This is a continuation of our articles exploring the prospects of Tesla (NASDAQ:TSLA) as a business and attempting to value the stock. In the first article, we concluded that the quasi-lease obligations had the potential to adversely impact future earnings if:
1. A substantial number of lessees returned their vehicles after 36 months; and
2. Tesla had to pay a substantial amount to re-market the returned vehicles; or
3. The resale price for returned vehicles was substantially below Tesla's cost in acquiring them.
In the second article, we revised our viewpoint. We concluded that, at a high growth rate for sales of model S's and realistic resale values for returned vehicles, the quasi-lease obligations were not likely to have a material adverse impact on future earnings.
Now I want to explore further. My own view, as expressed on the boards several times, is that if Tesla can produce and sell a mass market vehicle in volume in the $35,000 to $50,000 price range, it will come to dominate the automobile industry. Thus far my view has been based on nothing more than instinct. Now I want to do some quantitative analysis to see if that view is reasonable.
The $64 billion question is:
Can Tesla make and sell profitably enough model Es to dominate the mid-priced automobile market?
So the first question is whether Tesla can produce them at a low enough cost to be able to sell them profitable or not. I have only a clue or three. My clue is that Toyota (NYSE:TM) can produce a Prius for less than that price and make a profit selling it. True, Toyota has decades of experience in mass producing vehicles and Tesla does not. Perhaps that will make the difference between success and failure in profitably producing model Es. There's no way to know for certain at this point but I think not. The techniques are known, the equipment is available; and Elon Musk is a smart man. A nicely equipped Prius plug-in sells for $41,008. Since Toyota makes a profit selling it at that price, it must cost less to produce.
The Chevy Volt sells for $34,995 before incentives and the Nissan Leaf for $34,830 for the SL model with no extras. The mass production costs of those cars are approximately the same as for a Tesla model E, plus the cost of the battery, minus the cost of an internal combustion engine.
How approximately? I don't know. As you must all know by now, I'm fond of speculating about the unknown based on educated guesses from oversimplified assumptions. For exploration purposes, I'm going to assume that those costs are equal and compare the cost of an internal combustion engine ("ICE") to that of a Tesla battery. I'm going to use $3,000 as the typical cost of an ICE. I'm also going to assume that Tesla has gotten up to scale producing model S's and model X's so that it now knows how to efficiently mass produce vehicles.
Now I'm going to compare the cost of the Tesla battery to the cost of the ICE. You may have read in Barron's or the New York Times that a Tesla battery costs $400 per kilowatt-hour. Hogwash. Tesla does charge $400 per kilowatt-hour if you upgrade from the 65 kWh battery to the 85 kWh battery, but that is not what it costs Tesla, just what it charges. Panasonic (OTCPK:PCRFF), which is Tesla's battery supplier, advertises 3100mAh cells at about $1.70 to $2 per cell, which comes to about $153 to $180 per kilowatt-hour. So all other things being equal, the battery costs $8,700 more than the ICE.
But all other things are not at all equal. $153 to $180 per kilowatt-hour is the cost of cells in small quantities. The largest battery cell purchaser in the world must get better pricing. Say, 20% better. This brings the price down to $122 to $144. Also, the cells in question are not equal. The cells Panasonic sells to everyone other than Tesla have individual expensive safety mechanisms built into them to prevent fire. This is necessary for laptops but not for Tesla batteries which have a fire safety mechanism installed for the whole battery. Removing the redundant fire safety mechanisms built into each cell lowers the cost. The bottom line is that Tesla is paying no more than about $1 per cell for a battery-pack, or about $90 per kilowatt-hour. Since the model E will be 80% shorter than the model S, it will use a 6,000 cell battery-pack for a total cost of about $6000. So the battery will cost $3,000 more than the ICE, all other things being equal.
That accounts for the extra cost of the battery pack. But we have not accounted for the fact that the other components of an electric vehicle are a lot simpler, and therefore less expensive, than those of the ICE vehicle. How much less expensive? I don't know. I'm going to guess $3000 less expensive just because it produces simple comparisons at equal production costs. I did say that I was fond of speculating about the unknown based on educated guesses from oversimplified assumptions, didn't I?
So we have a presumed cost differential between the cost of production of an ICE vehicle and a 65kWh battery powered Tesla-designed middle market vehicle of zero.
What does it cost Toyota to build a hybrid Camry?
