In the lobby of SolarCity’s glass and stone headquarters in San Mateo, Calif., six young men in Dockers and solid shirts—Silicon Valley’s uniform—sit waiting, glued to their smartphone screens. A middle-aged manager ambles through a door into the waiting area and announces, “I’m here for the new hires.” All six stand.

This is a typical day at SolarCity, which is hiring more than 350 engineers, technicians, and salespeople a month and now employs a total of 7,500, more than Facebook. Launched eight years ago, the company has become the nation’s largest installer of residential solar systems. So far SolarCity has installed systems for more than 140,000 customers (it’s No. 2 in commercial installations), and its sales, which doubled in the latest quarter from the year before, are on track to hit $250 million in 2014, according to Credit Suisse. Wall Street has noticed. Since going public two years ago, the company has seen its stock jump from $8 a share to a recent price of $59.62—giving it a market cap of $5.3 billion—even though it has yet to post a profit. (It is reinvesting all its cash flow in growth.)

With economies of scale the cost of solar power has plunged much faster than many people expected—and some are actually arguing that solar will become cheaper than fossil fuels within the next few years. Now the difference is only about 25% after subsidies, a margin so small that people like Lyndon Rive, the co-founder and CEO of SolarCity, can taste victory. As he stated flatly in an exclusive interview with Fortune: “The core mission of the company is selling clean energy at a lower cost than fossil fuels. Period.”

SolarCity SolarCity warehouse in Foster City, CAPhotograph by Gregg Segal for Fortune

The question is, How does he get there? Can he? Is there snake-oil salesmanship afoot? Some of the shorts seem to think so—they now hold 22% of the float of SolarCity’s stock. On the one hand, Rive is a guy who as a teenager ran a homeopathic-medicine business, which really is snake oil in the modern sense. On the other, he’s part of an illustrious business dynasty as the cousin of Elon Musk, who was himself dismissed in the early days of Tesla as a hype artist, and who made doubters eat their words.

Solar power still amounts to less than 1% of the nation’s electrical-generating capacity—coal produces about 40%—and its proportion will stay in the low single digits until it becomes cheaper than fossil fuels. Rive, a hyperenergetic, fast-talking 37-year-old, says he has a plan to get there, which includes building one of the world’s largest solar-panel plants in Buffalo—a move even many of his supporters say is crazy at a time when solar manufacturers have been struggling. He has also sped up the time to install a solar system on a house from two days to, typically, a half-day, and is slicing the cost of acquiring customers. Says Rive (rhymes with “five”), a rough-hewn executive who favors shirts with the yellow-and-white SolarCity logo on the pocket: “The future will be primarily solar. Just do the math. The industry is growing at 40% to 50% a year. At that rate 50% of our electricity will be solar in 15 years.” That may sound as though he’s smoking something, but in September the International Energy Agency, a group not known for its green leanings, issued a report predicting that by mid-century, solar would be the “dominant” source of our electricity, accounting for 27% of production—more than coal, oil, wind, or anything else.

Gov. Andrew Cuomo, Lyndon RiveNew York Gov. Andrew Cuomo with Rive at the groundbreaking for SolarCity’s manufacturing plant in BuffaloPhoto: Courtesy NYS Governor’s Office

Perhaps better than anyone, Rive understands that this won’t be easy. SolarCity’s success so far has come quickly. Last year it controlled about a third of this $8 billion business. Now great forces are lining up against him: a utility industry threatened by a technology that is siphoning off customers, and the gradual expiration of federal subsidies for solar. Little wonder that short-sellers are circling.

On their way to Burning Man in an RV, Musk set his cousin Rive on a new path with one-word suggestion: “solar”

Saved by Underwater Hockey
Rive comes from a long line of South African adventurers. His great-grandmother was the first female chiropractor in South Africa, and his grandfather was an African explorer. When he was in high school in Pretoria, Rive started a business distributing homeopathic medicines. At 17, he got into trouble with his principal for skipping classes. According to Rive, the principal thought he was goofing off and called him into his office to threaten expulsion. Rive said he couldn’t go to school because he had to run the company: “I showed him my financials,” recalls Rive, “and I said, ‘Let me get this straight. You want me to leave all this so I can finish high school, go to college, and then get a job where, if I’m lucky, I’ll be making one-quarter of what I’m making now?’” They made a deal: Rive could stop attending classes and graduate if he could pass his exams. He did.

Rive’s mother and Elon Musk’s mother were twins. Musk—a founder of PayPal, SpaceX, and Tesla—is the chairman of SolarCity and owns 22.5% of its stock. In the late 1990s the family saw opportunity in America, and Musk was the first member to leave South Africa for California. Then Rive’s mother left, and then his brothers, and finally he followed in 2000 with a bankroll he had made from his homeopathy business. With his brother Peter, with whom he was later to found SolarCity, he launched an IT outsourcing software company called Everdream.

