Tesla's Surprise Quarter: A Case Of Accounting Adjustments, Not Operational Improvements

Nov. 06, 2019 12:57 PM ETTesla, Inc. (TSLA)1.36K Comments66 Likes
Sunil Shah profile picture
Sunil Shah


  • Although several articles have questioned the operating miracle in Tesla’s September results, the details just released in the 10Q form shed more light, especially into the "how."
  • Through an investigation of the three 2019 10Qs and the reserves movements, one can distinguish how much of the profit jump stemmed from operational improvements or from accounting adjustments.
  • This article demonstrates the profit jump was far more due to revisions of reserve balances than an improvement in operations. Sell Tesla while the delusion persists.

Tesla's (NASDAQ:NASDAQ:TSLA) third quarter results certainly caught the street by surprise. From a consensus expectation of a loss at $0.23 per share, the profit of $1.86 was the very definition of an earnings beat.

This article posits the results were far more a result of accounting revisions (reserves and used-vehicle values) rather than a genuine improvement in operating profitability. No Tesla bear needs reminding the share price reflected the earnings beat. The market's misinterpretation offers a valuable shorting opportunity.

During the September 3Q 19 conference call, the CFO, Zachary Kirkhorn described the quarter (my bold italics):

Yeah, thank you, Elon. Q3 was a great quarter for Tesla. We've made terrific progress, and yet again, we realized margin improvements in nearly every aspect of the business. There are three key points I would like to highlight.

First, we returned to profitability in Q3, aided by improved gross profit, reduced operating expenses and the absence of negative one-time items that weighed on our financials in the first half of the year. GAAP automotive gross margin improved sequentially to 22.8%.

A number of articles by SA fellow contributors have questioned the results, but these authors did not have the benefit of the quarterly report which divulged detailed reserve movements and enabled a deeper investigation.

This article focuses on the profit statement up to the gross profit level, where the margin jump and most revisions in the measurement of accrual charges were made. The adjustments that boosted results occurred in three categories:

  • Estimation of the value of closing stock, namely the value of used vehicles for resale.
  • Under-providing – when compared to the first two quarters - for an impairment in the value of "vehicles repurchased" in the buyback program, (linked to closing stock).
  • Under-providing (again, when compared to the first two quarters) for warranty expenses for repairs.


It’s important to note that these adjustments may be reasonable – management is entirely justified in making these subjective revaluations, if changing circumstances so warrant. However, I cannot but notice a conspicuous lack of both, a rationale and an elaboration, in the presentation of the results.

Closing Stock

For those not versed in accounting, this is a basic computation of how gross profit is calculated for a period, be it a quarter or a year.



Cost of Sales:

Opening Stock




Closing Stock


Cost of Sales:



Gross Profit


As a Chartered Accountant who spent many years auditing companies, I learned the easiest way to fine-tune the gross profit is by adjusting the Closing stock. The figure is unique with regard to how it may be audited: Most metrics have a counterpart which can be investigated (e.g a sale would have a counterpart in debtors or cash received for the sale). But the closing stock is a balance that's booked directly to the balance sheet and profit statements (cf a counterpart entry), and apart from a physical count (mostly done on a samples basis at year-end and hence prone to imprecision), it’s the easiest way for a company to "fine-tune" profits. This is especially so between quarters (note quarterly results are never audited, just the year) while ensuring the closing stock account is accurate for the year-end.

My curiosity on Tesla’s results first arose on examination of the 10K submission. It seemed unusual that more cars were sold than produced (845 vehicles), but the closing inventory was higher than at the quarter’s start. I performed a qualitative check on production/deliveries of vehicles in each quarter of 2019 (Q1,Q2 and Q3 data), and compared this with the change in the inventory balance. As seen from the table below, for Q1 and Q2 the inventory change (magnitude and sign) accords with the net addition/depletion of vehicles. However the vehicle change in Q3 (845 more vehicles sold than produced) - which should lower closing stock - actually led to an increase of $199m in inventory.

Qualitative Check on Inventory





77 100

87 048

96 155


63 000

95 200

97 000

Add/Reduce vehicles

14 100

-8 152


Value of Inventory close $m

3 837

3 382

3 581

Value of Inventory open $'m

3 113

3 873

3 382

Change in inventory $m




$ Change Reflects Vehicle +/-




But it’s only on analysis of the 10Q form published on October 29 that alarm bells starting ringing. The table shows the detailed inventory account at the end of each quarter in 2019.

