Napkin math: some fundamental analysis -- my style (too long for anyone to read: conclusion at bottom):
2019 Tesla margin:
500,000 cars/year (run rate 2018 plans, so 2019 deliveries, pessimistically 0 growth (false), no delays (probability offsets 0 growth pessimism; lack of quantification of offsetting is why this is one napkin))
100,000 S & X/year @ $15,000 margin
400,000 3 & Y/year @ $ 8,000 margin
$4,700,000,000/year margin
Google says 20 Gigawatts of solar panels operating in USA, which "can power 4.6 million homes". I'm going to assume the following:
- 2 PowerWalls per home
- That this extrapolates to similar PowerPack numbers for utilities per solar wattage
- Utility scale is 25% of margin of home scale
- Utility is similar to commercial
- Utility & Commercial is 75% of market
Worldwide GWp cum. estimate for 2015 is 233. Extrapolating 20GW:4.6million::233GWp: that's 53 million homes. Per year: 55GWp predicted install 2015. 233GWp:53GWp::53million: 12million homes usage install each year.
Assuming average of 2 PowerWalls per home that install, and at that average, about 5% of homes install storage, and Tesla gets 10% of the market share: .05*.1*12,000,000=60,000 per year. Right now there's pent up demand from utilities and older homes that already have solar and wind energy, so that might be one quarter of sales, so 80,000 per year. Assuming 75% of that is 60,000 utility and 20,000 home, so utility (PowerPack) 25% margin: a normalized 35,000 per year install worth of PowerWall margin (a virtual number that fills for both PowerWall and PowerPack). I forget the calcs for margin for Tesla on these, so assuming 1,100 margin each, is $38,500,000/year margin.
Adding 38,500,000 to 4,700,000,000 is not even a rounding error:
4,738,500,000.
I'm doing something wrong. If I take
Elon Musk thinks Tesla can sell batteries much faster than cars at its word, I get:
$5G/y * 2 = $10G/y -- but wait, automotive margin is not the same as roof or battery margin; the cars are more expensive (by far).
From Tesla's announcement:
Gaining Momentum with Tesla Powerpack
"To date, nearly 300 MWh of Tesla batteries have been deployed in 18 countries,"
How much Wh is a PowerWall? 13,500Wh. So, 300MWh/13,500Wh=22,222 PowerWalls to date. Assuming doubling every 2 years during GigaFactory output growth phase in the near term (2019), that's:
2016 22,222 that year (assuming 0 installs in November & December despite 93MWh planned for install in only 2 sites)
2016 29,111 that year (without above assumption)
2017 44,444 - 58,222 that year
2018 88,888 - 116,444 that year
2019 177,776 - 232,888 that year.
Ok, so assuming normalized 1,100 margin per PowerWall, and that that causes PowerPack margins to be out of wack, that's .4375 * $1,100 = $481, or $85,554,700 - $112,019,128/year margin in 2019.
Home building peaked in 1972 in USA (
!). Currently, it's at 1,323,000/year. Let's assume Tesla Roof can garner 2% of the new roof market by 2019: that is 26,460/year. Let's assume that Tesla Roof costs 12,000 per home average, and margin is 2,000/roof, so that's $52,920,000/year margin USA. Multiply by 4 to get worldwide, and it is $211,680,000.
So far, $5 billion margin per year. Why do I think Solar and Battery will have so little margin compared to Auto? Because I see actual Auto numbers, and no full production battery and solar numbers.
What if I extrapolate out into 2021? Now, I assume battery sales slow to about even as they are understood by the marketplace and just go for new installs and utility sales, although having enjoyed huge growth, will also saturate, so let's assume around one quarter of all new homes built with Tesla Roof get one PowerWall and an equal amount purchased by utilities. Tesla Roof will have around 5% of all new homes in 2021, so at a run rate of new homes of about 5,000,000/year, that will be 250,000 getting Tesla Roof, and 125,000 PowerWalls that year. That seems like excess PowerWall. Let's check the capacity of GigaFactory 1 that does batteries:
Wiki says "Its projected capacity for 2018 is 50 GWh/yr of battery packs and its final capacity upon completion of entire factory is 150 GWh/yr." When is full Hamster Heaven? I'm going to skip looking it up. I did anyway.
LYBIO.net Is A Movement For Internet Online Accuracy For Speeches, Text, Words, Quotes and Lyrics. "2020 or so". So, 150GWh/yr. Half to PowerPack & PowerWall. Extrapolating using my normalized PowerWall scheme: half of 11,111,111 PowerWalls/year is 5,555,555 PW/y. Using my margin multiplier of .4375 * $1,100 * that number, I get $2,673,611,111 margin/year in "2020 or so". Let's say that's 2021.
But, how many PowerWalls is that? That's 5.5 million/year. I said "125,000 PowerWalls/year is excessive". Wait a minute. This doesn't add up. Is it 125,000/year or 5,555,555/year? That's 44x more in one estimate than the other.
On a log scale, that's 5.09 vs 6.74, the middle of those two is 5.915, so 10^5.915 is 822,242 PowerWalls/year (using my weird scheme), or $395,703,962/year PowerWall + PowerPack margin.
Returning to 2021 roof, we get $500,000,000/year margin out there.
I forgot auto. How many auto is Tesla doing in 2021? Let's say it's 3x 2019, since there's new competition then. That's 1,500,000. Let's assume the margin is 3x as well: $14,100,000,000/year.
The napkin 2021 estimate margin is 14,995,703,962/year, or around $15G.
