Photo by Petmal/iStock via Getty Images
When Tesla (NASDAQ:TSLA ) was in the process of acquiring SolarCity in 2016, another Elon Musk controlled company, the purchase was largely justified by SolarCity's "cutting edge" solar roof product, which was about to be launched.
In October of 2016, Musk did a presentation on the former television set of the "Desperate Housewives," as part of his effort to get the deal approved by stockholders despite SolarCity's poor financial performance at the time. As a result, I sometimes fondly refer to the location as the set of the "Desperate Accountants." A Los Angeles Times article discussing it can be found here while a YouTube video of the full presentation is here.
The presentation contained four different styles of tiles, which Musk stated were all solar roofs and they would start delivering them to customers the following summer. It eventually turned out that none of the four was actually functional solar roofs and almost five years later, only one style is even available; in fact it has only been since late 2019 that there have been appreciable deliveries.
I discuss Tesla's pricing strategy with respect to the solar roof and the apparently quite poor economics of it based upon information in trade publications as well as in Tesla's financial statements. I raise the question as to whether the solar roof can ever be a profitable product for the company.
Zachary Shahan of CleanTechnica wrote an article last September which contains some extremely helpful information Elon Musk provided regarding Tesla's economics with respect to traditional solar panel sale and installation. The article was titled: "Elon Musk Explains Why Solar Power is so Cheap: CleanTechnica Exclusive"
The article begins by indicating that Tesla's price for standard solar panels is $2.01/watt (it quoted a $1.49 cost to the purchaser after the 26% federal tax credit).
In it, Elon Musk told Shahan:
Solar panel cost is only ~50 cents/Watt. Mounting hardware, inverter and wiring is ~25 cents/Watt. Installation is ~50 cents/Watt, depending on system size.
The other solar companies spend heavily on salespeople, advertising and complex financing instruments. We do not.
So, there you have it, direct from the "horse's mouth": Tesla's direct costs are $1.25 per watt. Shahan then goes on to say:
Another 76 cents a watt covers some additional soft costs while also presumably providing a small profit.
So, what's that small profit? 25 cents/ watt? Possibly a bit more? If in fact there is a profit, even if small, it makes no sense for Tesla to now only do traditional solar panel installations if combined with a Powerwall purchase, as is their new strategy. This causes me to be a bit skeptical as to whether Tesla is making any money at that price on standard solar panels.
This recent article from Solar Reviews provides some further useful math. It states that Tesla decreased its price for the solar roof to $2.01 per watt in September. The article further states:
Tesla’s active solar roofing tiles cost $2.01 per watt, and the inactive shingles cost $7.65 per square foot. Tesla estimates that a 10 kW solar roof will cost a total of $31,133, including incentives.
Interestingly, Tesla is charging exactly the same per watt for an active solar shingle as for a solar panel. However, Tesla's cost for shingles must be much higher. Tesla currently buys its solar panels at 75c per watt (mainly the QPeak Duo Black solar panels from Hanwha according to this Solar Reviews article.) Presumably Tesla discontinued producing solar panels in Buffalo with Panasonic because it couldn't produce them at the same cost.
Manufacturing costs for the solar shingles Tesla produces must be well over $1 per watt. Installation cost must be a multiple of the 50c per watt that Musk provided for the solar panels; installing a few panels on a roof is obviously much cheaper than on-site installation of the solar roof. These costs must be substantially more than the $2.01 that Tesla is charging. It is difficult to see how Tesla could ever decrease the cost enough to earn a profit at this price.
The one major factor we do not know is Tesla's cost to manufacture (or possibly purchase) and install the inactive shingles. If it is a lot less than the $7.65 per square foot that Tesla was charging when the article was written, then it could conceivably offset the loss on the active panels. (It would be a bit ironic if it turned out that Tesla Energy's most profitable product were standard roof shingles.)
According to this roofing "calculator" article, the typical cost nationwide is $3.50-$6.00 per sf, while it indicates that the price in higher cost areas, including much of California, is in the $4.50 to $8.00 per sf range. Presumably there is some profit contained within these quotes, but a vast majority of Tesla's roof installations are likely in these higher cost areas, so profit is most likely less than $2 per sf; likely unable to make up for any loss on the active shingles.