Rive’s strategy in starting a business is to seek out a niche. He saw that players like IBM and Hewlett-Packard were doing IT outsourcing for big corporations but sensed there was an opening to service PCs and laptops for smaller companies. After surviving the dotcom meltdown in 2001, Rive built Everdream into a thriving business. Then, in 2004, Rive and his cousin Elon were driving to that geek lollapalooza Burning Man in an RV, and he told Elon that he was “sick of the grind of doing IT support.” Musk simply said to him: “solar”—sort of like Dustin Hoffman being told “plastics” in The Graduate. Only this time the result was better.

When Rive and his brother started planning the business in a spare bedroom at Peter’s San Francisco house, they searched for another niche to fill. “We spent two years looking on the value chain,” Lyndon recalls. “At the time most people were focused on solar manufacturing and almost no one on delivery. We started the company with the idea of making solar affordable.” At that time U.S. and foreign manufacturers of solar panels abounded, but no major installer existed—just mom-and-pop players. By the time Dell bought Everdream in 2007 for $120 million, Rive had already launched SolarCity. Its first official day was July 4, 2006, to symbolize independence from fossil fuel.

If it weren’t for a twist of fate a few years earlier, Rive wouldn’t have been around to start SolarCity at all. He and his wife, Madeleine, are underwater hockey players. The sport is like ice hockey except the checking and scoring take place at the bottom of a swimming pool. Two teams of six players hold their breath for as long as they can and maneuver a three-pound puck around the pool using a footlong stick.

Lyndon Rive playing underwater hockey Lyndon Rive (with puck) playing underwater hockeyPhoto: Courtesy of SolarCity

Rive and his wife were in the U.S. on J-1 visas that were running out. It was right after the 9/11 attacks, and the Feds were not renewing many visas, so the Rives were about to be deported. Around that time they happened to attend a practice session in San Jose, where they met an Asian swimming coach who explained that she got her visa through a clause in the immigration law that allowed athletes, musicians, and actors who had “exceptional ability” to stay. Both Lyndon and Madeleine applied, but only Madeleine, who had played for both the U.S. and South African national teams and had won many national competitions, was granted a visa. “She was better than I was,” admits Rive. As a spouse, he was allowed to remain in the country to practice his entrepreneurial skills.

Bringing Power to the People
The notion that solar power could overtake fossil fuels is disputed, of course, by big oil companies like Exxon Mobil and Shell. Exxon’s research shows that 75% of the world’s energy will still come from fossil fuels in 2040. After all, $28 trillion—yes, that’s trillion with a “t”—worth of oil, gas, and coal remains underground.

Rive argues that it should stay underground, that the oil majors look at the world through petroleum-coated lenses, and that the benefits of solar are so obvious that once the price becomes more competitive, fossil-fuel use will plunge. A carbon tax, which Rive favors, would hasten the adoption of solar, but he isn’t waiting around for Washington to act.

A lot of investors are betting that his vision is right. The research firm Mercom Capital Group says that private equity firms and venture capitalists are on pace to provide $5 billion in financing this year for residential and commercial solar projects, up from $3.3 billion in 2013. Nat Kreamer, chairman of the Solar Energy Industries Association and CEO of Clean Power Finance, which provides funding for solar projects, lays out the potential. Of the 125 million houses in the U.S., about 56 million are suitable for solar (some are too shaded or have other impediments). Kreamer says that adds up to a $1 trillion untapped market for solar installers like SolarCity. Add in warehouses, factories, and other commercial rooftops, and that’s another $200 billion.

RIV-10.27.14

In Rive’s vision, American homes—SolarCity has yet to venture outside the U.S.—will not only generate their own clean electricity but also have a Tesla plugged into a solar-powered charger in the garage. In addition, homeowners will have the option to store electricity generated from solar, solving the sticky problem of what to do after the sun sets. The company has a pilot project offering small backup storage batteries made by Tesla that, like emergency backup generators, allow homeowners to run some appliances and lights if the power goes out (they could also help utilities manage loads). The batteries are still too expensive to provide enough storage to run all the lights and appliances in your house all night, but affordability might not be too far off. Musk’s Tesla is building a $5 billion battery plant in Reno that could achieve the economies of scale to make home storage affordable.

