28th April 2019 by Michael Barnard
Recently, Elon Musk was interviewed by an MIT artificial intelligence researcher, Lex Fridman, and made a claim he had previously made.
Essentially, buying a car today is an investment in the future … The most profound thing is that I believe that if you buy a Tesla today, I believe that you are buying a value-added asset, not a devaluing asset.
At an Investor Event A few weeks later, Musk went on to say that Tesla owners could earn $ 30,000 in gross annual profits from their cars. That was qualified for a million miles and eleven years of vehicle life, but it's not that Musk has shown his math here.
We have decided to consider this claim in different ways. What is the background for this? For which Teslas did it apply? And has the used car market picked up the signal?
This is no news, Musk is currently only talking about it. The creation of a Uber rival, where its own Teslas drove itself, has long been on Musk's radar. With Tesla repaying $ 920 million in debt in the last quarter, delivery to other continents began (resulting in a delivery delay that impacted revenue), and he invested heavily in his next gigafactory in Shanghai for the Chinese and Asian markets (which led to a loss in the first quarter and expected losses in the second quarter), it is likely that a positive comment was required.
A relevant announcement from October 2016 was that all Teslas were manufactured from this point on with all hardware including sensors and neural sensors. Nets and actuators will be fully self-propelled after completion of the software component, and the upgrade to full autonomy would be an Over be the air update. They've optimized their processor ever since, but Musk has assured owners who buy full self-drive capabilities that the processors will be replaced for free. And new vehicles will be delivered with the new processors.
~ 6 months before it's in all new production vehicles. No change to sensors. This is a simple replacement of the autopilot computer. Will be performed free of charge for those who have ordered full self-drive.
– Elon Musk (@elonmusk) October 16, 2018
"Ordered" vs. "Reordering" could mean some extra upgrade fees for those who have not preestablished, but we'll see how that affects. Regardless, this is not important for further analysis.
Assessment of Claim
When complete autonomy is available, customers may choose to have their cars used for communicating with passengers when they are not being driven. Since the average car is driven at 5% a year, it's time to earn money instead.
Let's assume that 50% of Uber prices are not available to pay the driver That's about $ 1 a mile. Suppose you receive a limited number of 10 days of flight per day, an average of 3 miles and 5 days per week. This translates into $ 30 a day, $ 150 a week, $ 600 a month, and $ 7,200 a year.
30 miles require about 10 kWh of electricity for an average of $ 0.12 in the US, which is $ 3.60 per day for extra costs worst (most electric vehicles charge in low billing times in the middle of the night or for free available chargers). That costs $ 18 a week, ~ $ 72 / month and ~ $ 864 a year.
They double the average daily US trip with 30 miles extra per day almost, so there's extra wear, cleaning and hoop costs. Call it another $ 1,000 a year.
$ 7,200 – $ 1,100 – $ 864 = $ 5,336 gross profit per year for something you do not have to think about. Double the rides to 60 miles a day, that's $ 10,000 a year. Triple them to 90 miles per day, $ 15,000.
Average taxis in New York drive in each 12-hour shift 180 miles or 360 miles in one day. Most of the time they travel at fares, not with paying customers, we assume that 60% of the miles are below the fare, leaving 216 miles per journey flown. Take away the 5% of personal use and that's close enough for 200 miles per day under the fare. That could mean a profit of over $ 30,000 a year, assuming Tesla has not made a significant cut in the lead. Even that would be manageable with a reasonable adjustment of prices.
Tesla Model 3s ships with all of the $ 35,000 in autonomous hardware. Call it $ 45,000 with autonomy turned on. Used Model S 60 and 70 Tesla models, built after October 2016, are in the low price range of $ 50,000.
Most vehicles lose about 30% when shipped. Suppose your Tesla lost $ 15,000 in the first year, then probably another 10% per year. Let's work out some numbers to see if you even get a break-off on depreciation.
Yes, we could be ahead of depreciation with moderate carsharing pretty fast. And of course, these are depreciation rates that seem reasonable, but these would be hampered by the greater benefits.
If I could buy a used, one-year Tesla Model 3 for $ 30,000, which brings me $ 30,000 a year, that's actually an amazing deal. It is questionable that people pay a premium for it. Even a used model S 60 or 70 at this price would be very interesting.
Does the used car market show any signals here?
One of the places we would expect to have smart people The opportunity lies in the used car markets.
In early January CleanTechnica released two pieces on the used car market and value for Tesla Model 3s. The first found out that Tesla Model 3 was hardly present on the used car market – based on an analysis by Autotrader we compared the Model 3 with its competitors Audi and BMW and also with cheaper cars with the same volume.