September 30,


June 30,


March 31,


Raw materials




Work in process




Finished goods




Service parts








The balances for "Raw materials," "Work in process" and "Service parts" seem fine. However my auditor’s eye was drawn to the balance for "Finished goods."

Let me explain why. This balance aggregates both new vehicles ready for sale AND used vehicles that have been repurchased, either for resale or those employed in the showrooms. As Tesla produces primarily to order, the balance comprises mainly of used vehicles for resale that stem from Tesla’s buyback program. From the detail in the 10Q page 10, we learn about the buyback options program:

At the time of revenue recognition, we reduce the transaction price and record a sales return reserve against revenue for estimated variable consideration related to future product returns. Such estimates are based on historical experience. On a quarterly basis, we assess the estimated market values of vehicles under our buyback options program to determine whether there will be changes to future product returns. As we accumulate more data related to the buyback values of our vehicles or as market conditions change, there may be material changes to their estimated values. Due to price adjustments we made to our vehicle offerings during the first half of 2019, we estimated that there is a greater likelihood that customers will exercise their buyback options that were provided prior to such adjustments. As a result, along with the estimated variable consideration related to normal future product returns for vehicles sold under the buyback options program, we adjusted our sales return reserve on vehicles previously sold under our buyback options program. The net impact was $104 million reduction in gross profit for the nine months ended September 30, 2019.

To simplify the above, Tesla buys back vehicles under their buyback options program. These are held as used vehicles for resale, but their value is adjusted to estimate the fair realisable value. Any downward adjustment would equate to a reduction in closing stock, hence a reduction in gross profit. Conversely, any upward adjustment would translate into an increase in gross profit. We also learn that Tesla registered a net impact of $104 million reduction in gross profit for the nine months ended September 30, 2019.

Note this total for nine months is all that’s disclosed deep in the 10Q.

It's only on examination of the net charge taken in the two previous 10Qs that one may deduce how that $104m charge was split into the three quarters.

Quarter 1 $ 91.7m

Quarter 2 $ 14.5m

Quarter 3 -$ 2.2m

9 months total $104m

I question the charge taken in Q3, both its magnitude the its presentation in the September results.

There's no explanation which warrants an upward revision to used vehicles. In fact one could easily argue that the mass production of Model 3s would reduce the appeal/value of used Models S/X within the buyback program and this charge should logically rise.

As for disclosure, the September 10Q simply notes the cumulative charge for the nine months, not the reversal by 2.2m in Q3. I can only term this disingenuous.

(Update, November 14, 2019, 9:30 a.m.: The following section has been added in response to reader comments and questions.)

Does Tesla Revalue Inventory to boost profit?

This subject has invoked much debate and is clearly a grey area. The crux of the controversy is that accounting standards do not permit upward revaluation of inventory to boost gross profit. This, of course, is true, and Tesla does not revalue existing inventory. However, the company has revalued future inventory it may be obligated to purchase through their vehicle buyback option program. That reassessment of the future value or bought-back vehicles, versus the current book value, passes through the profit/loss account, It did so in Q3, boosting quarterly gross profit.

This does indeed obviate the standard. The revaluation is done by reducing the future purchase inventory guarantee reserve; the contra credit boosts gross profit. In essence, Tesla re-appraises the realisable value of cars they will be obliged to purchase when a Tesla owner exercises his or her buyback option. The difference between the buyback guarantee and Tesla's assessment of the prevailing realisable value is released to gross profit as an adjustment -- both up and down.

This is precisely what transpired in Q3, where gross profit was boosted by $2.2m, as shown in the three 10Qs for 2019 to date. Q1, Q2, and Q3 are shown as a cumulative nine-month charge of $104m, whereas Q1 and Q2 aggregated to $106m.

(End of update)

If Tesla had pro-rated the $104m charge equally into the three quarters, it would result in a reduction of gross profit of $37m (a charge of $34.6m rather than an add-back of $2.2m). Given the nature in accounting for closing stock and the difficulty in auditing it, Tesla’s treatment/reporting of each quarter is hardly likely to be investigated. Remember, as long as management ensure that the final balance is correct when the auditors visit at year-end, the issue poses no problem.

