theologyandfishing
New member
There's so much to cram into this DD/model that I'm going to try to keep each point brief.
Table of contents:
1) First, what are the risks; Quick Notes
2) Master equation
1) I'll start with the risks to my valuation model:
In my opinion, the biggest risk to Tesla is failing to execute on the growth plan. The biggest hindrance I can see for Tesla's valuation will be failing to achieve their 50% YoY growth targets.
If there's an actual economic recession where consumers have to drastically cut back spending, then Tesla would definitely suffer. But this risk is obvious, and probably applies to 99% of growth stocks. Side note: Tesla currently makes such good margin on their EVs, that they have the flexibility of dropping prices to meet demand; however, this would greatly cut into earnings and reinvesting into growth.
And of course there's the silly, but still very valid, risk to the valuation of companies that the market takes an irrational dive of 20%+, and investors lose confidence in stocks for the foreseeable future. However, Tesla's growth target might still be attainable in this economic environment.
I do *NOT* see new EVs from other companies entering the market as a risk to Tesla. There are 80M auto sales globally every year. My model calculates that Tesla obtains 12.1M of the sales (in 2027). This leaves plenty of room for legacy automakers and new EV manufacturers to sell millions of their own EVs.
Quick Note: This model is based *only* on EV sales. This does not include solar and energy storage, AI, robotics, or FSD. I see 2027 as the peak year of profitability for Tesla EV sales (we'll see why in my model in section 2ai)
Another Note: My model below has a base case (based on 50% YoY growth as Tesla aims for), a bear case (based on 40% YoY growth), and a bull case (based on 60% YoY growth) i.e. in the event that Tesla fails to execute on 50% YoY growth, achieving only 40%, then the bear case shows how the numbers could be affected
2) I'd like to start with a simple equation, and expand on the details of the equation. This is for year 2027:
$12,000 gaap earnings per EV * 12,100,000 EVs delivered * 50 P/E = $7.2T market cap or approximately $7000/share
Sections 2a, 2b and 2c will be the reasoning behind $12k gaap earnings per ev, 12.1M EVs delivered, and 50 P/E respectively
2a) This gaap earnings per EV metric will be the most difficult to explain. I'll start with my model and an explanation
2ai) Here's my model: https://docs.google.com/spreadsheets/d/12_Qp2gXP1HV3EC5UOWBytx6Kn5dfNftqkTo0EwP0Ua8/edit?usp=sharing
Note: non-gaap earnings per EV through 2021 by quarter: $5628, $8060, $8622, $9208
Explanation of model: Warning, this is incredibly detailed. I'm going to try to keep the explanation as simple as possible, and I'm more than happy to answer any questions.
Note: I wrote this DD before Q4 earnings came out, and I decided to leave that data point so you can actually see that my prediction was $2.48 non-GAAP earnings (not bad, huh? actual was $2.54)
The black numbers are known numbers (from 2020 Q3 to 2021 Q3). The red numbers are unknown, but projected numbers. From known numbers in the ER reports I was able to calculate ASP (average selling price, i.e. total auto revenue/evs delivered), COG per EV (cost of goods per EV, i.e. ASP*(1- gross margin)), earnings per EV ((non-GAAP EPS*outstanding shares)/EV delivered), and I was also able to calculate a sort of "operational/manufacturing cost per EV" by this equation:
ASP - COG per EV - earnings per EV = "operational" cost per EV
Here's a beautiful illustration of this equation:
So now that we have ASP, COG per EV, earnings per EV, an "operational" cost per EV and EVs delievered.. how do I project to get the red numbers? I extrapolated using exponential decay. Based on the rates that ASP, COG per EV and "operational" cost per EV were already falling, I took that rate and extrapolated out using an exponential decay model. I also assumed 50% YoY growth in EVs delievred (or 10.66% QoQ compounding).
The math for exponential decay was pretty annoying. For anyone interseted, here's the recursive equation: P1 = (P0 - b)*(1 - r) + b where P1 is one iteration past P0, b is the asymptote and r is the rate of decay. Solving for r gives r = 1-(P1-b)/(P0-b), and so I calculated the average r from the past 4 or 5 iterations to "use" the existing rate of decay. Interesting fact, I had to assume some 'asymptotes' for ASP, COG per EV and operational cost per EV, but if you look at my bear, base and bull cases where I've modified them, you'll see a very little difference in the outlook. The main difference between the base, bull and bear cases is simply growth of sales.
So now that I've extrapolated ASP, COG per EV, "operational" cost per EV and EVs delivered.. I can now calculate the project non-GAAP earnings per EV via the equation I used above:
earnings per EV = ASP - COG per EV - "operational" cost per EV
this one:
After finding earnings per EV, divide by the projected EVs delivered, and then by outstanding shares and you have the non-GAAP EPS estimates.
