovement to 1m per Gigafactory. This is part of my margin of safety.
4. Capital. Tesla is no longer a pre-revenue company led by a guy that sounds nutty. It's now a company with unprecedented historical revenue growth, not only leading but dominating market share in the segments in which it competes, two proven successful products and third one with a long waiting list, led by a guy that sounds nutty. Further, I also expect lenders to soon start shying away from existing manufacturers, due primarily to declining used ICE car value estimates
which back lease debt on balance sheets (think subprime crisis where declining home values backed mortgage debt, but on a smaller scale and more focused impact on the economy). This will free up capital to be lent to the only growing company in the industry at a cost of debt that Tesla has never enjoyed before. If the stock rises 50% in 2017, which is possible if the Model 3 ramp-up goes smoothly as so far seems to be, this would bring Tesla's market cap to $75 billion, so $5 billion
non-convertible debt, which is not dilutive, at an interest rate of 8% or below, to fund the start of Giga 3/4/5 would keep debt-to-market-value-of-equity ratio to a very reasonable level. I expect the majority of Giga 3/4/5 cost to be financed using Model 3 gross profits, which I expect will grow significantly more than operating expenses due to operating leverage, throughout the next three years. This math excludes any gross profits from Tesla Energy, which is part of my margin of safety.
I know 4m in 2020 sounds crazy on the face of it, but when I break it down to its components like this, considering all the layers of margin of safety that I explained above, it sounds reasonable. This is why 4m in 2020 is my base case.
