During Tesla’s last conference call, concerning the first quarter of 2018, Mr. Deepak Ahuja, Tesla’s Chief Financial Officer (CFO), while answering a question from Rod Lache, an analyst at Deutsche Bank Securities, Inc., said:
“Rod, we are very CapEx-efficient, overall. Let me just start from that point. And if we look at our depreciation costs on a per unit basis at steady run rate of 5,000 or so cars per week, we are in my mind well below most of our competitors – well below $2,000 per unit depreciation cost.”
I was blown away by this statement. Here in a public meeting was the CFO of a manufacturing company stating publicly and officially the actual cost of capital per unit of production of his company at a specific level of production, and to make it even better, he had just compared his cost of capital against the entire industry. This piqued my curiosity. The numbers just sounded too good. He had just opened up Tesla’s curtain enough to give me a chance to peek inside.
Recently, I explored the necessary level of Model 3 car sales and the corresponding number of vehicles for Tesla to break even. I presented three different break-even levels. These varied by average selling prices of all three models and the volume of Model 3 produced (the volume for Model S and Model X was kept constant).
However, these numbers were explored at the aggregate level. That is to say, I took the numbers from Tesla’s financial statements without further examining them. For instance, I took the quarterly aggregate revenue numbers from the income statement and I worked with them. But how did these numbers get there?
When I presented the third simulation, I made the remark that:
At these higher volumes, the price constraint on Model 3 becomes less relevant. In fact, all three models could be sold at the lowest permissible prices — or, looking at it from another perspective, higher prices will bring in profits.
Now, I would like to explore breaking up some of these numbers, and perhaps it might be of interest to take a look again into Tesla’s financials and their relation to depreciation.
What is depreciation? Investopedia defines depreciation as “an accounting method of allocating the cost of a tangible asset over its useful life. Businesses depreciate long-term assets for both tax and accounting purposes. For tax purposes, businesses can deduct the cost of the tangible assets they purchase as business expenses; however, businesses must depreciate these assets in accordance with IRS rules about how and when the deduction may be taken.”
BusinessDictionary provides four definitions for this term:
1. Accounting: The gradual conversion of the cost of a tangible capital asset or fixed asset into an operational expense (called depreciation expense) over the asset’s estimated useful life. The objectives of computing depreciation are to
1. reflect reduction in the book value of the asset due to obsolescence or wear and tear,
2. spread a large expenditure (purchase price of the asset) proportionately over a fixed period to match revenue received from it, and
3. reduce the taxable income by charging the amount of depreciation against the company’s total income. In effect, charging of depreciation means the recovery of invested capital, by gradual sale of the asset over the years during which output or services are received from it.
Depreciation is computed at the end of an accounting period (usually a year), using a method best suited to the particular asset. When applied to intangible assets, the preferred term is amortization.
2. Commerce: The decline in the market value of an asset.
3. Economics: The decrease in the economic potential of an asset over its productive or useful life.
4. Foreign exchange: The reduction in the exchange value of a currency, either by a government or due to weakening of the underlying economy in a floating exchange rate system.
From a practical everyday point of view, depreciation, for me, accounts for the time when I need to start looking into replacing an asset. For instance, my car might be working fine for over 10 years, but the number of times I have gone through a whole set of tires starts to tell me that if so many sets of tires have been worn out in the same car, then perhaps I should start looking into trading in my car. In this regard, accounting for time and use provides me with a valuable tool to determine when to change something.
How does Tesla account for time and use of its assets, and can this accounting have an effect on profits? The following is Tesla’s way of succinctly summarizing this:
As part of this summary, it shows a systematic way of accounting for the historical wear and tear under the line item called accumulated depreciation and amortization.
Now, in my opinion, there are three possible ways of accounting for this wear and tear:
1. Just right
2. Too little
3. Too much
In addition, I need to know how to account for it. Tesla provides a summary of this as well. Using the information from these tables (see above and below), I constructed the following to see how much depreciation needs to be added to the cost of each car by applying the maximum possible number of useful years for each asset class:
The logic behind this is that each car requires three basic things: materials, labor, and capital. Capital, in this case, for example, is the land where Tesla’s factories are built. Capital is also the use of robots for painting, assembling, welding, etc. However, Tesla’s cost of capital is massive: $12,511,744,000. It cannot all be used up at once. Rather, a little portion of it is “used up” in every car.
To extract a portion of this large cost of capital, I used Tesla’s statement, “Depreciation is generally computed using the straight-line method over the estimated useful lives of the respective assets, as follows:”
In addition, Tesla uses the following rules to determine depreciation (see below, following a few more of my notes):
I got a large number for use of capital, $888,661,100, from this exercise. This is the amount of capital used per year with an average useful asset life of 14 years (ignore the other value in the table). If I divide this by all the cars produced in a year, then I will get an idea of how much capital is used per car.
Please note that this number includes the assets for the other divisions or segments of Tesla, such as the energy generation and storage segment or the automotive leasing segments, and it is not being “watered down.” If it was “watered down,” then a quick way of doing this could be based on Tesla’s revenue of between 75% to 80% due on account of the automotive segment (more on this later).