Tesla's recent financial performance in the Energy Generation and Storage (EG&S) segment supports the conclusion that Tesla is likely losing large amounts on solar roofs. As the volume of solar roof installations have increased, the gross margins in the segment have decreased and then become negative.
The pattern was particularly stark in Q1, when the gross loss in the segment increased by $66 million, from negative $35 million to negative $101 million, and the gross margin plummeted from minus 5% to minus 20%, a result which was not at all "tony." Although Tesla does not provide us with a detailed breakdown, the trend suggests that the problem is due to a negative gross margin in the solar roof sale business rather than anywhere else. The solar panel leasing business is an annuity which generated $111 million of revenue in Q1 and appears to always provide a positive gross margin (at least according to "TSLA GAAP"). This means the $383 million of EG&S sales revenue (per p. 10 of the 3/31 10-Q) must have had a really "ugly" negative gross margin of considerably greater than 20% in Q1 on its sales products.
Tesla has provided figures and commentary in its SEC filings and disclosed decisions elsewhere suggesting that the negative sales gross margin is mainly due to the solar roof rather than energy storage. Among some of the indications are that the sales gross margin deteriorated significantly in the most recent quarter as energy storage sales materially decreased and energy generation sales materially increased. The fact that Tesla is now selling solar roofs only to customers who also agree to buy a Powerwall or two is another strong indication. (Only sell a loser if you can dump it with a modest winner.).
Among Tesla's specific commentary in the most recent 10-Q (p. 35) is the following:
Gross margin for energy generation and storage decreased from 4% to -20% in the three months ended March 31, 2021 as compared to the three months ended March 31, 2020 primarily due to a higher proportion of Solar Roof in our overall energy business which operated at lower gross margins as a result of temporary manufacturing underutilization during product ramp, increased service maintenance costs on solar energy systems where we are the lessor and lower gross margins in our energy storage business as we are ramping Megapack.
The 2020 10-K (p. 35) provided further supporting commentary that the gross margin for the solar roof must have been significantly negative:
Gross margin for energy generation and storage decreased from 12% to 1% in the year ended December 31, 2020 as compared to the year ended December 31, 2019 primarily due to a higher proportion of Solar Roof in our overall energy business which operated at lower gross margins as a result of temporary manufacturing underutilization during product ramp
The standard convention in these explanations is to list the largest cause first, and of course in both descriptions it was the solar roofs. Tesla blames the poor (clearly negative) margin on "temporary factory underutilization" but as sales and presumably production have increased, the losses have become even larger. Tesla's pricing during the past two quarters, as I discussed above, makes it hard to see how the solar roofs could ever be profitable, irrespective of volume.
Tesla did significantly increase the cost of the solar roofs in early April. It has even been cancelling or attempting to revise firm contracts under the former pricing, additional indications that TSLA was losing substantial amounts per roof, losses which would likely not be made up for in volume.
A May 9 Solar Reviews article stated that the newly revised pricing is as follows:
According to Tesla, the active solar roof tiles cost $2.01 per watt. The cost of the inactive shingles varies, based on the complexity of your roof. Tesla divides roofs into three groups:
Although when someone buys a "solar roof," it ultimately does not matter how the individual components are priced, this pricing structure strikes me as odd. Tesla is keeping the active panels at the original low price, while raising the price substantially on the inactive ones. Maybe there is a logical marketing reason for pricing them this way, but it could lead to "adverse selection," with a disproportionate number of customers choosing to have small outbuildings with simple roofs covered with a high percentage of active shingles.
A good example of the kind of situation Tesla could be facing relates to a solar roof the company recently installed for the former U.S. Assistant Secretary for Alternative Energy, Dan Reicher, as discussed in a April 29 New York Times article. (Although there is a very curious caption under Mr. Reicher's photo in the article which says: "Mr. Reicher, a former assistant energy secretary, bought his solar shingles from a wholesale contractor and not from Tesla directly.") The roof is being installed only on a new addition with a simple roof, not the entire house. The article contains a photo of the roof being installed.
The NYT article mainly focused on the recent price increase. The article mentioned a small portion of the April 26 earnings conference call:
During a quarterly earnings call on Monday, Mr. Musk insisted that demand for Tesla’s solar roofs “remains strong” even though the company had raised prices substantially. He described the last-minute increases as a teething problem.
“We did find that we basically made some significant mistakes in assessing the difficulty of certain roofs,” he said. “You just can’t have a one-size-fits-all situation.”
