Bloom Energy’s filing for an initial public offering this week has offered long-time watchers of the well-funded Silicon Valley fuel cell firm reams of hitherto unavailable data. 

Now it’s up to would-be investors to decide whether the data paints a clear picture of how Bloom will turn its $1.5 billion in equity investment into a public company that won’t also collapse under the weight of its business ambitions. 

Here’s our take on the key data points in Bloom’s S-1, starting with the top-line figures.

A $2.3 billion deficit and no profitability for the foreseeable future 

Bloom’s financials reveal that the company has continued to face large and significant losses in the past two years — $263 million on revenues of $376 million in 2017, and $279.6 million on revenues of $208.5 million in 2016. (Interestingly, the company told The Wall Street Journal that 2016 was a profitable year.)

In the first three months of 2018, it reported a net loss of $17.7 million on revenues of $169.4 million, compared to a net loss of $59.5 million on revenues of $72.2 million in the same quarter last year. 

Bloom had an accumulated deficit of $2.3 billion as of March 31, 2018, and expects “that we will incur net losses on a GAAP basis for the foreseeable future.” In other words, Bloom has no intention of quickly overturning former GTM Editor-in-Chief Eric Wesoff’s dictum that the number of profitable standalone fuel cell companies stands at zero. 

Bloom also revealed that its cash reserves fell from $157 million at the end of 2016 to $88 million at the end of March, with a burn rate of $37 million in the first quarter of 2018. This indicates the company’s need for investors’ cash, particularly in light of its debt situation. 

$950 million in debt — and not much room to raise more 

Bloom is also carrying a lot of debt. “As of March 31, 2018, we and our subsidiaries had approximately $950.5 million of total consolidated indebtedness, of which an aggregate of $593.7 million represented indebtedness that is recourse to us,” the company wrote in its S-1.

Of this, $356.8 million represented debt of Bloom’s PPA entities — the special-purpose vehicles it created under its Bloom Electrons program, which hold and manage their debt separately from Bloom itself. Another $245 million represents debt under Bloom’s “6% Notes,” which were issued by J.P. Morgan and Canadian Pension Plan Investment Board in late 2015 and transferred entirely to CPPIB last year.

Another $249.4 million represented debt under Bloom’s “8% Notes,” the subordinated secured convertible promissory notes issued to “certain investors” between December 2014 to June 2015. Of those notes, $215.9 million will convert automatically into Class B common stock immediately before Bloom’s IPO. 

Then there is the $94.8 million in senior secured notes issued in June 2017, the “10% Notes,” which come with certain “restrictive covenants” that could make it more difficult for Bloom to raise debt in the future. Key to those is a “minimum collateral test” that could restrict the company’s expansion outside its PPA subsidiaries: “If we do not meet the minimum collateral test, we cannot invest cash into any non-PPA subsidiary that is not a guarantor of the notes,” the S-1 notes. 

$1.5 billion in equity financing and voting control for existing investors

According to Bloom’s S-1, “as of March 31, 2018, the Company has completed several rounds of private financing with gross proceeds totaling approximately $1.5 billion.” That figure puts a floor on what the company’s investors are presumably hoping to recover from an IPO. 

Bloom’s most high-profile investor is Kleiner Perkins Caufield Byers, which took its first stake in 2002, and now owns 15.85 percent of its shares. Other major private shareholders include the Kuwait Investment Authority with 10.73 percent of shares, New Enterprise Associates with 8.78 percent, Alberta Investment Management Corp. with 7.53 percent, and Advanced Equities Financial Corp. with 6.55 percent. 

Other investors include Goldman Sachs, Credit Suisse Group, Morgan Stanley, GSV Capital, Apex Venture Partners, Mobius Venture Capital, Madrone Capital and SunBridge Partners.

Bloom’s S-1 presumes that outstanding shares of its convertible redeemable preferred stock will be converted into an aggregate of 107,610,244 shares of Class B common stock, and that another 8,382,757 Class B shares will be issued to the investors holding convertible notes. These Class B shares have 10 votes per share, while the Class A common stock being offered to investors has one vote per share. 

This, plus voting agreements between CEO KR Sridhar and “certain stockholders,” “will limit or preclude Class A stockholders’ ability to influence corporate matters while the dual class structure remains in effect,” the S-1 notes. This includes “the election of directors, amendments of our organizational documents, and any merger, consolidation, sale of all or substantially all of our assets, or other major corporate transaction requiring stockholder approval.”

