What is this thing?
This is an earnings model that I made. Earnings models allow you to value a company based on your own view of future company performance. Earnings models are used all over Wall Street - by sell-side analysts (the guys that write the buy/hold/sell reports) as well as by buy-side funds (people who buy stocks for hedge funds, long-only funds, etc).
This model mirrors many of the other Tesla models that institutional investors (the audience mentioned above) use. I’ve used my background on the sell-side as well as my experience in dealing with Tesla models to make this available to anyone who wants to play around with it.
Why are your estimates so high/low/stupid?
The model was made to reflect consensus estimates, not my own views. Consensus estimates are the average of the sell-side analyst estimates, and are used as a proxy for investor expectations. Estimates are updated as of the last update date on the model.
NOTE: Estimates beyond 2020 start to get hazy as much of the street does not model this far out. Additionally, estimates for vehicles beyond the Model S/3/X are difficult to come by as most of the street is either not baking future models into production estimates yet, or are using other line items (like the Model 3) as placeholders, making it difficult to separate their views.
How do I use it?
If you want to play around with estimates, go through the model and change the blue numbers (the inputs) to reflect your own views. By far, the biggest drivers in this model, and the ones you probably want to focus on, are:
Model 3, S/X, and future models deliveries
Model 3, S/X, and future models gross margin
R&D and SG&A expenses (modeled as a % of sales)
These drivers will make up most of the variance in the outputs. From there, you can go to the valuation tab that will then connect those outputs to target prices for the stock. You can use any of several methodologies that are listed within this tab to get to a target price for Tesla.
Another use of the model is to just look at historical data and/or look at where expectations are today for the company. Just make sure to expand the columns at the top of the sheet.
As I mentioned in the model, I am including only GAAP estimates. This differs a bit from the street, which offers both GAAP and non-GAAP EPS estimates. The primary difference between GAAP and non-GAAP is that stock-based compensation is stripped out of expenses because it is a non-cash expense. This was a practice that tech companies started, but some companies have begun to walk back as criticism has increased. The argument is that stock-based compensation (SBC) is still a cost to shareholders, and is recurring (because these companies rely on it to attract and retain talent). Therefore, it should be counted as an expense. GAAP estimates include SBC, and I have taken the approach to simply show GAAP and not non-GAAP, as I don’t believe there is validity to the non-GAAP numbers. Remember this when you see Tesla EPS numbers on the internet, as the difference between GAAP and non-GAAP is significant, and people must be careful in separating the two. In the future, I may bring this into the model.
I’ve included several valuation methods within the model. Note that the DCF approach has many more inputs that must be tweaked, in addition to the ones highlighted above, as it is a much more precise (but not necessarily accurate) way of modeling. You’ll have to adjust estimates going out to 2030, the WACC, and the exit multiple.
Take a look at the comments within the cells; they contain management guidance as stated in the press release or within the conference call.
Posting an excel model onto a website is a huge pain. Let me know if you know of a better way to do this.