Now for some interesting (but not very pretty) graphs. Again, take 2029 and 2030 with a grain of salt. The S-curve will flatten
ASP graph:
COG per EV graph:
Operational cost per EV graph:
This is my favorite graph. Through no design of my own, but only from exponential decay extrapolations of other metrics, the non-GAAP earnings per EV graph tops out in 2027 as the year we'll see peak profitability per EV.
non-GAAP earnings per EV graph:
https://i.imgur.com/jD9KdaR.png
non-GAAP EPS graph:
https://i.imgur.com/ZlEz6rV.png
ok, so why do we see dropping ASP, COG and operational costs?
2aii) Dropping battery costs can be summed up with Wright's Law (Cathie Wood was one of the early TSLA investors that spoke of Wright's Law). Battery costs were as high as $1000/kWh** when Tesla first started building the original Roadster. It has since dropped to $100/kWh, and is expected to drop to $50 or less per kWh.
**reference to $1000/kWh: J.B. Straubel, former CTO of Tesla, interview: https://youtu.be/aWR5-mo8f1g?t=1520
Here is an excellent reference to declining battery costs as presented by Tony Seba. This presentation was centered around upcoming innovative and disruptive technologies: https://youtu.be/Kj96nxtHdTU?t=405
2aiii) "Tesla isn't profitable!" we heard for years. This is because they were burning cash for R&D. They've made their manufacturing more efficient with innovations like the structural battery pack and gigacasting, and with 2 new factories we'll see even more benefits from Economies of Scale***. Also, as Tesla opens Austin to serve the east coast US, and Berlin to serve Europe we'll see decreased shipping costs. This will cause the "operational/manufacturing" cost per EV to drop.
***https://www.investopedia.com/terms/e/economiesofscale.asp
2aiv) regulatory credits were 13% of earnings in 2021 Q3, and 11% of earnings in 2021 Q4. But I'm not sure I need to make a bull case for collecting free money.
2b) Will Tesla deliver 12M EVs in 2027? This is a fundamental risk for my model, as stated above. If Tesla maintains their 50% YoY sales growth, then they will sell 12M in 2027. We'll see. However, one thing for certain will not change Tesla sales:
2bi) Competition. Tesla has *always* had competition. They had competition in 2013 to their Model S, and they have competition today. It's the ICE car, and Tesla is still managing to sell every EV they make. A couple points to make here:
I wont get into this point too much. It took Tesla 5 years from their first mass produced EV until they managed positive Free Cash Flow (2013 Model S until 2018). It's not exactly a one-to-one comparison for other companies, but Ford, GM, Rivian, Lucid and others will need at least a couple years to achieve positive free cash flow on the EV segment of their business. But let's say all of them succeed in making a nice commercially viable, and profitable EV..
My next point is that the EV market is expanding *rapidly*. New EV models will not likely take sales from Tesla. They will take sales from the Toyota Corolla, RAV4, Camry, Honda Civic, Accord, Hyundai cars, Volkwagen cars, etc. as we've already witnessed. There's no reason to believe that any new EVs entering the market will take meaningful sales from other EVs instead of gas cars.
2c) The P/E ratio will always be a guess. Tesla will be a company with a presence in EV sales, solar and energy storage, large commercial power storage, and potentially FSD software subscriptions and AI driven bots. I believe (assuming a bull market) that there will still be enough frenzy surrounding Tesla to see a 100 P/E in 2027. The conservative side of me calculates these numbers with 50. By the time the S-Curve flattens, the P/E could be 30.. or maybe 60 like Amazon? It's hard to say where Tesla will be in 2030, but they've always innovated and come up with new ideas.
There is one thing I confident about though; Tesla should *never* be trading at Toyota, Ford, or GM's P/E ratios. The ICE car is a dead technology, not dropping in price like the EV, and is proving to be far inferior to EVs in every way. Plus legacy automakers are not growing (some have *lost* sales for the last 2 years). It took Tesla investing billions over a decade to bring down battery costs, and make the EV commercially viable. Something legacay automakers have never done.
Anyway, I hope you guys liked this DD. I'll try to have discussion with you guys in the comments if you'd like. If you have a bearish counterpoint, then I'm all about discussing it; however, it will help me (and both of us) if you point out which part of this equation it affects:
$12,000 gaap earnings per EV * 12,100,000 EVs delivered * 50 P/E = $7260/share
so that we can establish the point of disagreement.
Thanks for reading!
