First Solar Inc. (NASDAQ:FSLR) is clearly the most valuable solar manufacturer in the world today. The company has built the most profitable business in solar energy on the back of thin-film technology that no one else could make work at scale and has become a go-to partner for utility-scale developers. 

The company is also going into its biggest business transition ever. It's spending about $1 billion to upgrade from a product called Series 4 to a new solar module called Series 6, which is a larger form factor, higher-efficiency solar module, and aims to lower costs to install than its predecessor. It's a big bet that First Solar has to get right, and here's where we can expect First Solar to be when the upgrades are complete. 

Utility scale solar installation with a partly cloudy sky in the background.

Image source: First Solar.

After strong 2017, 2018 could be choppy

Calendar 2017 has been dominated by strong demand for all of First Solar's projects. But it should be highlighted that the fact that management expects $1.55 to $2.20 per share in earnings is due in large part to the project side of the business. Module sales are going well, but almost the entire increase in guidance after the second quarter was based on project sales

Looking at 2018 and beyond, First Solar will reduce its focus on the development business and focus on selling modules and fully engineered utility scale solutions. By not holding projects to completion, it will vastly reduce the risk from changing interest rates or currencies, which aren’t risks solar manufacturers want to take.

That will mean the windfall of profits coming in 2017 will be gone and the company will have to make most of its money on the solar module and component sales. And with equipment being upgraded, there will be some changes in what's being produced. Here is a table of expected production based on guidance and management's comments.

Product 2017 2018 2019 2020
Series 4 production 2.6-2.7 GW  2.1-2.4 GW  0 GW  0 GW 
Series 6 production 0 GW  1 GW  3.5 GW  4.0 GW 
Total production 2.6-2.7 GW  3.1-3.4 GW  3.5 GW  4.0 GW 

Data source: First Solar earnings reports and public comments. 

You can see that production will likely rise in 2018, primarily because Series 6 will come on line faster than Series 4 will be shut down. Despite this fact, we may not see revenue growth. Component sales are less valuable per watt than selling entire projects and most Series 4 production will be component or module-only sales

Since solar systems cost about $1 per watt to build today, First Solar will likely generate sales of $0.60 to $0.80 per watt next year. That would put revenue in the range of $1.9 billion and $2.7 billion, down from the $3.0 billion to $3.1 billion in sales expected this year. With non-GAAP operating expenses expected to be $330 million to $340 million this year, the company may need to keep gross margins in the high teens to remain profitable. And industry-leading margins may not be a slam dunk going forward. 

How great of an upgrade is Series 6? 

The biggest question I have about First Solar's future is, where will Series 6 efficiency end up versus the competition? First Solar has said that Series 6 modules will be at least 18% efficient in turning the sun's energy into electricity, up from 16.9% in production today and that in one to three years, it plans to be over 19% efficiency. But we don't know what efficiency of a mass-produced module will actually be or when the jump from 18% to 19% will take place. 

I will point out that First Solar's efficiency numbers aren't perfectly comparable to silicon-based solar modules. Thin-film technology squeezes more energy from a panel by performing better in hot and humid conditions and is less affected by shading, making it a better solar option in many desert applications. But efficiency is still an important proxy for First Solar because it no longer has a huge cost lead over competitors, as it did in the 2000s. 

The challenge is that commodity silicon cells are already 19% efficient or higher. SunPower (NASDAQ:SPWR) says it will make a P-Series solar module that's 19% efficient with commodity mono-PERC cells by the end of the year for the utility solar market. Manufacturers like JinkoSolar (NYSE:JKS) and Canadian Solar (NASDAQ:CSIQ) are also building out their mono-PERC capacity in the hopes of increasing efficiency to 18% to 20% in the next few years. Without an efficiency advantage, First Solar will be forced to compete on price and that will put a cap on margins.

The balance sheet isn't quite what it seems

The balance sheet is often seen as a huge strength for First Solar, and it is. Expected net cash balance of $2.1 billion to $2.3 billion is almost half of the company's market cap, which seems great for investors. But it doesn't account for capital expenditures that will be paid for in 2018. There is at least $500 million of capital expenditures to complete the Series 6 upgrade and that's probably on the low side. 

This isn't to say that First Solar doesn't have a great balance sheet. But when Series 6 upgrades are completed, I would expect net cash to be more like $1.5 billion rather than where we see it today. That's only about a third of the market cap, a little smaller cushion than it appears today. 

A hidden asset

There is one nice asset that may be underappreciated by investors right now. First Solar is a co-sponsor of 8point3 Energy Partners (NASDAQ:CAFD) and owns 15,395,115 shares of the company's stock. And with the stake up for sale, the company could get as much as $400 million in cash from it

Now that First Solar is moving to component sales, it doesn't need the yieldco anymore and that could lead to a nice cash infusion for the balance sheet. 

Where does First Solar go from here? 

You can see that First Solar's future isn't as certain as it may appear on the surface. The balance sheet is good, but 2018 could be a cash drain as Series 6 equipment is installed. A move to component sales could also lead to a drop in revenue as early as next year. And if Series 6 doesn't perform significantly better than Series 4, we could see competitors start to catch up with this solar giant and may even pass it on an efficiency and cost front. It may take two or three years to sort out where First Solar really stands in the highly competitive solar market. 

Travis Hoium owns shares of 8point3 Energy Partners, First Solar, and SunPower. The Motley Fool recommends First Solar. The Motley Fool has a disclosure policy.