In my last post, I showed that the total cost of buying a Tesla with a loan (or cash) is easy to figure out. However the final value is critically dependent upon one thing: the depreciation of the car when the owner sells it. Unfortunately that quantity is uncertain, because it has not happened yet and none of us have a crystal ball.

Even though we cannot predict its exact value, we can use simulation to study the statistical distribution of the total loan cost.

For simplicity we assume the annual depreciation rate follows a normal distribution, e.g. one with mean 15% and standard deviation of 2%, i.e., Normal(15%, 2%) as shown in the following:

The resulting total loan cost has the following statistical distribution:

We can immediately see that there is a non-trivial probability that the total loan cost can exceed $18,769 (the total lease cost as shown earlier).

Again, I stress that the distribution is just a tool to assess the range of outcomes. The takeaway is that while buying is still a cheaper option, it is not guaranteed.

Moreover, buying brings uncertainties whereas leasing has much less of uncertainties. Perhaps that is one of the reasons leasing remains a good option for at least some people.

The spreadsheet contains the simulation can be downloaded here.

Here is the a video of me talking about it: