Short Tesla (TSLA-NASD: $250). Target Price = $12. Twelve Dollars. That would be a 95% drop from its share price today. There is a lot of emotion and hype around the company and stock right now, with its cheap electric car model, the Tesla 3, getting banner headlines every day on surging sales.
But Nathan Weiss, an institutional newsletter writer out of Rhode Island, runs the unit economics on the production costs of Tesla–and finds the stock wanting. He presented his vews at our Toronto Subscriber Investment Summit. The entire transcript of his speech is below, or you can click on the video.
He customized his powerpoint for Canada, showing–every few slides–how much beer investors would have to drink to believe each successive claim by Tesla management.
I subscribe to Nathan’s service, and his ideas over the years have made me and my OGIB subscribers more money than anyone–especially in ethanol stocks in 2013-2014. His research and analytical skills are second to none. The timing here could hardly be better to understand why investors should short Tesla.
As you can see we’re probably negative predisposed to Tesla Motors, this actually shows a recent fire of a car charging a (inaudible). We do like lithium and we’ll talk about it in the presentation.
A couple of quick disclaimers, this is for information only and not for investment advice, we may change our mind tomorrowplease don’t sue us.
A little bit about my background. We’re a Rhode Island based independent research firm and we try to come up with ideas and themes that we then write about and sell to our clients, which are primarily hedge funds. We just follow a few stocks at a time and we try to know them very well. We recently published a 99 page re-initiation report on Tesla Motors. We limit ourselves to 30 clients.
Today we’re going to talk about Tesla Motors and specifically some more of the controversial things surrounding the company, as well as a little bit of insight into what we do as a research process on a stock. We’re going to talk about the credibility, environmental aspects, subsidies, Tesla’s Giga Factory which is their largest lithium ion battery plant. The Model S and orders backlog and some of the financials and economics.
From the consumers point of view the Model S is a great car. Most controversial stocks and companies have a great product and it’s actually the numbers behind it that are troubling. I personally love Krispy Kreme donuts and the company went bankrupt.
So from the consumer’s point of view the Model S is a cutting edge car with a really cool UB. It’s green, it’s sexy, very high performance and the fastest models are sub 4 seconds, 0 to 60 and been billed as the safest car in America by Tesla and it’s just plain cool. People stop and take pictures and want to ride in one and it’s a really cool product.
From a stock analyst point of view…so the bullish analysts will tell you the production constraints they have more orders than they possibly can deliver, they have industry leading 25% gross margin, they’re profitable on a non GAAP basis and if you move some numbers around you can make money. There is a sustainable cost advantage and they’ll do 500,000 deliveries in 2020 up from 50,000 deliveries last year.
The bearish analysts will tell you they have zero credibility, an unreliable product, aggressive accounting and a couple use a stronger word but…heavily subsidized and they lose thousands of dollars on every car.
So if we quickly look at Tesla’s current products on the upper left we see the Model S and it’s generally $70 to $140,000 US dollar car and the average is around $92,000. As you see on the right, it’s got a massive 17” touch screen which other auto makers have been hesitant to put in for liability reasons.
They introduced the auto pilot which is basically their self driving highway mode where they can follow the lane and set the cruise control and follow the cars and kind of drive itself on the highway…again before most States are allowing it to happen.
They have auto extending handles, touch door handles and to date they’ve done 107,000 deliveries and again 50,000 last year.
The newest product they introduced on September 29th last year is the Model X SUV. The initial model came out at $137,000, which was a pretty negative surprise. Investors and buyers thought it would be about a $90,000 car. They are promising that some of the next models are being priced at about $112,000 US with a pretty good set of options.
It’s got a lot of the same features as the Model S except it has Go Wing doors, electromechanical doors that open themselves and look really cool. It has auto pilot and a windshield that actually extends over the driver’s head to compensate because of lack of sun roof because of the falcon wing doors.
They shipped just over 200 of them in the 4th quarter and we think they’re about at 2,000 deliveries today.
Then the power side of the business and given their expertise and seeing the amount of market and utility demand that they’ve got in the power business as well, they’re producing a 200 pound lithium ion battery that goes into the wall.
It’s really aggressively priced at around $3,500 or sells at $3,000 but by the time they actually install it and include the inverter and everything it’s about $7,000. So it turns out Tesla wasn’t actually as cost competitive as they thought, but that was their entry into a really rapidly growing market.