There are two reasons why this exercise (in spite of the above caveats) may still prove of use. First, the other divisions or segments are not as large as the automotive segment, and second, if the exercise proves to be successful, then this carries more validity, since it includes extra “dead weight” in the analysis. In other words, the result of applying this capital per car as such will carry the whole company in the car cost. An analogy to this would be if I was hunting in the woods and I wanted to make sure of not missing, I would use a shotgun rather than a rifle.
For the sake of consistency, I am applying similar numbers (almost the same) as in the previous “break even” article, and I am using the lowest possible average selling prices per car. In other words, I am stressing the business model as much as possible. The analysis assumes an annual rate of production of 554,688 cars. This includes all three of Tesla’s models.
The resulting amount is a cost of capital of $1,602 per car. Notice that there is a positive 5.8% difference in the cost of capital compared to the previous quarter. This is important since the depreciation from the previous quarter was smaller. The rate of change of this difference is 5.8%.
Now, $1,602 might not seem like a large number in relation to Tesla’s average selling price; however, I just accomplished allocating a capital cost of $1,602 to each car from a very large capital cost of $12,511,744,000. This is not too shabby in my humble opinion.
Tesla is on its way to produce 5,000 Model 3 cars per week by the beginning of July. It is currently producing approximately 3,000–4,000 Model 3 cars per week now. How would the unit cost of capital change when this next level of production is attained?
If I run the exercise by hand — (5,000)(13)=65,000 per quarter; (65,000)(4) + (25,000)(4)= 360,000 Model 3 cars per year; $888,661,100 / 360,000 = $2,469 — $2,469 is the cost of capital per car at a production rate of 5,000 Model 3 cars per week.
In this case, it is greater than $2,000, but as explained before, there is some “dead weight” or margin of safety in this analysis. Keep in mind that the weekly rate of 8,744 Model 3 cars per week as presented in the tables is greater than 5,000 Model 3 cars per week done by hand, and this helps to distribute the cost of capital among a larger number of cars.
If I was to “water it down” by 20%, then (5,000)(13)=65,000; (65,000)(4) + (25,000)(4) = 360,000; (($888,661,100) (80%)) / 360,000 = $1,975. With a level of 5,000 Model 3 cars per week, the cost of capital is less than $2,000 per car. Mr. Ahuja was right.
The cost of capital can, of course, be “watered down” a bit since part of these assets belong to other divisions, such as Gigafactory 2 and even a portion of Gigafactory 1. As you know, Gigafactory 2 is producing photovoltaic products capable of generating electricity and Gigafactory 1 is producing components to store electrical energy as well as lithium-ion car batteries. In any case, the “watering down” process is valid since not all of assets are involved in the manufacture of cars.
To explain this “watering down” path a bit more, when Tesla bought out SolarCity, it paid over $2.6 billion, and it took over $462 million in goodwill and $5.8 billion of solar energy system assets. These solar energy assets are not involved at all in car manufacturing. If I take only the amount paid for the assets, then this brings down Tesla’s assets by around 20%.
The following table (see below) shows the different category costs per car. It, first, shows the cost of materials as a percentage of the average selling price. Each quarter shows slightly different costs of materials per model. The latest quarter shows smaller percentages compared to previous quarter. This will later show the sensitivity of the analysis to slight changes (in this case 5%).
Next, it shows a cost of labor per each model. Model S sedan has a cost of labor of $4,000 per car, but Model 3 has a cost of labor of $1,000. Third, it shows the cost of materials based on the percentages already mentioned. This is followed by adding up the costs for labor and materials, and it shows below this the sum of each class of cost (materials, labor, and capital).
Notice the relative size of each of the cost categories in relation to the selling price. The largest is the cost of materials, followed by the cost of labor, and finally the cost of capital (not shown separately) in the case of Model S and Model X.
In the case of Model 3, this order is changed. The largest is the cost of materials followed by the cost of capital (not shown separately), and finally the cost of labor.
This discussion shows the concept of economies of scale in action. The cost of materials is the largest component in both cases. It also does not change much. With greater levels of sales, the cost of materials is not budging much. This is the variable portion of costs. The cost of capital is considered a fixed cost. Moreover, with higher sales, less cost of capital is allocated to each car. This is what economies of scale is all about.
The cost of materials in this model, by the way, is the result of knowing (as much as it is possible for an outsider to know internal costs) the costs of labor and the cost of capital. This helps to determine the cost of capital given the expected profit for each model. The blue boxes above are meant to show that these percentages are not fixed.
Below is the expected gross margin per car (in yellow) as well as the individual gross profit per car model. After this, the expected annual gross profits per model are shown.
This is followed by a reconstructed, traditional income statement for the automotive segment based on the above figures. It shows aggregate numbers for revenue, cost of materials, cost of labor, and cost of capital. It also shows a gross profit. Below this, the company-wide operating expenses are shown for research and development as well as selling, general, and administrative expenses.