It appears as though Tesla didn't do basic research on the topic of installing roof shingles or undertake a serious effort to test out and cost the installation process for various configurations before developing a pricing strategy for the product.
Much more puzzling, however, is that Tesla has claimed to be installing solar roofs since early 2018, if not a bit earlier.
In October 2016, we unveiled Solar Roof, which integrates solar energy production with aesthetically pleasing and durable glass roofing tiles and is designed to complement the architecture of homes and commercial buildings while turning sunlight into electricity. We recently commenced Solar Roof production at our Gigafactory 2 in Buffalo, New York, and are beginning to install them in customers’ homes.
In October 2016, we unveiled Solar Roof, which integrates solar energy production with aesthetically pleasing and durable glass roofing tiles and is designed to complement the architecture of homes and commercial buildings while turning sunlight into electricity. We have been installing this product at a slow pace to gather learnings about our design and installation processes, and plan to ramp the production of Solar Roof with significantly improved manufacturing capabilities during 2019.
In 2019, we commenced direct customer and channel partner sales of the third generation of our Solar Roof, which features aesthetically pleasing and durable glass roofing tiles designed to complement the architecture of homes and commercial buildings while turning sunlight into electricity. We are ramping the volume production of this version of the Solar Roof at Gigafactory New York, and are increasing our installation capabilities by training our personnel and third party partners.
In 2019, we commenced direct customer and channel partner sales of the third generation of our Solar Roof, which combines premium glass roof tiles with energy generation. We are ramping the volume production of Solar Roof at Gigafactory New York, and we are improving our installation capability and efficiency.
It makes absolutely no sense to me that Tesla could have been installing solar roofs since 2017 or early 2018 and finally discovered in the spring of 2021 that some roofs are more complicated than others making it much more expensive to install their product.
On the recent earnings conference call, Musk also said:
First of all, I should say, the demand for the Solar Roof remains strong. So despite raising the price, the demand is still significantly in excess of our ability to meet the demand to install the Solar Roofs. So production has gone fine, but we are choked at the installation point.
Tesla just asked for and received yet another extension, until December 31, on its requirement to meet minimal employment goals in New York State, so I would have to conclude that production is NOT going fine in Buffalo; it's simply going less bad than installation. The fact that Tesla is now restricting sales of solar roofs only to those customers who also buy a Powerwall suggests the economics are not "going fine" either.
If I were a New York State official, taxpayer or job seeker, I would not be happy that Tesla is restricting the potential market for this product in a way state officials likely never contemplated. Even if Tesla meets the requirements at December 31, I am certain NYS officials had high hopes that Tesla would greatly exceed the minimum requirements.
The acquisition of SolarCity by Tesla occurred almost five years ago. I have written about Tesla Energy a number of times since then, the first one being an article in 2017, SolarCity Adds to Tesla's Losses. In it, I estimated that in the first two full quarters of Tesla's ownership, SolarCity likely cost the company roughly $120 million.
Gross profits for Energy Generation and Storage have been at best merely modestly positive almost every year since (and in fact was negative last year). When SG&A and interest expense associated with the segment are subtracted, the results have been undeniably negative. In addition, Tesla has spent considerable amounts for R&D on the solar roof, a product which will likely never generate sufficient profit to pay for it.
Of course, there are also the VIE investors who are entitled to a material portion of the revenue and income from the leased solar energy systems, a topic I have written about a few times. My most recent article was in September 2019: Tesla's VIE Day of Reckoning has Arrived. In fact, in the most recent quarter alone, the VIE investors received $26 million of income and $32 million of the cash flow from these leased systems.
Solar roof strategy and pricing seems to be constantly changing and does not appear to be at all well thought-out or provide a likely path to profitability. The simple fact that Tesla was recently surprised to discover that installing a roof on a complicated surface is considerably more expensive than doing an installation on a simple one is truly astounding, supposedly after installing solar roofs on a limited basis for at least the past three years.
Solar roofs are not rocket science, after all. Tesla just seems to throw things against the wall to see "what sticks." To a certain extent, it reminds me of a Marx Brothers farce.
In contrast to energy generation, energy storage may ultimately be a profitable product for Tesla. Although Tesla does not provide specific breakdowns, the financial trends and 10-Q commentary I referenced earlier seem to suggest storage may have positive gross margins. Unlike solar energy leasing, storage also does not have the burden of project debt or VIE payments.