Peter Rive, Elon Musk, Lyndon RiveSolarCity COO Peter Rive, chairman Elon Musk, and CEO Lyndon Rive at Nasdaq exchange for the company’s IPO on Dec. 13, 2012Photo: Mark Von Holden—AP

Standing in the way of Rive’s vision are utilities that fear that the technology is putting pressure on their century-old business model of big centralized energy generation. Some utilities are starting to resist the implementation of solar because they say that while their costs for operating power plants and grid systems remain the same, they are beginning to lose revenues from solar customers who no longer need to buy energy from them. Arizona, for example, has proposed taxing solar installations. SolarCity is fighting back, suing to keep the state from imposing “grid” levies.

Dickon Pinner, a McKinsey director who heads up the consulting firm’s clean-tech practice, says, “The utilities for the first time are feeling that their business model is being threatened by distributed-power systems like solar and are starting to respond.”

The key is regulation. Most utilities operate in a monopoly market and enjoy a regulated 10% return on their invested equity. Rive calls this “insane.” As long as utilities stay regulated, it’s hard to see what would force them to innovate on cost and technology.

Making Solar Cheaper Than Fossil Fuels
In the long run, all this wrangling with utilities won’t do Rive much good if he can’t beat them on price. Installation companies such as SolarCity have two basic costs. One is solar panels themselves. The other is the cost of selling and installing them, which includes electrical hardware, labor, permits, interest payments, and overhead such as customer acquisition.

The good news is that over the past few years the price of solar power has dropped dramatically—the cost of residential installations has fallen 60% since 2000, according to the Lawrence Berkeley National Laboratory. Even so, an installation on a typical house still costs about $27,000. That’s a lot of money for most homeowners to shell out upfront. The innovation that has allowed the industry to take off is no-money-down solar leases. Today, according to Green Tech Media Research, some 70% of residential systems being installed are financed with leases. (The rest are bought outright or with loans.) With a lease (or a variant known as a power purchase agreement), the homeowner essentially buys electricity from the installer, which owns the equipment and does the maintenance. Deals usually pencil out to about a 10% to 20% savings on a typical utility bill. The homeowner gets clean, cheaper energy for 20 years. The installer gets a 20-year dependable cash flow.

Lyndon Rive Rive before photo­voltaic panels at the SolarCity warehouse in Foster City, Calif.Photograph by Gregg Segal

In most of California, SolarCity now charges its residential solar customers a flat rate of 15¢ per kilowatt-hour—one of the lowest in the solar industry for the residential market. (A kilowatt-hour is enough juice to run 10 100-watt bulbs for an hour.) But that’s not good enough. Today electricity in the U.S. costs about 12¢ a kilowatt-hour. That 12¢, however, is an average price. In some coal states in the Midwest and the South, electricity costs even less than that, and it will be a long time before solar is competitive in those regions. In the 14 states where SolarCity operates (including California, Arizona, and Massachusetts), the average price is more than 15¢.

Complicating the price picture is that starting in 2017, the 30% federal investment tax credit (ITC) is scheduled to drop to 10%. (Some in the industry believe the 30% rate will be extended, but SolarCity can’t count on that happening.) To stay even at a 10% subsidy, the company will have to cut costs about 13%. By 2020, Rive aims to get another 20% or so of costs out, which would let him hit that target of 12¢ a kilowatt-hour without subsidies.

How is he going to do it? One place SolarCity is looking for cost savings is in the solar panels themselves. Over the past few years the cost of photovoltaic solar panels has dropped some 70%. Starting about a decade ago, the Chinese, who now control 90% of the photovoltaic (PV) manufacturing market, built up huge overcapacity that has been driving down prices. Over the past two or three years a bloody shakeout has occurred, with many solar manufacturers going bankrupt, including some in the U.S. (like Solyndra, which received $535 million in Department of Energy loan guarantees). Today the price of panels has stabilized at about 70¢ a watt, according to Bloomberg New Energy Finance. To put that in perspective, in the mid-1970s, PV panels cost $79 a watt. With economies of scale there’s still a lot of room to drive down the price of the panels.

SOLAR PRICES PLUNGE Since the 1970s the cost of photovoltaic panels has decreased from nearly $80 a watt to an average of $.63, driving a solar boom worldwide. SOLAR PRICES PLUNGE Since the 1970s the cost of photovoltaic panels has decreased from nearly $80 a watt to an average of $.63, driving a solar boom worldwide.Graphic Source: Paul Maycock; BNEF

In June, Rive shocked the solar world by announcing that SolarCity would acquire Silevo, a California maker of solar panels, for $200 million in shares, plus another $150 million if cost and volume targets were met. Many in the industry thought this was either a gutsy move or an insane one. Why would SolarCity, which now buys its panels from multiple vendors, want to get into a cutthroat business dominated by the Chinese, where profits were negligible at best? First, Rive is worried about supply. At its current growth rate SolarCity will need enough panels to produce two gigawatts of electricity by 2017. Using Silevo’s technology, SolarCity is building a massive manufacturing plant in Buffalo that will be able to make about one gigawatt worth of panels a year. Raymond James analysts estimate that it will cost between $350 million and $400 million. The next facility Rive plans to build, he says, will have a production capacity of between five and 10 gigawatts. “Now you get true economies of scale,” he says. SolarCity broke ground on the factory in late September, aided by $750 million in subsidies from the State of New York.

