[Editor's Note: The author has provided a brief update at the end of the article.]
One of the Tesla (NASDAQ: TSLA) metrics that receives the most attention are quarterly delivery numbers as that appears to be a way of both possibly tracking overall demand and also, given the initial manufacturing problems for the Model X, a way of monitoring progress on Tesla's manufacturing issues. There are a lot of ways to "manage" quarterly delivery numbers, however, where the details of what Tesla may be doing have little visibility.
Tesla's overall financial reporting also provides very little visibility about anything given the wild swings in quarterly gross margins, large and unpredictable increases in other expense categories, and a very complicated and also unforecastable balance sheet from quarter to quarter. A somewhat arcane portion of the balance sheet are the "Resale Value Guarantees" (RVG) accounts which have both a short-term portion (of guarantees that have less than one year before expiration) and a long-term portion (for the portion of guarantees that have more than one year before expiration).
The amounts listed in these accounts are also quite large and were almost $1.8 billion at the end of the first quarter (using Muskian precision!) and may be seen on page 4 of the 2016 Q1 10-Q filing. Since the accounts are on the liability side of the balance sheet, the full amounts in the accounts may also appear to be an additional layer of debt. Another SA contributor (Bryce_in_TX) wrote a helpful blog post in May describing the accounting behind the accounts to clarify that the full amounts listed are not debt to be repaid such as any short-term or long-term debt whose full amount listed is actually owed to creditors.
Bryce also described, however, that accounting is based on assumptions when there are future events whose outcome is not yet certain or known and that is actually the critical factor in monitoring and analyzing the balances in the RVG accounts. The reason the RVG accounts are important for monitoring ongoing deliveries is that for a company that claims to spend no money on advertising, marketing, or promotion, analyzing the RVG accounts can provide insight into one of Tesla's very widely used but little recognized promotional activities to drive delivery numbers.
Although the accounting behind the RVG accounts is somewhat complicated, the basic concept is simple: Tesla is guaranteeing resale values to "purchasers" of its vehicles and its financial statements need some way to track how all the money may flow in such transactions within standard accounting procedures. The reason that I have put quotation marks around the word purchaser is that effectively, when an owner has the right to return the vehicle, the vehicle has not actually been purchased but is essentially being leased. That is why there is also a very rapidly growing account on the asset side of Tesla's balance sheet with a title of "Operating Lease Vehicles, Net."
The standard Tesla lease is for between 36 and 48 months and since periodic accounting is based on ongoing assumptions that result in what are typically monthly accrual entries based on future projections, Tesla needs to estimate what will be the value of its vehicles after 36 to 48 months of use. For a relatively new company with an evolving base of experience with what before the Model X was just one model, such assumptions may still be uncertain at this point - especially since we are only now starting to see the first wave of Model S vehicles with leases expiring. Those expiring leases will both trigger the RVG provisions in each lease and start testing assumptions about actual resale values in the used vehicle market.
An interesting detail in the notes to Tesla's financial statements (on page 53 of the 2015 10-K) is that the assumptions behind the RVG amounts are only analyzed annually! So, even in the best case concerning the possible future value of used Tesla vehicles, for the company's financial reporting a whole year could go by before Tesla is willing to incorporate real world experience from what is happening in the marketplace for its used vehicles.
There are some interesting trends that can be seen on Tesla's ongoing quarterly balance sheets (which all may be accessed through the SEC's EDGAR system) which shows that the internal RVG assumptions are actually changing very rapidly on a quarterly basis - and those assumptions are also probably a prime driver of Tesla's delivery numbers each quarter.
As Bryce_in_TX has stated earlier in his post about Tesla's accounting for the RVG accounts, the entire amounts are not debt for Tesla to repay. What will possibly trigger future payments are differences in Tesla's original assumptions for resale values at the start of the lease and the actual vehicle values at the end of each lease.
If the future resale values of leased vehicles are less than the assumed resale value in the original contract, there will be contingent liabilities (and cash payments) for a portion of the amounts listed in the RVG accounts. Why this is important may be seen in the following spreadsheet, which also shows how rapidly the RVG assumptions are apparently changing each quarter:
As you can see in the spreadsheet, there is another item also listed that is from the Deferred Revenue accounts on the balance sheet. The reason why this is important, as Bryce has already described in his blog post in May, Deferred Revenue is the residual amount originally booked that is the difference between the original list price of the vehicle and the resale value guaranteed in each lease contract.