A $158 million hit to 2017 revenues: How the ITC affected Bloom’s fortunes

As we’ve noted in our Bloom IPO coverage, a big factor in the company’s poor performance last year, and in its hopes for a revival this year, lies in the federal Investment Tax Credit for its systems. “We expect that any Energy Server deployments through financed transactions (including our Bloom Electrons programs, our leasing programs, and any third-party PPA programs) will receive capital from financing parties who derive a significant portion of their economic returns through tax benefits (tax equity investors),” Bloom wrote. 

After Congress failed to include fuel cells in the list of approved technologies to be reauthorized for the ITC, the credit expired at the end of 2016. “In order to offset the negative economic impact of that lost benefit to our customers and financing partners, in 2017, we lowered our selling price to customers,” Bloom wrote. “Because many customers or financing partners would monetize the tax credit upfront, the actual impact to our selling price was generally greater than 30 percent.” 

The result for Bloom was a $158 million hit to revenues in 2017 as it cut prices and saw reduced business in the wake of the ITC’s expiration. But the retroactive reauthorization of the credit passed by Congress in February helped provide a $43.9 million one-time product revenue benefit in the first quarter of 2018, and has played a key role in the company’s decision to follow through with its IPO this year. 

The fading role of California’s SGIP in Bloom’s business 

Bloom acknowledges that its business “depends on the availability of rebates, tax credits and other financial incentives,” including state incentives in California, Connecticut, Massachusetts, New Jersey and New York. Of these states, Bloom has deployed most of its fuel cells in California, where the state’s Self-Generation Incentive Program has helped cover a large portion of the upfront cost of its fuel cells along with batteries, thermal energy storage and other distributed energy resources. 

But recent changes to the SGIP, including limits to how many awards any one company can receive and requirements that an increasing number of fuel cells be powered by biogas to receive the incentive, have reduced the opportunity for Bloom to capture this funding stream. The company’s billings derived from customers benefiting from the SGIP represented approximately 36 percent of total billings for 2016, but only 12 percent for 2017, and 18 percent for the first quarter of 2018, the S-1 reported. 

(A side note on biogas: While Bloom’s fuel cells are only carbon-neutral when using biogas, only about 9 percent of the company’s units were using it, rather than natural gas, as of March 31, 2018, the S-1 reported.) 

More than half of revenues from two biggest customers 

Bloom’s customers include 25 of the Fortune 100 companies, including AT&T, Equinix, Home Depot, Kaiser Permanente and The Wonderful Company. But it relies on a few key customers for a majority of its revenues, and that reliance is growing. 

In 2017, 53 percent of its revenues came from two customers: Southern Company, whose subsidiary Power Secure partnered with Bloom in 2016, at 43 percent; and Delmarva Public Service, the Pepco subsidiary that installed 30 megawatts of fuel cells, at 10 percent. 

In the first quarter of 2018, Southern Company made up 53 percent of total revenue, driven by a big 37-megawatt data center deployment, while new customer Korea Energy represented 17 percent. Bloom’s product sales backlog of 108.2 megawatts also shows its reliance on big customers, with four customers accounting for about half of that total. 

A move away from Bloom Electrons and PPAs in favor of direct orders

The other key factor to note in Bloom’s backlog is its almost complete reliance on direct sales to customers, at 98 percent of future business, versus its alternatives such as PPAs, leases and managed services, which stand at zero, zero and 2 percent, respectively. That’s a big shift from 2016, when the company’s portion of acceptances of revenues for direct purchase contracts stood at only 10 percent, compared to managed services at 31 percent, and Bloom Electrons at 53 percent. 

But over the past two years, Bloom has been shifting away from these options, which recognize revenue “ratably” over the lifetime of the contract, and toward direct sales, where revenue is recognized upfront. “Starting in 2018, the vast majority of our revenue is recognized at the time of acceptance,” the S-1 noted. 

This also means that Bloom Electrons — the 2010 program that financed customers’ installation and use of fuel cells through power-purchase agreements and has driven about $1.1 billion in financing through its five distinct PPA entities — is no longer part of Bloom’s focus. Bloom’s S-1 notes that “we expect no additional future investment in Bloom Electrons,” meaning that “electricity revenue and related costs will stay relatively constant for many years into the future."