Tesla Autonomous Driving Case Study: DCF Vs Probability Based Valuation
Ashwath Damodaran, is an iconic professor of valuations at NYU Stern & has trained a generation of corporate finance specialists by generously making hisspreadsheets & methods available online to non-NYU students long before the days of the MOOC & the Khan Academy. He recently wrote two posts on Tesla’s valuation: The first post in Sep 2013 had a valuation of $67 in Sep 2013 & the second post updated the model in Mar 2014 to give an estimate of 2014. Both these models rely on a single Discounted Cash Flow Valuation (DCF) method which projects revenue, profit & cash flows into the future & discounts them back to the present.
In a high growth company like Tesla, the most important assumption in any DCF model makes is around the potential growth in revenue at the end of the forecast period. Prof. Damodaran made a critical choice by assuming that Tesla would like Audi in the future. Here is the below chart where he makes that assumption.
However, the DCF model does not take into account powerful optionality that companies can extend into new markets or develop new technologies that could change the market trajectory. Damodaran himself has discussed a probability weighted scenario analysis like the real options methodology but he did not use that method in the Tesla analysis.
But, what do we do when we have truly innovative technology like autonomous driving that could make Tesla blow past projections? Shall we attempt the probabilities of these different scenarios & be approximately right than exactly wrong?
Goldman Sachs attempted a probability based approach for five scenarios in March 2014 to value Tesla. The table is given below:
The “Maytag Repairman” scenario assumes a convenience based model like dishwashers that altered the perception of consumers value. Similarly, could autonomous driving alter the perception of convenience of a Tesla vehicle based on Autopilot features. Goldman’s “Maytag Repairman” valuation was at $329 for Tesla’s automotive business but only at 8.3% probability of happening. Each individual scenario’s valuation was done based on a DCF model but the aggregate valuation was derived through a probability weighted approach. Therefore, the aggregated probability weighted model was $180 for the automotive business.
However, after our recent research on Tesla’s Autopilot hiring patterns, the probability of the “Maytag repairman” scenario will likely be more than 8.3%. We changed the probabilities based on the Goldman model & outlined a different probability based valuation.
The valuation when we changed the probabilities of the “Maytag Repairman” scenario to 39.2% & correspondingly decreased other scenarios, jumped to $255 for Tesla’s automotive business and $275 after accounting for the Grid storage optionality.
The autonomous driving technology illustrates a contentious issue for finance practitioners. Shall we ignore the Autonomous Driving possibility & only value the company based on current reality? Or Shall we factor in the probability of the autonomous technology working & appropriately adjust our world view?