I don't intend to duplicate Bryce's already excellent description of the accounting behind these leases, but I think the following example is interesting showing the actual accounting entries behind various assumptions and provisions that may be in Tesla's leasing contracts. The following spreadsheet shows the entries behind different assumptions for RVG percentages:
A possible seemingly intuitive interpretation of the numbers above is that a higher RVG percentage may even be considered "conservative" for revenue recognition as it results in a lower deferred revenue amount and then lowers periodic amounts included in revenues throughout the lease term. Given Musk's mumblings on conference calls about the difference between GAAP and non-GAAP numbers, someone might even think the company is being somehow short-changed by such accounting treatment. A very short answer to the idea that Tesla is being short-changed in all of this is that it is clear in the example given that Tesla is receiving the full purchase price of the vehicle up front in cash.
The reason that a lower deferred revenue portion is also not "conservative" is that a very uncertain event 36 to 48 months later is behind the RVG amount. As you can also see from the "q1 increase" column in the first spreadsheet above, the apparent RVG percentages for new leases in the first quarter of 2016 was AT LEAST 70 PERCENT. The reason that I say it was at least 70 percent is that older leases with lower RVG assumptions would have been expiring during the quarter and so the calculation of a percentage of the changing balances between Q4 2015 and Q1 2016 would actually be slightly skewed below what the RVGs may have been on new leases during the quarter.
Tesla also doesn't provide any help about such activities with its completely inadequate financial disclosures about its leasing programs and so insights into such programs have to proceed through analyzing balance sheet accounts. From those accounts, a combination of understanding the basic accounting entries for each lease and basic math results in the glaring fact that new leases in Q1 2016 had to have had an RVG of at least 70 percent. As you can also see from my first spreadsheet, the relationship between RVG balances and Deferred Revenue balances has also steadily increased from 50 percent in Q4 2014 to 64 percent in Q1 2016, and that is definitely not "conservative."
As I have also described above, in the notes following Tesla's financial statements, they describe that they review RVG balances annually. As you can see from my spreadsheet, however, those percentages are rapidly changing on a quarterly basis. Used vehicle prices are also continually changing (particularly in cars with whompy wheels or for people who supposedly live down long dirt roads!) and so reviewing RVG assumptions annually is completely inadequate.
The increasing quarterly trends also deserve a huge amount of scrutiny given what can already be seen in Tesla used vehicle prices. This morning I used Kelley Blue Book to get an estimated value of a 2013 Model S 85 that with options probably had an original price of at least $90K (current Model S pricing may be seen on the Tesla design page). The Kelley Blue Book value for that car with an assumption of 36,000 miles was $50K, or 55 percent of its original value. As such, original leases that were apparently written with RVGs of less than 50 percent were probably somewhat conservative. Now that the RVG percentage of the entire portfolio of leased vehicles was 64 percent by the end of Q1 2016, however, and that new contracts in Q1 were apparently written with RVG assumptions of at least 70 percent, those assumptions are probably now overly aggressive.
How this affects deliveries
Here's why all this matters concerning deliveries. As Jeb Bush once notably said "people like free stuff" and so, just think of this deal:
"Buy" a new Tesla and after 36 months we will guarantee that you can sell it back to us for 70 percent of what you paid for it!
Since Tesla also prides itself in selling its cars to sophisticated owners, you can just imagine the number of people tempted by such a deal!
Possible Accounting Problems from RVG Assumptions
The problem with all that is that at the end of the lease (or guarantee period), the actual values need to be reconciled in Tesla's financial statements. The following spreadsheet also shows possible payments due to owners or leasing partners depending on various assumptions for resale values at the end of leases based on the reported balances at the end of Q1 2016:
The numbers you see on the "Change from Q1 RVG balance" line are possible future write-offs that would be recognized as additional expenses if the actual future values of used vehicles have the possible RVG values listed above in the table.
Although a possible future write-off of the amounts listed above (which Tesla probably wouldn't report in the non-GAAP numbers!) would be considered a rounding error for a company whose CEO has now predicted that it will have a $1 trillion market value and which will likely burn through $2.8 billion of cash in just this fiscal year, there is a more important principle which is overall financial statement integrity. If your assumptions are wrong, your financial statements will be wrong and that is misleading to investors.
In the sleight of hand approach to its overall financial reporting there are also lots of other things that are inadequately reported and need more detail but in the example below shows that the RVG assumptions could result in an even higher level of write-offs in the future:
Since new leases with RVGs written in Q1 2016 apparently assumed at least a resale value of at least 70 percent of the original purchase price, if the future value of Tesla vehicles are actually 55 percent, that would result in a $75 million write-off just for the leases originated in Q1. Although that number doesn't sound that large relative to such a colossal world-changing enterprise which just collected another $1.4 billion from gullible investors, that annualizes to being $300 million of annual write-offs from overly aggressive RVG assumptions - and that figure is close to five percent of total projected revenues for 2016 which definitely makes such an amount very material.