The 2014 Camry hybrid has a base price of $26,140. The identical Camry non-hybrid has a base price of $22,235, or $3,905 less. That gives us an idea of how much a consumer will pay for a greener, more fuel efficient car. It also gives us an idea of what it costs Toyota to substitute a small battery and electric motor for the ICE in the vehicle.
But let's use the non-hybrid price because that is where the high volume sales are. That $22,235 price is for a car with no options. No one would buy it that way but it is perfect for estimating the cost of making it. So let's use it for that. That $22,235 price has to cover dealer markups and a profit margin for Toyota. Let's say 10% for dealer markup and 10% for Toyota's gross margin. That leaves a cost of 80% or $17,788. Let's say Toyota is 20% more efficient than Tesla. That puts Tesla's cost back to $22,235. Then let's add another 20% for whatever we forget. That brings Tesla's production cost to $25,793. At a $35,000 selling price and the same cost, Tesla would have a gross profit of $9,207 or 26.3%. These numbers are no doubt off; and they could be way off; but they do give us insight into the feasibility of manufacturing the model E at a cost-effective price.
My conclusion is that it is feasible for Tesla to make $35,000 model Es which it can sell at a profit, if it can make them in sufficiently massive volume.
What is sufficient massive volume? Tesla's current production facility has an annual production capacity of approximately half a million cars. A half a million cars is clearly sufficiently massive volume. A much lower number may also do but let's go with a half million vehicles and see what happens. Can Tesla produce 500,000 vehicles across all models combined? That is the next question.
I think it depends. It depends on whether Tesla can buy or make enough batteries for 500,000 vehicles.
What about the batteries? By this point in time, we are already making about 80,000 model S's a year and about 30,000 model X's a year. That's a lot of batteries. Is there any capacity left to make model Es with?
Prolific SeekingAlpha author John Petersen, a very smart man and an expert on the lithium-ion battery industry, is of the opinion that Tesla cannot procure enough batteries for more than about a hundred thousand vehicles a year. For his reasoning, look here and here. John's viewpoint is not to be dismissed lightly. He and his wife are intimately involved with Avicenne Energy, a leading consultant to the lithium-ion battery industry.
I respect both the man and his viewpoint but, after some research of my own, I find myself disagreeing with him on this point.
John points out that the present unaccounted for capacity of lithium-ion cell production is theoretically enough to support only a hundred thousand vehicles a year. I accept that. John also points out that much of that unaccounted for capacity, for one reason or another, is unusable by Tesla. I also accept that. John is of the opinion that for Tesla to compete for the presently accounted for lithium-ion cell production, it would have to compete with purchasers (like tablet battery manufacturers) who could much more easily pass on an extra nickel a cell to their customers than Tesla could. I also accept that.
At this point, our views diverge. John sees no reason that Panasonic (Tesla's current supplier) "will spend another $200 billion of shareholder capital to expand manufacturing capacity so that it can sell batteries at a loss in order to help Tesla ..." I agree. But John has grossly miscalculated how much Panasonic would have to spend. $200 billion is two orders of magnitude too large. Panasonic had a grand total of $43.6 billion in buildings and machinery and equipment, at original cost before any depreciation, on its balance sheet, across all of its lines of business combined. Batteries were about a pittance of televisions, not to mention all the other lines. $200 billion is at least 100 times what the plant would cost.
It will cost Panasonic $200 million to reopen closed battery plants (plural) and install new production lines for small and large format cells. My authority for that statement is John Peterson himself. He says so in that same article. He also says in that same article that Panasonic went from a $500 million to $1 billion loss in its lithium-ion battery division, when Tesla was not buying battery cells from it, to a $41.6 million operating profit, thanks to Tesla purchases. I accept that. That gives Panasonic and others such as Samsung (OTC:SSNLF), LG (OTC:LGEIY), and Sony (NYSE:SNE), a large profit incentive to build plants and make lithium-ion cells for sale to Tesla at that same profitable price. Well, maybe not quite that same price. John points out that the Panasonic plant had been fully depreciated, so it had no depreciation cost built into the battery cells it sold to Tesla. John seems to think that fact was the only reason those sales were profitable.
I disagree. If Panasonic could reopen at least two closed battery plants and install completely new production lines for $200 million, then the cost of that is no more than $100 million a plant. What's missing? The land and an industrial box. What could that cost? My guess is no more than $20 million. So we're talking $120 million in original cost for a lithium-ion battery cell plant which, depreciated on a straight-line basis over 20 years, gives you depreciation of $6 million a year. Panasonic made $41.6 million without depreciation. Taking depreciation into account it would still have made $35.6 million. I think John is wrong. I do not think that the only reason Panasonic made money producing lithium-ion cells for Tesla is because it did so in fully depreciated plants.
