“If we can send a Roadster to the asteroid belt, we can probably solve Model 3 production.”

As expected, Tesla saw some pretty big losses this quarter, mostly related to Model 3 delays. The company reported $675 million in losses attributable to shareholders in Q4, wrapping up 2017 with almost $2 billion in losses for the whole year.

But, unlike other doomed companies posting dire losses quarter after quarter, Tesla revenues have been sizable. Just this quarter, Tesla earned $3.3 billion in revenues from automotive sales, leasing, energy products, and services. For the year, the company reported almost $12 billion in revenue. People want Tesla products, but Tesla can't stop spending more money than it has.

The company (naturally) contends that all losses are temporary. "At some point in 2018, we expect to begin generating positive quarterly operating income on a sustained basis," the company stated in its investor letter. Musk specified in the accompanying earnings call that he expected the company to be profitable by the more stringent Generally Accepted Accounting Principles this year.

That line may sound familiar: Tesla claimed similar projections in early 2016, and while the company did indeed have its second-ever profitable quarter that year, it resumed its loss-making habit shortly after.

Model 3

Tesla now has to crank up Model 3 production. In early January, Tesla reported good delivery numbers for Model S and Model X vehicles, and the company's energy business certainly seems healthy (Tesla installed 143MWh of battery storage in Q4 2017, not counting its massive installation in South Australia). But investors are concerned about the Model 3 because it was supposed to be the mass-market car that would fly off the assembly line and make money for the cash-burning company.

The entry-level electric Model 3 debuted in July 2017 to great fanfare, but Tesla has struggled for months to ramp up production to meet demand. The quarter after the Model 3 was introduced, Tesla revealed that it had only delivered 266 vehicles. The company said the issue was bottlenecks on the production floor, and it bought an automation and machining company called Perbix, presumably to handle those bottlenecks.

But by the end of Q4, deliveries seemed only marginally better. In early January, the company released a statement on Q4 Model 3 deliveries ahead of today's Q4 earnings call, and Tesla said it had delivered just 1,550 Model 3s in Q4. At the beginning of the year, the company said it expects to churn out 2,500 Model 3s per week by the end of Q1 2018 and 5,000 per week by the end of Q2 2018. But the investor letter published today seemed to hedge that estimate even further.

"It is important to note that, while these are the levels we are focused on hitting and we have plans in place to achieve them, our prior experience on the Model 3 ramp has demonstrated the difficulty of accurately forecasting specific production rates at specific points in time," the letter said.

Musk's attitude seems to be that there's no doubt Tesla will achieve its ramp numbers and that minor delays are negligible in the big picture. "If we can send a Roadster to the asteroid belt, we can probably solve Model 3 production," he said on today's earnings call, referencing the SpaceX Falcon Heavy launch yesterday.

The extra stuff

Tesla never seems to want for side-quests. The company builds stationary lithium-ion battery storage for residences and utilities: it built 330 new Superchargers last year, and it just entered a new deal with South Australia to build a huge 250MW virtual power plant. Last fall, Musk revealed a semi truck slated for production in the near future and a new ultra-luxury Roadster. Musk has intimated that, after the currently-unannounced Model Y is revealed, Tesla will develop a pickup truck.

In today's investor letter, Tesla hinted at continued development of Autopilot, an automotive feature that saw a great deal of controversy in previous years. The company revealed that it had completed "an extensive overhaul of the underlying architecture of our software" enabling "a step-change improvement in the collection and analysis of data and fundamentally enhanced its machine learning capabilities."

Musk also denied that LiDAR was strictly necessary on cars for autonomous driving. "We have to solve passive optical extremely well," he said, referencing running image recognition based on camera data. "At the point that you have solved passive optical extremely well, what is the point of active optical?"

As for energy, the company reported that the South Australia installation is working as planned in its investor letter. "This project is already generating substantial benefit by meeting high summer demand when supply is limited and by instantaneously responding to unplanned interruptions or frequency drops in the grid," Tesla wrote. "Due to the success of this project, we're seeing an increase in demand for Powerpack, our commercial energy storage product."

Musk briefly addressed the semi truck on the phone call today as well. He said that a run rate of 100,000 units a year would be a reasonable expectation four years from now (although given previous years-out projections for the Model 3, it's safe to take that with a grain of salt). "I think we may be able to exceed the specs that we announced last year, which would be exciting," Musk added. Tesla surprised many by announcing a 500-mile range for the semi last year.

Musk also talked about the Fremont factory and the Gigafactory in Sparks, Nevada. He noted that, on Model 3 production, the company's semi-automatic lines are working faster than the fully-automated lines, but Tesla's engineering firm in Germany has developed a fully automated line that will be delivered to Tesla's facilities at the end of March. "The competitive strength of Tesla is not going to be the car; it's going to be the factory," Musk said. He added that a primary objective for Tesla is "productizing the Gigafactory."