This, again, is a stressor on the analysis since the operating expenses are not allocated only for the automotive segment.
Finally, the interest expense is also shown. This shows the possibility of obtaining profits.
Keep in mind the kind of applied stressors:
1. The capital cost of the whole company was allocated to the automotive segment.
2. The lowest possible average selling price was applied per model.
3. The company-wide operating expenses were applied (again) to only the automotive segment.
The analysis provides the possibility of covering all costs. The amount of profit before taxes (3.4% of revenue) is modest. However, the proposed model is a relatively high-volume strategy with low sales prices. It is a positive thing that the analysis allowed for a profit.
If the model would have come negative, then this would provide support that the business is not feasible, and this is shown in the fourth column. The two differences between both models is a lower cost of capital of $1,514, shown above in green, and a 5% increase in the cost of materials, shown above in blue.
This shows how sensitive the business model is to the cost of materials. Tesla has been criticized for using a vertical integration approach for building cars. According to Max Warburton, a respected car automobile analyst, “The risks of what Tesla is doing are extraordinary. … It’s not only the most vertically integrated tech company you’ve ever seen. It’s also the most vertically integrated car company since Ford in the 1920s.”
The small difference in the cost of materials shows that vertical integration is the correct approach for maintaining control over material costs. I have included below, for those needing it, Tesla’s latest income statement as a reference.
If you found this discussion on Tesla’s capital of interest, let me know. Perhaps I can provide a presentation on materials or labor for the next episode.
I hope the above information was of value to you.
Below, I present the model used above by applying the “watering down” process applied to capital using both quarters with different weekly production levels of Model 3 units as discussed above. In other words, the weekly production rate of Model 3 cars is kept at 8,744 in the last quarter and at 5,000 in the previous quarter.
To begin, I previously showed that at higher average selling prices, Tesla will break even at lower weekly production levels than those shown in this article.
Second, in this article, I previously showed the effect on Tesla’s profitability of producing higher volumes of Model 3 cars when allocating 100% of the capital invested at very low average selling prices, and I showed the sensitivity of Tesla’s business model to a rather small explicit change in material costs (5%) and small implicit changes in capital unit costs (increase of rate of change of 6.8%).
The small increase in capital cost was solved by showing profits. However, in spite of smaller unit capital costs, the increase of material costs was not solved and losses were incurred. This makes sense when you compare the relative size of unit capital costs to material costs. In the case of the Model 3, capital costs are about 7% the size of material costs. Therefore, a smaller percentage change in material costs has a greater overall effect than a larger percentage change in capital costs.
Finally, in this appendix, I show the effect on profits when allocating only 80% of the capital invested. When increasing the weekly level of production (from 5,000 cars to 8,744, a 54% increase) of Model 3 cars, Tesla will make a profit when selling cars at the lowest possible average selling prices.
At a level of 5,000 Model 3 cars, things are in fact pretty good considering everything. With very low average selling prices (a worst case scenario), Tesla covers all of its expenses except for about 10% of operating expenses plus the interest expense. Keep in mind that this means that it has broken even at the gross profit.
What was the effect on the bottom line of a 20% reduction in Tesla’s assets while keeping everything else the same?
Well, if Tesla increases volume by 54%, it can reduce capital unit cost by 31%, and this allows it to reach profitability.
Now I can see that the company can produce a profit at higher car volumes with lower sales prices. However, there are still one or two problems left to solve, and I look forward to discussing these in the near future.
By the way, someone might question my motivation in spending so much time and effort in financially analyzing a car company since it does not seem to tie directly to the theme of solar, wind, geothermal, biomass, biogas, tidal, and wave energy.
Well, you are right. Traditionally, a car company would be the opposite of these clean energy generating sources. However, several years ago (perhaps more than some of you have been alive) there were a few car companies that brought to market electric cars, and everything was well with the universe … until it was not. These same car companies decided to literally crush these first electric cars.
I was really bummed. Nobody else, with the exception of a few rather smart and enterprising individuals, were tinkering (in private) with electric cars. Then a few years later a small company decided to give it a go, and they tried it with electric cars that had swappable batteries. I hated the idea of swapping my car’s battery, but it was a start. Right?
Well, without going into further details, the company was also killed from the top, and again guess what? I was bummed. Then, after being bummed for so long on electric cars, Tesla came along. They came in assertively and with the hope of making a real change, and my spirits were lifted, and in the last year, in particular, all sorts of sad comments have surfaced questioning the credibility of the company and its CEO.
These small articles are my way of contributing in whatever infinitesimal way I can to seek the truth. This is the reason why I do it. Luckily for me, my analysis has proven so far on Tesla’s side, but its financials are a bit of a mess — and I see it as my job is to untangle this mess.
Disclaimer: I have, sadly, no positions in the stock of this company [TSLA]. The above is not a recommendation for investment. If you seek investment advice, consult with a professional, and if you do invest, then do it wisely and without allowing the loss of any of your investments to injure you financially. Remember that money invested in the stock, bond, option, futures, and real estate markets can and might go to zero.