Samsung, which produces almost as much lithium-ion cells as Panasonic, is negotiating with Tesla to supply it with lithium-ion cells right now. And there is also LG, Sony and other smaller producers in the game.
So the supply is not there now. But it will be. The law of supply and demand will, as always, prove valid. If they want to buy 'em at a price I can make a profit selling 'em for, I'm going to make 'em. If I don't have what I need to make 'em, I'm going to buy it or build it. I believe that the battery constraint argument is a tempest in a teapot. The supply will materialize at a workable price.
Might cells cost more. Possibly. Supplies cost more all the time without putting the manufacturer out of business. Your suppliers don't want to put you out of business. They'd lose a big customer. And they can't replace this customer with anyone at all. No one uses lithium-ion cells in this quantity. Put Tesla out of business and you have put yourself out of business. The converse is not true though. There are multiple lithium-ion cell producers. But only one major buyer. Who do you think has the bargaining power in that situation?
I haven't even considered the fact that, if push came to shove, Tesla could always vertically expand and make its own lithium-ion cells. After all, if Panasonic stops supplying Tesla, and no one jumps in to fill the gap, all those technicians are now out of work and it only cost $120 million to make a plant. But push will not come to shove. The fact that Tesla could do it, means it won't have to. Been there, done that.
Now we are up to the $63 question. (That's not a typo - there's always at least one more question we all forgot)
Can Tesla sell enough model E's (together with model S's and X's) to justify its present valuation?
I think that it depends on whether Tesla can eliminate range anxiety by building an extensive SuperCharger network that allows a Tesla driver to go most anywhere without wasting a lot of time planning or deviating from his intended route or sitting for long periods at slow charging stations. I know some of you think differently. Those of you who think there is no range anxiety issue can skip to the last page of this article because what follows assumes that there is range anxiety and analyzes the costs of resolving it.
What does it take to resolve it? A SuperCharger network that spans the relevant geographic area with SuperChargers every 180 miles along major highways and a few in each large city. How many is that and how much does that cost? Let's figure that out.
In the U.S. (including Southwestern Canada) I calculate it takes 136 SuperChargers to crisscross the country every 180 miles north to south and east to west. At $300,000 each that comes to a capital expenditure of $40,800,000. Now let's put five Superchargers in each city with a population over a million. There are nine of them. Three in each city with a population over a half-million but under a million. There are twenty-four of them. And one in each city with a population over two hundred thousand but under a half-million. Another eighty-three. That comes out to 197 Superchargers. That's another $59,100,000 for a total capital cost of $99,900,000. Less than $100 million to blanket the country with SuperChargers. Everywhere you go you pass a SuperCharger - or three. $100 million isn't much for a company with $750 million in cash; $400 million in quarterly GAAP revenue; and which recently raised a billion dollars publicly. Tesla can do it.
Tesla's newly published SuperCharger road map goes further. It calls for 247 SuperChargers throughout the US and Southwest Canada by the end of 2015. This is an ambitious schedule and Tesla has already proven that it knows how to fall behind. My slower, less ambitious schedule will support sales just about as well and is more feasible.
At the present time, Tesla isn't doing either. It has come to a dead stop in expanding the SuperCharger network. Today's map of SuperChargers actually shows one less than the current number last shown as already open. One in Texas went missing. Tesla is falling behind in its scheduled SuperCharger expansion. From its website, it seems that Tesla did not build any during the summer and missed the projected number completely. Tesla has put off until Fall its next milepost which it now call "most metropolitan areas."
Looking at the map for Fall 2013, it does not look like "most metropolitan areas" to me. It looks more like West Coast, East Coast, Chicagoish, and Texas. I'm from New York and we are known to be provincial but even I know that there are many metropolitan areas in the middle of the country besides Colorado.
But that aside, why did Tesla build no SuperChargers at all during the second quarter? One possibility is that it was so busy building model S's and selling them that the SuperChargers fell by the wayside. I think that is a big mistake and very short sighted indeed.
Another thing worth noting about the SuperCharger road map is that many of the SuperChargers shown as already existing on the present Fall 2013 map do not exist on the "Today" map. What they are is the SuperChargers that were supposed to have been built during the second quarter but were not. They amount to eight missing SuperChargers. And one of the two that were shown as existing previously in Texas seems to have disappeared. Details, details, details. I hope that they don't lose billions of dollars as easily as they lose SuperChargers.