What's the market cap supposed to be? Well, P/E ranged from 4.78 in 1920 to 44.20 in 1999. The geometric mean (aha! useful idea -- n-th root of the product of the n elements) was 14. So, if earnings is 15 gigabucks, then the market cap should be 210 gigabucks, or in today's almost 150,000,000
Tesla Motors Shares Outstanding | TSLA Nasdaq - Macroaxis it is $1,400 per share. Using 2016's average P/E ratio, it is ... wait, that's not a good idea per
Is the Stock Market Cheap? - dshort - Advisor Perspectives so I'll use PE/10 of 27 in August 2015, which indicates that in 2021 Tesla shares would be 2,700 per share. That's a market cap of $.4T.
What if it's not all profit? Like, they just keep expanding. Well, actually, none of it is profit if they keep doing that, right? It's just all future investment. So, the P/E ratio would always be something like 0 if they keep expanding at equal to the rate of income.
Let's say they expand until every car in the world is EV. Let's assume all cars sold in 2021 are EV, to be pessimistic about Tesla's opportunities. 60,000,000/year cars made, a billion in the world: replaced every 16 years. So, let's see: 2021 + 15 (let's assume SOME progress 2012-2020) = 2036. (I think it will go faster, but this is just throwing out napkin calcs.)
FUD writers say "Tesla will have 40 competitors in 2021". Let's say Tesla just plateaus in 2021, then: they make only $225,000,000,000 during that time, at which time, the market will only have electric vehicles left. Let's say that Tesla says "job done" and closes shop THAT DAY. Well, that means they have to cash out, and give shareholders $1,500 per share. That's very close to my original 14 P/E ratio.
But, what if they keep expanding that whole time, and never make any money until 2036, but then don't close shop? Let's say competition makes them expand only by 25% a year rather than 50% a year. This is the 3rd time I'm seeing this 44 number in this napkin practice: 1.25^17=44, so that's a vehicle build rate of 22,000,000/year. Does Ford do that? Um, no: 2,500,000/year. Wordlwide it is 60,000,000/year, and how many of us think Tesla will get a third of that? Sure, many of us do, but I'm trying to be conservative and pessimistic here. So, let's give them the same market share as 1/4 of Ford, so not many more than in 2019. In fact, less than in 2019. What the heck?
Tesla already outsells Mercedes in their local name brand class. So, let's say that it is not inconceivable Tesla could do 1/2 as much as Ford. Well, Tesla is an international brand, so worldwide it might be closer to 1/1. 2,000,000/year could be obtained by 2026 in this manner. Let's say that's also when they plateau. And let's say Tesla cars built then are so good they last 200 years each. So, they'd have 2026-2036 free and clear profit, so 10 years at ... let's pessimistically say margin is only 3x what it was when it made 1/4th the cars, which is a horrible assumption, but they will have competition, so ... $15G/year, like I said above. That's $150G profit until they've filled the world with cars that last 2 centuries, and for 150M outstanding shares, that's $1,000 per share. But how many people retire and start spending all their money in 2036? I do: I turn 65 that day. But that's not a safe bet on how people will value the company going forward today. Oh, I forgot Tesla Energy. Their portion was $800M margin in 2021 (that's extreme lowball), and could be huge growth. Let's say 4x that, or 20% of all new roofs for a rule of thumb (which confers battery and other sales -- not just roofs -- just a rule of thumb). That's $3.2G/year. Add that to the $15G/year, for 10 years, and that's $1,213/share.
The geometric mean of my pessimistic share price calculation for Tesla sometime in the next two decades (ranging from today to 2036) is $1,470. The market cap would not be a trillion dollars, like many have said here, but it would be over a fifth that. Log of my implied market cap is 11.34, and log of a trillion is 12. Log of today's market cap is 10.44. The geometric mean of those 3 numbers is 11.24, or 174,457,642,731, or a share price of $1,163.05. Notice that my conservative number is closer to the wild unrealistic number given by many who say trillion dollar cap, and further from the current market cap. So, who's being wildly unrealistic right now? Of course, everyone wants to be cautious, and a lot of this is based on "show me the money", proof of profit, proof of future events. Short term indeed. Let's say the investment timeline is 5 years: then, they'd be thinking that they could rake in only around whatever profit the next 5 years' investors would be considering achievable less the current share price, so let's say $150G, or $1,000/share, so a profit in 5 years of only $820/share when they sell 5 years from now. So, do they value the stock price at $820/share today? The idea is to make more than 10% each year, so they'd have to pay less than $500/share to make more than a 10%/year growth investment.
If their short term profit motive was more than 10% and less than 5 years, then ... what? Let's say one year is the view. Tesla would earn $20/share in one year if they stopped investing in their future, which they're not going to do. The value of the company is valued for their long term plans. 2021 is a better short term thinking valuation. So, that would be $100/share profit per year, and growing (see Tesla Energy). Let's assume everyone gets totally un-trusting, and only keep share price at its annual profit over a multiple of the next two years, so 3x that: the share price would be $300 today. But, headed to above $1,000/share.
I'll look at it another way: their growth has been about 25% every year in cars sold. Their margins are around 25%. So, it's basically a growth rate in projected earnings of somewhere around 25% per year plus optimism minus pessimism. 25%-10%=15%. How much above 10% makes it worth it? Half of the implied potential above comparison? So, 7.5%? That would mean growth of 17.5% is acceptable, and the stock price would have to be $211 within a year, and $248 within two years. It would be $403 in 4 years, and $902/share in 10 years. By 12 years, it would be $1,246/share, which is right around some of my conservative predictions above.