Warranty Reduction

To those not accounting savvy, Tesla sells a large portion of its cars under warranty, where most repairs are not charged to the Tesla owner but covered by the company under warranty. To account for future potential charges, a warranty provision or reserve is created. Actual repair costs are then deducted, as incurred, to this reserve. Naturally this requires an estimate to be made for the provision, as there will obviously be time lags between the creation of a warranty (covering say, next five years) and when the actual expense is incurred. The timing gap is matched by adjusting the warranty reserve account.

As per Tesla’s 10Q:

We accrue a warranty reserve for the products sold by us, which includes our best estimate of the projected costs to repair or replace items under warranties and recalls when identified. These estimates are based on actual claims incurred to date and an estimate of the nature, frequency and costs of future claims. These estimates are inherently uncertain given our relatively short history of sales, and changes to our historical or projected warranty experience may cause material changes to the warranty reserve in the future.

Rather than rewrite the treatment, I paste the analysis of Barron’s accounting expert Bob Willens, who’s covered the subject thoroughly.

Tesla recognized $153 million in warranty expense for the second quarter of 2019. It sold $5.2 billion worth of cars, so warranty costs represented about 2.9% of that revenue, but the ratio fell to about 2.7% for the third quarter.

Is this a big deal? If the warranty-expense rate hadn’t changed, Tesla’s pre-tax income would have been about $12 million lower. Tesla reported a pre-tax profit of $176 million during the second quarter.

Tesla also reduced its overall accrual for previously existing warranties. It doesn’t think it will have to provide as many warranty repairs as it once believed. It added another $37 million in third-quarter pretax income.

The cumulative impact of these warranty-related revaluations resulted in a boost of $50m to profit. That's material, compared with the $143 million profit Tesla reported.

To reinforce my argument, reserve balance may be viewed from another perspective. For each quarter in 2019 a roughly similar number of vehicles were sold. So each quarter resulted in a similar number of new vehicles covered under warranty. As seen in the two tables below that are drawn from the 10Qs, the reserves accruals account was raised by $96m and $98m in the first and second quarters, respectively.

One MUST ask why it increased by a mere $42m in the third quarter. Even if new data gleaned from the quarter supported a downward revision, in the interests of accurate reporting, is Tesla not obliged to highlight it? The downward revision of an expense reserve can scarcely be termed an "operating improvement."

September Quarter

June Quarter


Again I must state that these accounting changes - if reflective of reality - are a fiduciary management responsibility. Yet I question if underlying circumstances have changed to warrant them. For instance, surely the realizable value of used model S/Xs that Tesla repurchased would decline after the mass entry of Model 3, a cheaper and highly-desired model? Why would Tesla reverse an impairment charge, rather than raising it?

I firmly believe Tesla’s portrayal of the quarter is at best inaccurate, at worst a misleading representation.

I urge you to revisit the CFO’s opening summary for the quarter in the CC, copied at the top of article.

I believe a fairer and more accurate version would be:

We've made good progress. We returned to profitability in Q3, in order of importance by:

  • A revaluation in our warranty repairs accrual which resulted in a lower charge, when compared to the first two quarters
  • A lower impairment for our used-vehicles value that we hold for resale. For the previous quarters this led to a reduction in profit of $92m and $15m in Q1 and Q2 respectively, whereas in this quarter the impairment adjustment boosted profit by $2m
  • Note the above items are not one-time charges but a regular part of our business
  • And lastly in order of importance, by a gross margin improvement as well as reduced operating expenses

I have little doubt that Tesla spiked due to the impressive earnings beat, a surprise the market construed as a landmark to an invigorated path of rising operating profitability. Alas, downward adjustments in reserves are far from sustainable, they're limited by their very nature. The market's misinterpretation of the quarter, the 30% spike, offers the opportunistic investor a great short. Sell Tesla at these prices while you can.

This article was written by

Sunil Shah profile picture
A fund manager who cut his cloth in Schroders London. He joined Coronation South Africa in 1998, running the Smaller Companies Fund which had the best 5-yr record in the sector during his tenure. In 2005 he left Coronation to pursue his passion in writing (and invest without constraints). He recently completed his first novel, a financial thriller called "White Man's Numbers" see www.whitemansnumbers.com - highly acclaimed rollercoaster! See excerpt and reviews on website link.

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.

To ensure this doesn’t happen in the future, please enable Javascript and cookies in your browser.
Is this happening to you frequently? Please report it on our feedback forum.
If you have an ad-blocker enabled you may be blocked from proceeding. Please disable your ad-blocker and refresh.