Anyway, let's assume that Tesla gets with it and builds, builds, builds. Let's say that, by the time the model E goes into production, the SuperCharger network is built in the US; midway to being built in central western Europe; and started in Asia.
Let's limit our SuperCharger network to the European Union since that is where the money is. Eastern Europe has some very large, but also very poor, cities. For convenience, when I talk about Europe from now on I mean the European Union. Europe is smaller than the U.S. but it has more cities with high populations. Putting five Superchargers in each European city with over one million population would take 90 SuperChargers. There are eighteen of them. Let's put three in each city with a population over a half-million but under a million. There are 41 of those, so another 123 Superchargers. And 45 more to put one in each city with a population of 300,000 or more, but less than 500,000. Europe is crowded with cities. There are not so many wide open spaces which do not go through a major city as in the U.S. There's also a lot of large bodies of water in the way. That limits the North-South and East-West driving routes. So covering the major highways between cities will only take another 18 SuperChargers. That's a total of 276 SuperChargers at a total cost of $82.8 million.
Tesla has announced that it intends to have all of Central Western Europe except for France (Germany, the Netherlands, Switzerland, Belgium, Austria, Denmark, Luxembourg and Southern Sweden) covered as well as 90% of England and Wales.
In Asia, I will ignore cross-nation travel since many bordering Asian nations are not friendly to each other. Asia really needs to be done city-by-city. Asian cities are big. There are 146 Asian cities with a population greater than one million. Many are poor and poor candidates for automobile sales. I'm going to limit Superchargers to the 10 larger, more affluent cities at five each for 50 Superchargers total. That's another $15 million.
Obviously we don't have to do this all at once. We can concentrate on the U.S. and Europe, where the money is, and leave Asia with just Hong Kong for awhile. We can put six in Hong Kong and stop there. Hong Kong is small. What if we built about twenty SuperChargers each quarter in the U.S. until we had saturated the country, then five a quarter after that. We build thirty-five each quarter in central Europe until we have saturated it and then five a quarter after that to get France, Spain, Portugal, Italy, Finland, Ireland and other western European markets. We build six in and around Hong Kong and then five each year in Asia after that picking affluent cities to place them in like Singapore. This tracks Tesla's newly published SuperCharger road map pretty closely. Let's see what that looks like. Here is a spreadsheet screen-shot showing each fourth quarter.
You can download the full spreadsheet here.
Why continue building after we've saturated these areas? Because we've saturated them at 180 mile distances. It takes a while to SuperCharge for a 180 mile leg. But you can charge to half the battery capacity in 20 minutes - a long bathroom break. So I would keep going and putting SuperChargers midway between the ones that are 180 miles apart. Not in Asia since we are not driving cross country there. In Asia, we add cities. We build five SuperChargers a year there.
Notice that I am not taking account of any advances in battery technology. Will there be advances? Of course there will. I expect battery technology to continue to improve at a capitalized annual growth rate of 4.5%. How will it improve? The same sized battery-pack will hold 4.5% more usable charge each year than it did the year before. So the SuperChargers will get 4.5% closer to each other, measured by the amount of the vehicle's charge you have to use to reach it. And the same capacity battery pack will get 4.5% smaller and lighter than it was the year before. Let's call this the shrinking law.
If the shrinking law proves true, in no more than sixteen years it will take only half as much charge to reach the next SuperCharger. At that point, charging enough to get to the next SuperCharger will take only 20 minutes. But sixteen years is too long a time for me to wait. So let's go on.
OK, so now it's the end of the first quarter 2016 and we've invested $160.2 million. We've sold a total of 200,000 vehicles, so our federal tax credit for customers has expired. Our share count has increased by about 8.7 million shares as a result of stock option exercises and/or public or private financings. Either way, the purchasers paid for the shares. Let's say the stock held steady at $150. That means we raised about $1.3 billion. That covers the cost of building out the SuperCharger network and probably the cost of a second plant in Europe. If that second plant costs more, we sell more stock or wait longer until more employees exercise more stock options.
At this point, we have SuperChargers all over the U.S. and Southwestern Canada; all over central western Europe (except France); and in Hong Kong and Singapore. We've been making and selling model Es but now we can sell a whole bunch more. Let's go make and sell a whole bunch more model Es. Wait a minute. Can we do that?