One of the first things we’re going to discuss is whether Tesla’s management team is credible. If you believe they are then you probably want this, this and this. It takes about 4 to 5 beers to think about Tesla being credible.
So one of the first things we do when we’re looking at a company, whether it’s me or one of my analysts, I tell them to go back and look at the last few years. What happened? There are some things with permits or energy or commodity prices that’s outside the company’s control and we don’t fault management for having as bad a guidance as everybody else on that.
But there are things that are under their control. In terms of raising capital which is clearly under Tesla’s control; there have been multiple instances, 2 of the most egregious on this slide where in 2013 they promised they didn’t have any plans to raise capital right now and we spent no time on that at all and 7 days later they raised $230 million in equity and $450 million convert.
They did the same thing in August and the CFO implied on a conference call that we’re comfortable with our cash levels and 9 days later they did an equity offering that was up sized to $785 million.
They’ve done similar things in production. They’ve got a Roadster production and the time tables have been delayed. They originally said they’d build 650 Roadsters in ’08 and they built 140. The bottom bullet point, they actually missed every year’s delivery except for 2013 in terms of their initial guidance.
And the bottom of the slide shows the Model X which we showed a couple slides previously was originally to come out at the end of 2014 and came out late in 2015. At the time in late 2013 they said they were at the final brush scopes of the design and completely ended up revamping the front end and making major changes in mid 2015.
Lastly on the delivery guidance they made some ridiculous financial statements. The first one in May 2010 they said they weren’t profitable because they had Model S expenses when developing the vehicle. They did $20 million in revenue and almost $4 million of gross profit but then had $16 million of SG&A.
One of the things that I think is really important what management says on a call or conference they’re not really held accountable. If you put something in a SEC or government filing that’s actually important, it’s on the record and reviewed by lawyers.
They had guided in early 2015 that they were going to hit a 30% gross margin and deliver cars at a 100,000 vehicle annual rate at the end of 2014. They delivered 17,000 cars at a 22% gross margin, arguably we think the actual car’s gross margin is only 17%.
Is the Model S green? A lot of these things are particular to Tesla but it doesn’t mean all EV’s aren’t green but if you believe the Model S is then you probably need a few beers.
On Tesla’s website they calculate the effective CO2 emissions of the Model S sedan are about 176 grams per mile driven and that compares to a small gasoline car at 240/280 and the Jeep Grand Cherokee at 440.So when you go into EIA’s data and start to actually verify the claim there is about 572 grams of CO2 emitted nationwide in the US when you include the 7% transmission loss on average from power plant to customer.
When you multiply that by miles driven the average efficiency we get about 216 grams of CO2 per mile. That’s not bad and a little higher than what Tesla stated but pretty efficient.
However, lithium ion batteries are inefficient when they charge and you get about a 15% charging loss partially because of heat generation and then just the chemical reactions that take place. So when you add that on top of it you end up getting 255 grams per mile so it’s kind of in the range of a small passenger car.
Taking that a little further, the Model S in particular uses a lot of power when it’s idle. So they’ve taken the role of using over 7,000 small lithium ion battery cells that they then actively circulate coolant through to warm and cool it even when the car isn’t being driven.
It monitors the temperature and there’s tons of fire detection and fire suppression equipment on board. Those total units of consumption of that monitoring and temperature compensation system is about 3 ½ kilowatt hours per day more than a small refrigerator. When you divide that out over miles driven that adds another 61 grams per mile for your effective CO2 emissions and that gets us up to 316 grams.
And in the constructional lithium batteries themselves is very CO2 intensive, over 100,000 mile battery life that adds another 153 grams per mile of effective CO2 emissions and we’re not even including things like the car actually weighs over 4,000 pounds, has aluminum fenders and so it’s a very energy intensive car to build.
So we think the total effect of emissions is about 469 grams a mile, which if we go back to our Jeep Grand Cherokee is actually more than that at 443 for the Jeep Grand Cherokee.
Other emissions matter and so Nitrogen Oxides are perceived a lot (10:10 audio skips) after the Volkswagen scandal. The EPA reports it’s about .508 grams of NOX emissions per kilowatt generated by the US grid.
Based on our previous driving efficiency and power consumption calculations the Model S emits about .27 grams a mile and that’s up against the EPA’s Tier 2 standard you’re past your vehicle mile limit of .05 grams per mile.
Sulphur dioxide is also similar but roughly where it’s given approximately 30% of the US power grid’s pull you end up with sulphur dioxide emissions of about 182 times the gas powered automobile. So 20,000 Model S’s in the US emit the same sulphur dioxide as 3.6 million passenger vehicles, gas powered.
If you look at the bottom the Volkswagen Passat diesel which is one of the cars being recalled emitted .25 grams a mile of NOX, so Volkswagen took a $17 billion recall position and facing EPA fines up to $37,500 per car. The Model S we calculated emits .278 so a little bit higher and Tesla has received a $7,500 Federal tax credit, $2,500 State credit and other auto makers including Volkswagen have been forced to pay a total of $412 million in zero emission credits from Tesla.
Should Tesla be subsidized? Probably not.
I always like to use the Amtrak example here and in the US Amtrak is constantly cited as a big wasteful bloated organization losing hundreds of thousands a mile. Newspapers have gone as far as to say on some of the longer routes it would be cheaper to buy the passenger airline tickets and shutting down Amtrak. It would be lower costs.
So every year Amtrak gets drug before Congress and they testify why we need it and begrudgingly end up getting funded. The most recent fiscal year is a $289 million operating budget funding.
For 2016, we estimate and it’s actually below consensus but we estimate 35,600 Tesla vehicles will be sold in the US and for that the Federal tax credits will total $267 million, almost equal to Amtrak’s subsidy and also about $2 per household and so everybody in the US has paid a couple of bucks for the Model S over the year.
In addition, we mentioned those greenhouse gas EVD credits and so various States, particularly California, require that a certain percentage of your vehicle fleet be 0 tailpipe emissions and if you don’t hit that you have to buy credits from other auto makers. To date Tesla has received a total of $583 million credit revenue from other auto makers.
They’ve also received other nice tax subsidies such as a $1.2 billion incentive package from the State of Nevada for floating their Giga Factory. From the State of California they get a lot of tax exemptions and accelerated depreciation write off and the like.
Is the Model S reliable? We think you have to drink a lot to believe the Model S is reliable.
It doesn’t mean people don’t love the car still and people endured some incredible reliability issues and still had raving reviews about the company. But if you look at some of the independent road tests that have been done CNN did a road test where the touch screen and the door handles refused to work and they had to call Tesla and have them send an update.
Consumer Reports initially gave the Model S a 99 rating tying the top rating of a Lexus that they ever gave and recommended to buy it. that said as they drove the car at 15,000 miles the article said they had more than their fair share of problems, including door handle problems, a blank center console, a seat belt buckle that broke, all types of noises and they replaced the 3rd row seats, a 12 volt battery, the HVAC filter and the power train coolant light.
They actually took Tesla off the recommended list and saying the 2015 model had lower quality than the previous years.
Edmund’s was even worse. In February 2013 they plunked down $105,000 for a top of the line performance version of the Model S85. They were going to do a $20,000 road test which they then extended over 30,000 miles.
In the first year of ownership the touch screen had to get replaced, the sunroof leaked, rear tires wore out prematurely which is a common problem and importantly they actually had 2 driving unit replacements which in the Model S the main battery assembly the motor, inverter and transmission effectively a 1 gear transmission all get replaced in a single unit.
They actually had a 3rd replacement at 30,000 miles on the odometer as well as right height sensor, motor mount.
Their summary was the car has 30,000 miles on it and hasn’t been out of service for 30 days. In the wrap up review, which is still available online they reported 28 total issues.
Much like Consumer Reports there is an organization called True Delta which tracks actual repair visits from repair centers and they show here the ranking of US auto makers with models available in the US and the brackets beside the names is how many cars are available by that auto maker and so Tesla shows 2 cars available. They actually rank as the least reliable car in America.
Then Plug In America which is an interesting group and it actually is a Pro EV group. It’s transformed into a California non-profit and they are quite a large organization now where EV owners submit repair data and efficiency data and form a general community about all their respective EV’s and not just Tesla but it’s for all plug in vehicles.
With the idea that they can also help promote electric cars and have some really cool data.