Rive argues that when the factory is up and running it will be able to produce high-output PV cells at about 55¢ a watt—about a 20% reduction from current prices, which will help Rive get closer to his goal of making solar cheaper than fossil fuels.

Squeezing Costs
In a suburban neighborhood in San Mateo, I watch Victor Fernandez, the foreman of a SolarCity installation crew, direct workmen placing rectangular black panels on the roof of a ranch house. The workmen are using an innovative installation system devised by a company called ZEP, which SolarCity acquired last year for $155 million. Before ZEP, long metal rails were attached to a roof, and then each panel was bolted to the rails. With ZEP’s system the installer drives a few bolts through fixtures into the roof, and then the panels snap on like Legos. Says Fernandez: “We can now install two systems a day, as opposed to one every two days using the old system.”

As SolarCity races to cut its costs, some investors are starting to wonder about the company’s business model. As we’ve seen, SolarCity sells 20-year leases to its solar customers. GAAP rules don’t allow the company to recognize all the income from a 20-year lease in the year it’s sold. Investors, then, must value SolarCity stock based on the present value of that 20 years of cash flow from leases. The amount of cash flow, however, depends on some important assumptions. If someone sells his home, will the new buyers be willing to assume the solar lease? If not, what’s the value of a used solar panel? The company is saying most homebuyers will assume the lease, but the shorts believe SolarCity is being optimistic. Also, a significant part of the stock’s valuation is based on solar deals that have been booked but not delivered. Customers can back out of a contract without obligation, raising the question, in some minds, of exactly how many of those contracts are real.

The shorts talk about another issue: The Treasury Department is investigating whether SolarCity and other solar companies inflated the value of their installations to drive more tax credits to their investors. SolarCity says it has followed all the rules of the tax credit program.

Rive anticipates that his company will keep doubling in size for the foreseeable future. That’s hard to do for any company as it gets bigger, and the threat of competition—archrival Vivint raised $330 million in its October IPO—could make his task harder. But Rive says he’d like to see so much solar installed that his business is no longer a niche. “In that scenario humanity wins. Right now it isn’t winning.”And losing is something that Rive—and his competitive clan—can’t stand.


SECURITIZING THE SUN
How Investors can buy a piece of the solar boom.

Solar CityA SolarCity Corp. employee carries a solar panel being installed on the roof of a home in the Eagle Rock neighborhood of Los Angeles,Photo: Patrick T. Fallon— Bloomberg

SolarCity eats capital. Lots of it. This year the company will need to raise about $1.5 billion to finance its solar installations. Until now it has raised most of its funds from big banks and corporate investors such as Google. Solar­City typically pays these investors an 8% to 9% return on their investment. Recently Rive decided that the time was right to securitize his solar installations.

Think of solar bonds as mortgage-backed securities, only with—one assumes—a more dependable cash flow. SolarCity’s bundled securities are backed by 20-year solar leases held by homeowners, who pay their electric bills to SolarCity instead of to a utility. Solar bondholders are investing in this cash flow.

Unlike mortgage-backed securities, solar bundles do not include subprime loans. For homeowners to qualify for financing, they must have a FICO score of at least 680. In July the institutional investors that bought $201 million of SolarCity securities received 4.3% in annual interest. Says Nat Kreamer, head of the Solar Energy Industries Association: “Securitization is driving down the cost of capital and increasing its availability, and that’s the No. 1 priority of the U.S. solar industry.”

So far SolarCity’s securities are being sold only to so-called qualified investors—wealthy individuals and institutions. The company also plans to offer solar-backed securities to retail investors. In December, SolarCity acquired Common Assets, a web-based investment platform where individuals and organizations will be able to participate directly in solar investments.

There are risks. Interest rates are at historic lows, and a sharp spike in rates could drop the value of solar bonds. Also, though consumers typically stop paying their utility bills last, it’s not clear what the default rate on their monthly payments might be—and if SolarCity should ever go out of business, who would be around to collect those monthly fees?

SolarCity CEO Lyndon Rive believes that the risks are minimal and that solar bonds could supercharge the rate of adoption of this clean technology.

Rive’s timing might be right. Some university endowments are considering divesting fossil-fuel stocks. Investors everywhere are searching for higher yields. It could be a powerful combination.


This story is from the October 27, 2014 issue of Fortune.