OK, we can build them. Can we sell them? At $35,000? That's another important subsidiary question.
A nicely equipped Camry hybrid goes for about $31,000 list. A Buick LaCrosse starts at $33,135 and goes to $40,280 nicely equipped. A plug-in Prius at $41,008. So $35,000 is competitive with moderately priced ICE and hybrid and hybrid plug-in cars price-wise.
But is it competitive feature-wise? Cool-wise? Mileage-wise? What car is Joe and Jane America going to want to drive. If the model S is any indication, the model E will blow away the competition in safety, features and coolness. OK, maybe it won't. But it could. Let's see what might happen to Tesla's valuation if it does.
In exploring this scenario, I'm going to make assumptions on the optimistic side in general because I want to see what might happen. This is ultimately a risk/reward ratio assessment, so I want to look at the better side first. If it's not good enough, I save myself a lot of work since I don't have to look any further. This is my usual approach to relatively higher risk investing.
The assumptions that I am making are all stated in the spreadsheet on the input worksheet. As usual, you can change any assumption and the numbers will update across the whole spreadsheet to reflect the new assumption you made. Also as usual, I'm looking forward to hearing about the logic or calculation mistakes I made. I'm sure I made some. I always do. And lastly, also as usual, I'm looking forward to your feedback in telling me what you think I got wrong and what, if anything, you think I got right.
I used non-GAAP numbers since they are easier to work with and since I think they will converge to GAAP numbers over time.
So here are my input assumptions:
1. Tesla model S will grow sales at a rapid pace annually until it comes to dominate the super luxury worldwide car market with about a third of the total sales in that market by 2020.
2. Tesla model X will grow sales even more dramatically reaching over a quarter million vehicles sold in 2020.
3. Tesla's model E will grow sales even more dramatically than model X and will reach annual sales of about 120,000 vehicles annually with no end in sight.
The Average Selling Price of the vehicles are:
- Model S: $90,408
- Model X: $80,408
- Model E: $35,000
Each vehicle's annual growth rate starts at 100% and slows up by 30%, 20% and 10% annually for the model S, X and E respectively. The net profit margin on each car is 10%.
Outstanding shares start out at 126,729,861 and grow at 3% annually for stock option exercises and public or private placements. SuperCharging stations cost an average of $300,000.
Here is a screen shot of the results:
With those inputs, the results I get are disappointing to me. I get EPS in 2016 of $8.94. I state the three-year-out projection since projecting further than that becomes an exercise in creative magic. But I fancy myself a bit of a magician so I did it anyway. I get EPS of $28.43 in 2020 with a still healthy annual growth rate in EPS of 24% annually. This might justify a PE in the neighborhood of 20 for a stock price of about $560 a share. That's a potential gain of 250%. The key words for me are potential and optimistic. Remember, I said that I thought my assumptions were on the optimistic side.
I compare that with the potential risk. Of course, there is always a possible risk of a total loss on any stock but I don't think that has a significant probability. More realistically, there is a risk that the growth rate will flatten out at 20%; or the model E will not blow away the competition; or the model X will flub; or Elon Musk will have a heart attack; etc., etc., etc. Should one or more of those risks materialize; or any one of hundreds of others, the stock could reach an equilibrium of mid double digits. My personal requirement for higher risk investing requires more than a potential 250% profit potential over seven years to justify the risks. Those are my personal requirements because I do rather well and so I have a high opportunity cost.
So I'm still on the sidelines on Tesla. I love the car. But I would not buy the stock yet nor would I touch a short position with a 20 mile pole. I'm still watching and waiting for a catalyst to trigger my interest in placing a wager. A large pull-back in the stock would do it. A $35,000 model E that I absolutely love would also do it.
I do have to tell you that I carried out the "projections" all the way to 2032. This is truly fairy tale make believe based on mathematics. But I'll give you the results anyway. The results are that by 2032, Tesla has become the number one automobile manufacturer in the world totally dominating worldwide automobile production from the mid range to super luxury markets. This, of course, assumes that everyone else stands still and just lets Tesla take away their market shares. I've never seen that happen before and I don't believe it will happen here. How much market share can Tesla take with Ford, GM, Toyota, Porsche, Mercedes, etc. kicking and screaming all the way. Maybe we'll all find out.
Your mileage may vary. Actually, I'm sure it does. That's why the spreadsheet is configurable.
Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article. I may initiate a volatility spread at any time by buying out-of-the-money calls and out-of-the-money puts.