Additional Details from Financial Statement Notes
All of this would be helped if Tesla would provide more detail about its programs and transactions that result in the operating lease arrangements and assumptions for RVGs over the operating lease portfolio but that is not the case. There is a curious disclosure in its financial statement notes (also on page 53 of the 2015 10-K) that describes that the accounting and financial reporting for operating leases is different depending on whether the lease is through what is described as "commercial banking partner" which finances the purchase directly by the customer or a "bank leasing partner" which directly buys the vehicle from Tesla and leases it to the customer.
For the purchases directly by vehicle owners but which are financed through a "commercial banking partner", both the deferred revenue and RVG balances are amortized on a straight-line basis over the life of the financing period. As such, there is an ongoing "matching" of reported revenues from deferred revenues and a declining RVG balance in line with the projected path toward an ultimate RVG balance for each vehicle at the end of the financing period.
For the lease where a "bank leasing partner" is buying the vehicle from Tesla and then leasing it to a customer there is a strange quirk in their reporting in that the deferred revenue portion of the transaction is classified as a short-term liability, which is generally recognized in one year but that the RVG portion of the transaction is classified as a long-term liability to be amortized over the life of the lease term. There is then some other possible contradictory language that says that the deferred revenues are "accreted" over the life of the guarantee period but that contradicts placing the entire deferred revenue balance in the current liabilities listed on the balance. Again, however, there is no further detail which would be helpful as to revenues recognized each period from such arrangements.
If the deferred revenue portion of the "bank leasing partner" transactions is treated as a normal short-term liability and amortized over one year, however, that then creates an accelerated and asymmetrical method of recognizing revenue from the transaction. After one year, all the revenue from the transaction would be recognized but the uncertain "tail" of the RVG would still be on the balance subject to subsequent possible revisions if the RVG assumptions are incorrect.
There is another interesting detail in the Q1 10-Q (in the notes on page 10) that lends further support to my suggestion that aggressive RVG assumptions are now being used and offered to customers and that is from the disclosure that at the end of Q1, the deferred revenue portion of the "bank leasing partner" transactions was $179 million and that the RVG portion was $716 million. As such, the RVG portion of the combined balances was 80 percent at the end of Q1, which is even a lot higher than the imputed Q1 RVG percentages for new leases originated during the quarter of 70 percent.
As I've already commented, there is no disclosure of the actual terms and conditions of any of the bank financing arrangements or external leasing programs in Tesla's SEC filings. Given that $1.8 billion is now listed as a liability on the balance sheet, there is no excuse for not having more detail as such an amount would definitely be considered material.
The volume of business through such external financing or leasing partners would also be considered material and such arrangements and agreements are actually required to be listed as exhibits in 10-Q or 10-K filings or even in an 8-K if a significant new externally financed leasing program is put in place. As with all of Tesla's financial filings and disclosures, however, this is another area where they are pushing the envelope in deciding on "materiality" or in using the "confidential treatment" loophole offered by the SEC. Maybe the SEC staff that previously was responsible for monitoring Madoff's operations has been reassigned to monitoring Tesla. That is not that far-fetched as speculation given the travesty of the recent ill-fated secondary offering that was shoved down buyers throats two days after filing while using the "shelf registration" loophole in the SEC procedures.
In any case, although the full balance of the Deferred Revenues and RVG will never have to be repaid and could be "misleading" to someone who doesn't understand accounting, I've provided some examples of possible contingent payments in the future if the RVG assumptions are incorrect. Those possible payments would likely grow at a much faster rate than revenues, however, as vehicle operating leases are growing much more rapidly than overall revenues.
Depending on the ultimate value of the cars at the end of the operating leases, Tesla may also be misstating both sides of the balance sheet as well as expenses on its ongoing income statements. The asset side of the balance sheet would be misstated as the account "Operating Lease Vehicles, Net" is being amortized based on current assumptions about future resale values. As such, if the future resale values are less than the current assumptions for resale values, then Tesla will have understated the amortization of the leased vehicles over the lease term.
There is also another dissembling aspect of all of this on the part of the company and that is the convoluted and incoherent comments about non-GAAP and GAAP financial reporting. Poor Elon who bemoans that the company is not allowed to report all the "revenues" from their deliveries each quarter also seems to imply that there will be some pot of gold at the end of the rainbow when the operating leases run their full course. As I've shown in an earlier example of the accounting entries, that is not the case as ALL of the cash that Tesla will ever receive from these transactions have already been paid to the company by either the commercial banking partner or bank leasing partner. Instead of a pot of gold, investors are likely to actually receive a lot of very large lumps of coal if the RVG assumptions are unrealistic and too aggressive.
