When it bought SolarCity late in 2016, Tesla (NASDAQ:TSLA) may have taken on a manufacturing challenge that makes automobiles look easy. The solar industry has left a string of bankrupt manufacturers over the past decade, and a price drop of over 30% in 2016 alone could be a sign that more revenue pressure is ahead.
Amid this backdrop, Tesla is trying to make new technology work at scale, is bringing in a new partner, and is launching the solar roof. That's a lot to take on all at once, and it could be more than even Tesla can handle.
Tesla's manufacturing plans are complicated, at best
SolarCity was in the process of building out manufacturing capacity using technology it acquired from Silevo when it was acquired by Tesla. The acquisition meant bringing Panasonic, Tesla's battery partner, and its HIT technology into SolarCity's plant.
Panasonic's technology has similarities to Silevo, but is different enough that they're not a natural fit. There's also the small complication that Silevo uses a 156 mm cell and Panasonic uses a 126 mm cell. In other words, bringing SolarCity, Silevo, and Panasonic technology under one roof is far from a trivial task.
It's unclear how much equipment was installed before the Tesla acquisition, but there's certainly some Silevo technology in SolarCity's Buffalo facility. In its research, PV Tech also cited job postings that included language about Silevo technology as evidence that Panasonic won't be the sole technology in the plant.
We also know that Tesla will be making solar roof cells and shingles in the Buffalo plant. In other words, Tesla will need to manufacture three separate products in the facility.
Solar manufacturing is hard
Making one solar product cost competitive is hard enough. But Tesla is now trying to make three products simultaneously, and is facing a solar industry that currently is under pricing pressure. The price of solar panels fell over 30% in 2016, and prices are now around $0.40 per watt for commodity solar panels, which, not coincidentally, is what SolarCity is still installing as it ramps up solar manufacturing.
Neither Tesla or SolarCity has given any information about what efficiency or costs look like coming out of in-house manufacturing to this point. But given delays in launching solar panels and the fact that Panasonic had to be brought in as a partner, we can discern that it isn't going well.
The financial impact could be big if manufacturing fails. New York was nice enough to build the $750 million plant and buy equipment for it, but that came with a commitment to spend $5 billion in the state over the next decade, including hiring 1,460 high-tech workers. And SolarCity estimated it would still spend around $150 million in capital costs at the plant, plus cost overruns. If SolarCity, and now Tesla, fails to meet obligations for investment and job creation in New York it could be obligated to pay $41.2 million per year as a "program payment". If panels aren't competitive, it's safe to say the impact on Tesla could measured in the hundreds of millions in losses.
Why Tesla doesn't need solar manufacturing in the first place
At the same time that Tesla is trying to complete manufacturing plants to make three products, competitors are improving technology and lowering costs. On top of the already low and declining costs for commodity panels, many manufacturers are building out mono-PERC high-efficiency manufacturing capacity. Mono-PERC technology that's being built out is comparable in efficiency to what Panasonic was producing, so Buffalo may not have much, or any, efficiency advantage over competitors.
In other words, Tesla could be buying solar panels in the open market that would be similar to what the Buffalo plant would make. No headaches, no risk, no promises to invest in manufacturing capacity that may not be cost competitive at scale.
It looks like getting into solar manufacturing was an unnecessary risk for SolarCity, and now it's a problem for Tesla. Investors should hope it's not a distraction that becomes a financial drain, something the company doesn't need right now.