Table of contents:
1) First, what are the risks; Quick Notes
2) Master equation
Code:
a) gaap earnings per EV
i) model & explanation
ii) dropping battery costs
iii) increase in efficiency of manufacturing/innovation
iv) regulator credits will vanish. Tesla is insanely profitable without them.
b) EVs delivered
i) why competition won't affect Tesla's sales
c) P/E ratio
i) reasons and justification for a 50 P/E1) I'll start with the risks to my valuation model:
In my opinion, the biggest risk to Tesla is failing to execute on the growth plan. The biggest hindrance I can see for Tesla's valuation will be failing to achieve their 50% YoY growth targets.
If there's an actual economic recession where consumers have to drastically cut back spending, then Tesla would definitely suffer. But this risk is obvious, and probably applies to 99% of growth stocks. Side note: Tesla currently makes such good margin on their EVs, that they have the flexibility of dropping prices to meet demand; however, this would greatly cut into earnings and reinvesting into growth.
And of course there's the silly, but still very valid, risk to the valuation of companies that the market takes an irrational dive of 20%+, and investors lose confidence in stocks for the foreseeable future. However, Tesla's growth target might still be attainable in this economic environment.
I do *NOT* see new EVs from other companies entering the market as a risk to Tesla. There are 80M auto sales globally every year. My model calculates that Tesla obtains 12.1M of the sales (in 2027). This leaves plenty of room for legacy automakers and new EV manufacturers to sell millions of their own EVs.
Quick Note: This model is based *only* on EV sales. This does not include solar and energy storage, AI, robotics, or FSD. I see 2027 as the peak year of profitability for Tesla EV sales (we'll see why in my model in section 2ai)
Another Note: My model below has a base case (based on 50% YoY growth as Tesla aims for), a bear case (based on 40% YoY growth), and a bull case (based on 60% YoY growth) i.e. in the event that Tesla fails to execute on 50% YoY growth, achieving only 40%, then the bear case shows how the numbers could be affected
2) I'd like to start with a simple equation, and expand on the details of the equation. This is for year 2027:
$12,000 gaap earnings per EV * 12,100,000 EVs delivered * 50 P/E = $7.2T market cap or approximately $7000/share
Sections 2a, 2b and 2c will be the reasoning behind $12k gaap earnings per ev, 12.1M EVs delivered, and 50 P/E respectively
2a) This gaap earnings per EV metric will be the most difficult to explain. I'll start with my model and an explanation
2ai) Here's my model: https://docs.google.com/spreadsheets/d/12_Qp2gXP1HV3EC5UOWBytx6Kn5dfNftqkTo0EwP0Ua8/edit?usp=sharing
Note: non-gaap earnings per EV through 2021 by quarter: $5628, $8060, $8622, $9208
Explanation of model: Warning, this is incredibly detailed. I'm going to try to keep the explanation as simple as possible, and I'm more than happy to answer any questions.
Note: I wrote this DD before Q4 earnings came out, and I decided to leave that data point so you can actually see that my prediction was $2.48 non-GAAP earnings (not bad, huh? actual was $2.54)
The black numbers are known numbers (from 2020 Q3 to 2021 Q3). The red numbers are unknown, but projected numbers. From known numbers in the ER reports I was able to calculate ASP (average selling price, i.e. total auto revenue/evs delivered), COG per EV (cost of goods per EV, i.e. ASP*(1- gross margin)), earnings per EV ((non-GAAP EPS*outstanding shares)/EV delivered), and I was also able to calculate a sort of "operational/manufacturing cost per EV" by this equation:
ASP - COG per EV - earnings per EV = "operational" cost per EV
Here's a beautiful illustration of this equation:
So now that we have ASP, COG per EV, earnings per EV, an "operational" cost per EV and EVs delievered.. how do I project to get the red numbers? I extrapolated using exponential decay. Based on the rates that ASP, COG per EV and "operational" cost per EV were already falling, I took that rate and extrapolated out using an exponential decay model. I also assumed 50% YoY growth in EVs delievred (or 10.66% QoQ compounding).
The math for exponential decay was pretty annoying. For anyone interseted, here's the recursive equation: P1 = (P0 - b)*(1 - r) + b where P1 is one iteration past P0, b is the asymptote and r is the rate of decay. Solving for r gives r = 1-(P1-b)/(P0-b), and so I calculated the average r from the past 4 or 5 iterations to "use" the existing rate of decay. Interesting fact, I had to assume some 'asymptotes' for ASP, COG per EV and operational cost per EV, but if you look at my bear, base and bull cases where I've modified them, you'll see a very little difference in the outlook. The main difference between the base, bull and bear cases is simply growth of sales.