If everything else was right with Tesla, someone might view the amounts discussed here as minutiae and immaterial given their supposedly noble mission but this is a company which only exists based on a continual flow of new financing. Don't also be misled by Musk's comments about being "cash flow" positive as he apparently doesn't include things such as working capital growth, asset growth of vehicles under operating lease, and capital spending - all of which are required to support the company's growth plans - as using cash! My own financial model of the company does project that there will be around $300 million of positive operating cash flow in 2016 but, with $2.2 billion of capital spending and probably at least a $1 billion increase in vehicles under operating leases, the overall cash burn this year will probably be around $2.8 billion.
With the recently announced acceleration of production capacity to support the Model 3 launch and using standard metrics for productivity and throughput for automotive manufacturing equipment, the additional CapEx requirements for quintupling their production capacity will also result in probably needing another $6 billion in external financing in 2017 and 2018 to achieve their growth plans. The problems with "minutiae" such as RVG assumptions that could possibly trigger hundreds of millions of dollars in contingent liability payments over future years, is that such payments could also affect Tesla's ability to raise further financing or affect the cost of its bank financing or leasing programs.
I guess Musk may either be too arrogant to care about the details of all this or maybe he doesn't even understand financial statements (those are things for the little people!) but this week's outrageous offer to acquire SolarCity (SCTY) shows how quickly all of this could unravel given both companies needing continual heroin injections of external financing. Maybe Musk is oblivious to SolarCity debt now being priced at 20 percent yields but the current relationships with what are essential external financing entities to take the other side of purchases and leases could be threatened by a perception that the risks of Tesla debt and other financial obligations, including possible future contingent RVG payments, are now too great to continue the current financing programs.
As I said at the beginning of this article, the rapidly growing number of vehicles being delivered through the leasing and partner financing programs, promoted by what appears to be increasingly aggressive RVG assumptions, is probably a significant factor in the current number of vehicles being delivered to customers. As everything about Tesla is opaque and needs far more disclosure, I don't even want to make a guess as to what portion of current deliveries are being driven by high RVG offers but people do recognize a good deal when they see it!
All of this also does remind me of another interesting situation that I once analyzed which was when I noticed that WorldCom's gross margin percentage had been essentially unchanged during an eight-quarter period when there were brutal price wars and also a changing business mix to lower margin services. I then analyzed balance sheet accounts and saw that Other Assets had been growing by 20 to 25 percent annually over the previous two years although overall revenue growth had only been about ten percent during that period of time.
WorldCom's creative accounting during that period was to capitalize customer discounts as "marketing expenses" and put them on the balance sheet as other assets given the supposed value of their long-term customer relationships. A parallel with Tesla is that although the entire nominal amount of the RVGs are shown on the balance sheet, and also then essentially disregarded, what will be important are the potential contingent liabilities to be recognized as both future expenses and significant cash payments. Since we don't know what any of the terms or assumptions are behind the RVGs, we also don't know how much is possibly being hidden on the balance sheet similar to WorldCom hiding expenses on their balance sheet.
On one final note, I'm also stuck in this situation by Musk's dissembling pattern of selective disclosures and dismissive responses to valid questions on conference calls. "Yeah, you just don't get it" is pretty much how he responds to any attempt to ask detailed questions. The conference call after the SolarCity announcement was particularly appalling - especially when Musk either had no details at all about possible financial effects of the merger (they don't matter since I have still have dumb people to finance all of this!) or effectively didn't respond to a lot of questions.
Such a conference call brought back amusing memories of listening to Ken Lay's courtly dismissals of a lot of questions about Enron and Jeffrey Skilling's arrogance in suggesting that dumb questions were being asked. Since Elon is so into efficiency (eight years after becoming CEO, he now recognizes that he is only using two to three percent of his factory space when measured volumetrically!), maybe he is just trying to combine the characteristics of both Lay and Skilling.
There have been comments that there are various sources on Tesla's website that mention a Resale Value Guarantee of 50 percent - not the higher numbers discussed in my article.
The 50 percent figure that you see for the fourth quarter of 2014 in my table in the article is also in line with previous information from Tesla which does mention a Resale Value Guarantee of 50 percent. The examples on Tesla's website at this point are a link for just one variation of the Model S (the 60 from 2013) which shows a form mentioning the 50 percent guarantee and an obscure page that is not findable using any search function on Tesla's site which also lists a 50 percent guarantee.
The form from 2013 is no longer a relevant data point as the company has now had six different variations of the Model S since 2013 and that form is also from at least one year before the time period described in this article. The obscure page on Tesla's website which mentions a 50 percent RVG also has additional terms and conditions in that it is only applicable for "qualifying offers" which I would interpret as meaning as a trade-in value for the purchase of a new Tesla vehicle. With such conditions, that is not really a guarantee.
In any case, Tesla's financial reporting is the source of such uncertainties about actual average RVG percentages from each separate financing program (loans and leases) as they are unwilling to make any disclosures about the percentages for each program in their financial statement notes.
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.