They show in their data set much like we’ve seen in Tesla forums a lot of driving unit replacements. So when we were looking through the forums one of the things…we go through forums not to get anecdotal information but try to gather hard data out of forums. It’s really time intensive but with the right effort sometimes you can get some really cool conclusions. So we found between 30 and 40% of Model S owners had a driving unit replacement.
Based on the Plug In America data 77% of cars over 50,000 miles had a driving unit replacement. And 13 out of 14 in the data set with over 65,000 miles had 1 or more driving unit replacements. And we categorize the data and break it up by year and mileage, so the way you read this at 25,000 miles 27% of the 2012’s had a driving unit replacement, 23% of the 2013’s and 23% of the 2014’s. The 2015’s are still too new to have that many cars in the data set with over 25,000 miles but there is a large number that are at 10,000 miles now and look to be moving in track with previous years.
The Giga Factory which is Tesla’s lithium ion battery production center is unprecedented. So EV critics pointed out a couple of years ago that Tesla’s plan to sell 20,000 EV’s in 2020 would require more than 100% of 2013’s lithium ion battery output for cells, the types you use in laptops and power electronics.
This is actually one of the interesting things about the lithium market what with Tesla’s skeptics if you look at global lithium vehicle deliveries they will grow by more than 500,000 units between now and 2020. That growth will exceed everything put into consumer electronics today. So we are actually very positive lithium and positive lithium demand. We’ve recently seen some pricing quotes that lithium is trading at 13,000 in China.
This is what Tesla promised. It’s a massive Giga Factory, the largest building in the world and a $5 billion production center. Then there were a couple of little news articles and the union that was actually working on it said the project was cut back 80%. The CO of Tesla said it was a test project. This is a recent drone photo of what’s being built.
Are they production constrained? One of the main cases is they have more orders than they can deliver today and they can produce whatever they want and no worries about backlogs. That’s even a little harder to believe.
It’s a busy slide but what we do is go through forums and find online posts from groups like Plug In America where people ordered their car and in the good old days 2 years ago Tesla used to provide sequential production numbers.
If I ordered a Model S I might get production number 13,001 and Keith orders 2 of them and he gets 13,002 and 13,003, etc. and so it was really easy to track orders. They’ve since muddied the water and give unsequential numbers and skip thousands at a time. They do whatever they can to hide it, which to me is a very good sign they don’t want you to know something.
So when we go through their data we suggest they’ve actually been burning down backlog and we think they only have a 2,000 vehicle Model S backlog. Putting that graphic here which is a much more fun way to look at it this table shows that the red vertical bars are individual order clusters and the height represents the number of days from the time you order your Model S to the time it’s delivered.
One thing we do know with relative certainty is how many Model S’s are being produced each week. They frequently do conferences, trade shows and presentations where they talk about their current weekly production rate so we can track that pretty well.
You can see in the order cluster, so right around the middle of the chart was 2014 when they introduced all wheel drive and auto pilot. More recently production has been flattened by a low priced Model S70 and so a smaller battery but a really aggressive $70,000 price point.
And the black line shows how many days on average between the order and start of production which has been hovering around 10 to 25 days recently. So with current production rates for the Model S cars around 1,000 per week that indicates a couple thousand cars in backlog.
With the Model X they still provide sequential model numbers so we know up until September 30th anyway, we know at a hard and fast basis they have about 31,000 orders. They quit giving order guidance and quit sequential reservation numbers so they’re intentionally trying to muddy the water now that they’re in production.
The other thing is generally a company with more demand than supply raises prices and does things like cut back on SG&A and Tesla has kind of done the opposite. If you look at and in Canada you guys get a pretty good deal but in the US the dollar price of a Model S has gone from $99,000 on the far right in late 2013 for a fully loaded car to $84,000 today.
In Canadian dollars you’ve actually seen the price has gone from $104,000 Canadian to $110,000 despite the massive move in the currency. So they’re introducing a new lower price Model S.
The question is can they sell 400,000 cars a year and will that solve all their financial problems? I think that one speaks for itself.
So we’ve seen that people have been nice enough when they go through Tesla tours to take pictures even though they’re told not to. This is what we think the Model 3 looks like and that was actually a clay model seen in Tesla’s facility. It’s supposed to be a $35,000 car that does 200 miles per charge. They claim it’s going to be 400,000 in annual deliveries.
