Tesla's Q3 Vehicle Safety Report Reality Check

About: Tesla, Inc. (TSLA)
by: Quoth the Raven


Tesla just started reporting quarterly vehicle safety data, which it will do now going forward.

The company boasted about its vehicle's safety versus NHTSA averages, and specifically its vehicle safety with Autopilot engaged.

The truth is that Tesla's data may not be robust enough to compare to NHTSA averages.

In my opinion, it's important to put some context around these numbers.

One of the key pieces of good information that Tesla longs have been clinging to over the last couple of days has been the supposition that the company's on-road vehicle safety track record is significantly better than NHTSA averages.

Days ago, Tesla released its first ever vehicle safety report in which it stated that its vehicles, compared to NHTSA averages, had produced what appeared to be incredible results. The press release boasted these two key pieces of information:

  • Over the past quarter, we’ve registered one accident or crash-like event for every 3.34 million miles driven in which drivers had Autopilot engaged.
  • For those driving without Autopilot, we registered one accident or crash-like event for every 1.92 million miles driven. By comparison, the National Highway Traffic Safety Administration’s (NHTSA) most recent data shows that in the United States, there is an automobile crash every 492,000 miles. While NHTSA’s data includes accidents that have occurred, our records include accidents as well as near misses (what we are calling crash-like events).

Here's the data Tesla wants you to see, the way they probably want you to see it:

  • Tesla: 1 accident or crash-like event per 3.34 million miles with Autopilot on
  • Tesla: 1 accident of crash-like event per 1.92 million miles driven
  • NHTSA: 1 crash per 492,000 miles driven on average and this doesn't even include "crash-like events"

Extremely impressive looking, right? Off the bat it seems like you are almost 75% less likely to get into an accident or crash-like even simply driving a Tesla! That number appears to move to about 85% if you have Autopilot engaged! It's settled then, let's all get in our Tesla vehicles and bow at the altar of Autopilot, engaging it for any and all purposes, so as to remain as safe as possible!

But before we do that, let's just take a closer look.

As recent article from the LA Times pointed out, the raw data that Tesla started with to arrive at these numbers doesn't appear to be anywhere near as robust as the same data that the NHTSA uses to arrive at their numbers. Also, the Tesla data may be subject to several controls where the NHTSA data appears to be subject to significantly more variables. Therefore, in my opinion, the comparison between Tesla's data and the NHTSA data in the company's release simply doesn’t make sense.

For one reason, Tesla's data only includes newer vehicles. The Model S was officially introduced in 2012 and is likely to be the "oldest" vehicle included in Tesla's data. Theoretically, it'd be easy to assume that many of the vehicles Tesla uses in its data are newer than 6 years old. NHTSA data, on the other hand, includes vehicles that are sometimes 10, 20 and even 30 years old and, as a result, far more likely to be prone to accident-causing malfunctions.

Second, a study done by MIT found earlier this year that Tesla's Autopilot is most often used on highways. The NHTSA data includes drivers who are driving in all types of environments. Winding back roads and city traffic are likely to have a far different accident profile that a car cruising down the highway. The LA Times article commented:

Autopilot is recommended for use mainly on limited-access highways, but the total accident figures cover all roads and conditions. Tesla drivers have higher incomes than average and have much newer cars than average, which could skew results.

Third, the total number of vehicles on the road is significantly smaller for Tesla than it is for the NHTSA data set. Tesla is estimated to have somewhere under half a million vehicles on the road, total. The NHTSA numbers reportedly include about 271 million vehicles.

So let's be real here. The numbers Tesla presented can be looked at as a very promising looking start. However, scratching below the surface, it looks as though the presentation of the company's vehicle safety data, and future iterations of the same data, may not be best compared to NHTSA data for the company to make its point. I'm sure the company had no mal-intent for doing so, but in the future, why not present this data versus sequential and YOY data collected by Tesla using the very same controls? As the number of vehicles on the road begins to grow, we should start to be able to piece together a trend for whether these numbers will get better or worse.

Should Tesla start to "move the goalposts" and present this data in different formats and using different variables going forward, we'll have another reason to be skeptical. It is important for anyone following the Tesla story, as a supporter, reporter, analyst or investor on either side of the coin, to have proper perspective of this information.

It’s difficult to predict how this new perspective on this data could impact the stock price, other than to say how it is presented may be perceived as slightly misleading (assuming it wasn't done on purpose) or dishonest (assuming it was done on purpose). Given that Tesla trades at an extremely speculative valuation that is based mostly on sentiment, if investors believe that the company is further going out of its way to put a "polish" on numbers, it may negatively affect the company's ability to regain confidence from the market.

Disclosure: I am/we are short TSLA.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.