Data Trends for Investment Professionals


Tesla’s Model 3 Has Miles to Go

Manufacturing bottlenecks continue to plague the Tesla Model 3. In October, Tesla’s target of 5,000 car deliveries per week by Q4 2017 was pushed back to Q1 2018. Yesterday, Tesla pushed the target back again, to Q2 2018, after a disappointing quarter for production. Quandl clients had early warning of this shortfall thanks to our model-level insurance policy dataset. Read on for more details.

Tesla Stuck in “Production Hell”

A few years ago, CEO Elon Musk claimed that Tesla would deliver over 100,000 Model 3s in 2017. That aggressive goal was subsequently scaled back substantially, to a seemingly more achievable cadence of 5,000 cars per week by the end of the year. Yet that too has proved a moving target.

In October 2017, Musk offered the public a rare glimpse into the “production hell” that has plagued the Model 3. The main culprit, according to the official announcement, was the battery pack assembly: “We had to rewrite all of the software from scratch”, which added up to “about 20 to 30 man years of software in four weeks” for the new battery module. “This is what [we] spent many late nights at the Gigafactory working on.”

Although Tesla hinted at production bottlenecks at both the Fremont factory and Gigafactory, the delays seem to be a result of multiple weak links within the supply chain, both out- and in-source. These issues prompted Tesla to push back its timing for the 5,000 cars per week milestone, from Q4 2017 to Q1 2018. And now that too has been pushed back.

Tesla announced on January 3 its actual delivery numbers for Q4 2017. Despite an uptick in the last week of December, production remains stubbornly slow. Tesla delivered a total of 1,550 Model 3s in the entire quarter — an average of 120 cars per week. That’s a big miss from the company’s own target and lower than the already pessimistic forecasts from auto analysts, who estimated 2,900 vehicles over the quarter.

Tesla is now targeting 2,500 cars per week by the end of Q1 and has pushed back the critical 5,000 cars per week deadline to the end of Q2. Will they hit this deadline? Or will investors finally lose patience with these execution delays?

An Investor’s Dream… Or Nightmare

Tesla is an unusual stock for investors. Unlike most companies where “radical change” might mean a few percentage points up or down, Tesla’s performance is truly binary: all or nothing. In the most bullish scenario, given the company’s advances in battery technology and autonomous driving, Tesla could conceivably become the largest and most important vehicle manufacturer in the world. In the bear case, however, Tesla could very well go bankrupt if it is unable to resolve its production issues before running out of money.

Even more unusual is that these vastly divergent outcomes hinge on a single metric: the number of cars Tesla can successfully deliver every week.

The company is notably secretive about this number. While other auto manufacturers publish detailed sales figures every month, Tesla publishes its delivery counts only once per quarter, in aggregate form. Access to Tesla’s production facilities is tightly controlled, as is the company’s media profile. Tesla’s direct-to-customer sales model is another factor that makes estimating Model 3 deliveries difficult.  

Alternative Data Sheds Light on Tesla Production

For the last few months, Quandl’s customers have had an information advantage when trading TSLA stock, thanks to our Tesla Delivery Estimates dataset.

Here’s how the dataset works.

Automobile insurance is mandatory in the US. For this reason, new car deliveries are nearly perfectly correlated with the sales of auto insurance policies. Quandl has partnered with a consortium of auto insurance providers who command a statistically significant share of the US auto insurance market. Our insurance firm partners share their daily policy sales counts with us. These counts are then combined to calculate weekly Tesla car deliveries in the US.

The data is available at a weekly level for the Tesla Model 3 and at an aggregate level for all Tesla models. We have found that the dataset tracks actual Tesla deliveries with extremely high accuracy.

Using this dataset, we predicted that Tesla would miss the analyst consensus (of 2,900) weeks in advance of the announcement. We estimated that Tesla delivered 1,010 Model 3s over Q4 2017. This number was significantly lower than analyst consensus. Actual deliveries came in at 1,550, confirming our very low estimate.

Tesla’s Model 3s Coming Down the Road

While Q4 was an irrefutable production miss for Tesla, there is a silver lining. Both the company’s official press release and our own data suggest that there was a clear increase in production speed towards the end of the quarter, to the tune of several hundred cars per week.

If Tesla can maintain this increased cadence or even improve upon it, a delivery velocity of 2,500 cars per week by late Q1 or early Q2 seems entirely possible —  an outcome that would have seemed overly optimistic even a month ago.

Much will depend on how quickly Tesla can ramp up and sustain its production. We see several possible scenarios:

In the coming weeks, our Tesla dataset will validate one of the above production scenarios with high confidence — and well ahead of the rest of the market. Are you interested in gaining that same informational advantage? If you are a Tesla investor interested in tracking the company’s deliveries through 2018, contact us to see the data.

Learn more about this data.

Fix This
Created with Sketch.