We have just updated this assessment three months later to see only third-party model 2019 vehicles and to see if people did not sell their Model 3 in part because of a possible loss of their tax refund in a pre-2019 sale. Comparing apples to apples, Tesla Model 3 remains heavily underrepresented in the used car market. This data point may suggest that owners are thinking about the utility of the car, as many model 3 buyers are likely to buy the more desirable car. The several hundred thousand pre-orders for the Model 3 have indicated that this is desirable at the price points at which Tesla projected it, as well as its sixth-best-selling car status in the US in the last three quarters combined.
Then CleanTechnica pointed out that Kelley Blue Book has a grinning head and accepts the resale value of Model 3. The Kelley Blue Book described it as the best electric car for the expected resale value, but its own published figures made it clear that the Model 3 would hold its value better than any other car period, but has not yet dubbed this route. The second best electric car was the model X in this review. According to this thesis, this may indicate that people at KBB have recognized the value of autonomy, but if you read KBB's actual comments, there is no indication. It does not appear that the resale analysts rate it the way Musk claims it should, or the previous numerical work suggests it should not.
Another data point to assess is the recently released iSeeCars.com analysis of the resale of the Tesla Model S compared to other vehicles. They found that the S model was the fastest-selling used car in its class by a considerable margin, almost 10% faster than the much cheaper Audi A7 and A8 and with almost half of the Maserati Quattroporte's time to market. The similarly comparable BMW 7 Series takes almost two weeks longer, and the cheaper 6 Series is in the same boat. From this analysis, it is clear that Tesla's used cars are also high on the market, but does this send a signal about how buyers rate them on the basis of an autonomous taxi service?
We've reached iSeeCars to clarify a few issues to see if further browsing through the data would reveal something interesting.
The iSeeCars team indicated in its report that the Model S was the most discounted vehicle, but it was not clear how they came to this claim from the report itself. What they did was to look at the average of all used cars in this model to look for outliers. The Tesla Model S showed the biggest deviation in the used price.
However, this seemed potentially consistent with the Tesla Model S being offered in a vast range of new price points for various configurations, from the original Model S of $ 60,000 60 to $ 135,000 P100D. This is a much wider range of EIAs than its competitors, whose product lines are much narrower in segmentation and therefore include multiple offerings covering the same range. Could that be the cause of the deviation iSeeCars saw? They looked deeper and provided this interesting insight:
"P100d has the lowest percentage of all trim models that is 10% or more below the market value. The 60D had the highest percentage of prices that were 10% or more lower.
This seems to indicate that the 60D had the largest number of cars that were less desirable. This may have been a matter of fact for the future owners, but this seems to be the desire of the P100D. On the used market buyers want to get as much Tesla as possible.
Finally, the question arose whether the underlying data show anything about the vehicle value with full self-propelling characteristics. If this were the case, the statistics for cars would change significantly between 2016 and 2017. Once again we asked this special question from iSeeCars.
"Our analysis shows a decomposition of the model year or the Tesla manufacturer. 2016 and 2017 are almost identical for the market days: 2016 is 35.8 days, while 2017 is 35.9 days. 2018 are 29.0 days. "
Once again, the used car market does not show that buyers clearly assess the potential for full self-drive, but merely that they want the latest and greatest Tesla.
Where is that? us?
Musk's claim seems to be correct. It would be possible for a Tesla with full self-drive and a Tesla robotaxi autonomous carpool service to earn a lot of money each year for Tesla owners. However, there is no clear signal from the underlying used car market that everyone is interested.
Will everyone decide to take the car? Of course not. But will everyone get a family vehicle that could take any family member to where they need to drive without mom or dad? Yes, and that's worth a lot. In fact, it is quite possible that many families with two cars and children can have a single autonomous car that has the same benefits. Since the US government estimates the annual cost of a car is $ 9,576 per year, Musk may want to recommend one-vehicle families instead of Tesla Taxi!
Tesla's are highly sought-after used cars that sell faster than competitors and non-competitors In many cases, the resale value can be kept longer, but this does not seem to be geared towards earning money with the future Tesla Taxi capability. Of course this is like Henry Ford at the beginning of the 20th century and said, "If I had asked people what they wanted, they would have said faster horses."
Most people have not internalized the idea that theirs is Tesla's spinner instead of cost turns into money. When that starts to decline, the data may change. But by then, other vehicles will probably also have self-propelled vehicles, and companies like Uber will use them at competitive prices, leading to a similar rating on the used-car market. Perhaps.