In fact, in the recent earnings conference call, Elon Musk confirmed decent storage gross margins, at least for the Powerwalls:
"We're aiming for comparable margins in storage as in vehicle. But it is important to bear in mind that vehicle is more mature than the storage. So -- we're already are at margins with the Powerwall. But some additional work is needed for the Megapack to achieve good margins"
Of course, storage was never a SolarCity product, and could have been sold by Tesla without needing to spend over six billion dollars (including assumed debt) for its purchase of the company. In particular, Tesla's utility scale storage products have absolutely no relationship to anything SolarCity ever sold.
It is worth noting that I came to the Tesla story via SolarCity prior to its acquisition. It was clear to me then that SolarCity was in serious financial trouble despite the fact that much of its financial reporting was not transparent, and in many cases likely resulted in investors being misled.
In the subsequent four and a half years, all of my misgivings about SolarCity's financial condition and lack of transparency have been confirmed. Simply taking the reported Gross Profit for EG&S and making reasonable estimates of SG&A, interest expense, minimal R&D, required payments to the non-controlling interest investors, required principal payments, etc., there is no-way that SolarCity could have ever survived on its own.
Furthermore, not long after the acquisition, Tesla had to basically shut down its solar leasing business to avoid bankruptcy itself and cannot make any money in Buffalo despite having essentially a "free" factory. It has only been saved from $41 million in annual penalty payments as a result of negotiated changes to the original deal, changes which many, including the NYS controller, consider more than a bit questionable. The acquisition was undoubtedly a bail-out.
I believe that information provided by SolarCity to investors continues to risk misleading investors, as exemplified by the annual 10-K language I've cited above regarding the company's progress on the solar roofs. There are other aspects of Tesla Energy's financial performance I've discussed in the past, some of which I've referenced in this article, that raise serious questions about both its performance and the transparency of its financial information. There are some others issues yet to be addressed.
Some of my previous articles have touched upon not just the former SolarCity, but also Tesla's automotive activity, including its warranty reserves and its leasing. I frankly find the myth around Tesla vehicles not depreciating more than a bit humorous, as even a casual examination of the financial statements indicates otherwise; Tesla is able (required actually) to place its leased automobiles on the books at cost rather than market value (absolutely appropriate and GAAP compliant), but then still charges large quarterly depreciation expenses against them.
At lease-end, Tesla disposes of the leased cars which is part of the "Services Segment" gross profit/gross loss, which is ALWAYS a loss. If Tesla is selling these vehicles above depreciated book value, it can't be by much.
I am not at all a member of TSLAQ ("Q" is the letter added to a ticker after a company has filed for bankruptcy, for those who might not know) and in fact I ridicule some of the TSLAQ crowd at times. I even write "bankruptcy puts" (sometimes even valuation puts) against my short. In fact, as stated in my disclosure, I am long the TSLA 2025 bonds, which I consider "money good" and actually expect to see them redeemed early.
I simply see TSLA as absurdly overvalued; its current market cap is as much as most other auto manufacturers combined, which makes no sense to me. In addition, I sense a bit of "dry-rot" behind TSLA's wall when it comes to financial reporting and full disclosure.
A goal of mine is to raise the level of discourse here. I am frankly surprised that despite all of the articles on Seeking Alpha and in the general press, I always find topics that have not been sufficiently explored, which is the only reason I write articles here; I'm not simply going to rehash old news and analysis. I am admittedly fascinated by the Tesla story, and this lack of factual objective analysis, leaving "niches" available to me, is the aspect that has me most hooked.
Another aspect of this is the constant debate as to whether Tesla is making money selling vehicles, which usually devolves into discussions about regulatory credits. To my knowledge, no one has ever tried to look at Automotive vs. Energy to estimate where Tesla is making vs. losing money. Since I view Tesla Energy as a disaster, automotive might not look so bad, but since I have some reservations with the accounting there as well, I will never make that case.
I therefore challenge a few of you, preferably at least one bull and one bear, to try to separate those two businesses and make a case either for or against Tesla Automotive as well as Tesla Energy.
This article was written by
Disclosure: I am/we are short TSLA. 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.
Additional disclosure: I have a hedged short position in TSLA stock but am long the 2025 bonds.