So now that I've extrapolated ASP, COG per EV, "operational" cost per EV and EVs delivered.. I can now calculate the project non-GAAP earnings per EV via the equation I used above:
earnings per EV = ASP - COG per EV - "operational" cost per EV
this one:
After finding earnings per EV, divide by the projected EVs delivered, and then by outstanding shares and you have the non-GAAP EPS estimates.
Now for some interesting (but not very pretty) graphs. Again, take 2029 and 2030 with a grain of salt. The S-curve will flatten
ASP graph:
COG per EV graph:
Operational cost per EV graph:
This is my favorite graph. Through no design of my own, but only from exponential decay extrapolations of other metrics, the non-GAAP earnings per EV graph tops out in 2027 as the year we'll see peak profitability per EV.
non-GAAP earnings per EV graph:
https://i.imgur.com/jD9KdaR.png
non-GAAP EPS graph:
https://i.imgur.com/ZlEz6rV.png
ok, so why do we see dropping ASP, COG and operational costs?
2aii) Dropping battery costs can be summed up with Wright's Law (Cathie Wood was one of the early TSLA investors that spoke of Wright's Law). Battery costs were as high as $1000/kWh** when Tesla first started building the original Roadster. It has since dropped to $100/kWh, and is expected to drop to $50 or less per kWh.
**reference to $1000/kWh: J.B. Straubel, former CTO of Tesla, interview: https://youtu.be/aWR5-mo8f1g?t=1520
Here is an excellent reference to declining battery costs as presented by Tony Seba. This presentation was centered around upcoming innovative and disruptive technologies: https://youtu.be/Kj96nxtHdTU?t=405
2aiii) "Tesla isn't profitable!" we heard for years. This is because they were burning cash for R&D. They've made their manufacturing more efficient with innovations like the structural battery pack and gigacasting, and with 2 new factories we'll see even more benefits from Economies of Scale***. Also, as Tesla opens Austin to serve the east coast US, and Berlin to serve Europe we'll see decreased shipping costs. This will cause the "operational/manufacturing" cost per EV to drop.
***https://www.investopedia.com/terms/e/economiesofscale.asp
2aiv) regulatory credits were 13% of earnings in 2021 Q3, and 11% of earnings in 2021 Q4. But I'm not sure I need to make a bull case for collecting free money.
2b) Will Tesla deliver 12M EVs in 2027? This is a fundamental risk for my model, as stated above. If Tesla maintains their 50% YoY sales growth, then they will sell 12M in 2027. We'll see. However, one thing for certain will not change Tesla sales:
2bi) Competition. Tesla has *always* had competition. They had competition in 2013 to their Model S, and they have competition today. It's the ICE car, and Tesla is still managing to sell every EV they make. A couple points to make here:
I wont get into this point too much. It took Tesla 5 years from their first mass produced EV until they managed positive Free Cash Flow (2013 Model S until 2018). It's not exactly a one-to-one comparison for other companies, but Ford, GM, Rivian, Lucid and others will need at least a couple years to achieve positive free cash flow on the EV segment of their business. But let's say all of them succeed in making a nice commercially viable, and profitable EV..
My next point is that the EV market is expanding *rapidly*. New EV models will not likely take sales from Tesla. They will take sales from the Toyota Corolla, RAV4, Camry, Honda Civic, Accord, Hyundai cars, Volkwagen cars, etc. as we've already witnessed. There's no reason to believe that any new EVs entering the market will take meaningful sales from other EVs instead of gas cars.
2c) The P/E ratio will always be a guess. Tesla will be a company with a presence in EV sales, solar and energy storage, large commercial power storage, and potentially FSD software subscriptions and AI driven bots. I believe (assuming a bull market) that there will still be enough frenzy surrounding Tesla to see a 100 P/E in 2027. The conservative side of me calculates these numbers with 50. By the time the S-Curve flattens, the P/E could be 30.. or maybe 60 like Amazon? It's hard to say where Tesla will be in 2030, but they've always innovated and come up with new ideas.
There is one thing I confident about though; Tesla should *never* be trading at Toyota, Ford, or GM's P/E ratios. The ICE car is a dead technology, not dropping in price like the EV, and is proving to be far inferior to EVs in every way. Plus legacy automakers are not growing (some have *lost* sales for the last 2 years). It took Tesla investing billions over a decade to bring down battery costs, and make the EV commercially viable. Something legacay automakers have never done.
Anyway, I hope you guys liked this DD. I'll try to have discussion with you guys in the comments if you'd like. If you have a bearish counterpoint, then I'm all about discussing it; however, it will help me (and both of us) if you point out which part of this equation it affects:
$12,000 gaap earnings per EV * 12,100,000 EVs delivered * 50 P/E = $7260/share
so that we can establish the point of disagreement.
Thanks for reading!