If you look at their competitors today which is the Nissan Leaf which is sort of a boxy, hopefully no one takes offense, but kind of a boxy little car and they sell 44,000 copies globally. The I3 which I’m actually a big fan of and is a cool little car that gets a battery bump this summer and goes up to 120 mile range and also has a gas engine on board to back it up. It does 28,000 cars a year.
Is Tesla profitable? No.
So the far left column shows Tesla Motors financials as the report it compared to Ford and GM and we also threw in Renault and Volkswagen. So Tesla claims in their financials that they show a 25.5% gross margin compared to 11 to 15% for other auto makers.
Within that they also showed almost 22% SG&A and so their sales and overhead are incredibly high compared to 7 to 10% for Volkswagen.
So what’s going on? GM, Ford and pretty much every other auto maker the manufacturer sells the car to an independent dealer at about a 10% discount to MSRP and the dealer bears the cost of selling the vehicle to the customer, which is generally a break even transaction and they make money on service and extras.
Tesla because they sell direct to the customer they include that as an operating line rather than part of the cost of goods sold. So if we actually re-allocate 10% of revenue from SG&A up into cost of goods sold they look like a regular auto maker. So you get 11% of sales as an SG&A expense, which is better than Volkswagen and you get a 15.5% gross margin which is kind of in line with everybody else.
The problem is according to GAAP when you do development work for a car by factory equipment if it’s for a future model or production equipment it’s going to last multiple years and so you get to capitalize it so when you spend the money you don’t put it under income statement you call it an asset and then over the useful life of the asset, generally over 5 to 20 years you depreciate it.
When you’re doing development work and refining a certain car, let’s say you have a drive unit or door problem and you need to fix it, those expenses based on the current car production get expensed. So GM and Ford actually per GAAP accounting they include those expenses in the cost of goods sold and Tesla breaks them out as a separate line item and quite surprisingly that’s actually over 17% of sales is expensed R & D.
If we re-evaluate Tesla’s accounting we still get the same operating margins and so we’re not changing operating margins at all, but we’re re-allocating part of the SG&A and all the expensed R & D into cost of goods sold.
So we think they actually generate negative gross margins and those gross margins are deteriorating not improving. Then on top of that Ford, GM and Volkswagen spend 5 to 8% of revenue on capital expenditures new equipment and new models and Tesla spends 42%.
So there is definitely some flexibility and room there in what you call capital expenditures and I think there is a reasonable possibility that they’re putting production costs and other costs into that capitalized expense line.
So quickly the blue bars are SG&A expense per vehicle sold and that’s the most important on sort of a unit economic basis. Going from the far left to the far right the green line shows how many vehicles they’ve delivered each quarter. So deliveries are going up but SG&A per vehicle is also going up and again that’s not supposed to happen.
Most companies have positive operating leverage where the cost per unit goes down as you increase production. Again we think this shows they’re not a production constraint company and they’re actually doing everything they can to try and increase production and increase demand.
The same thing on the capital expenditures and we just show a couple of charts showing capital expenditure per car is actually increasing as deliveries increase.
And a quick point on the capital intensity of auto manufacturing and instead of building one car, Ford and GM have to spend about $35,000 on the total manufacturing equipment to support a single annual unit of output and so as of today Tesla has 100,000 cars of production capacity per year.
At Tesla’s recent spending of $51,000 per car they would have to invest $11 billion to get to 200,000 deliveries per year and the company currently has a $25 billion market cap.
If they reach their goal of 500,000 deliveries in 2020, we calculate they’re going to spend more than their current market cap, so almost $26 billion on capital expenditures to do so. We argue that Tesla (26:43 – inaudible) probably don’t want them to do that.
Lastly, a quick touch point on the unit economics which should sober us up a little bit. Again we like to look at the unit economics, the per unit cost of the car and this just shows what the previous slides have shown. Revenue per car is decreasing over time and so it goes from the left to the right. SG&A expenses have increased from $9,500 per car to over $20,000 per car in the 3rd quarter.
The same with expensed R& D and they went from $11,000 per car up to $15,000 per car. So Tesla’s loss in cash flow has been increasingly negative.
And like any good promotional company does in the 4th quarter they actually introduce a new cash flow metric saying on their own adjusting cash flow metric they’re positive.