SoftBank's $100 Billion Fund Is the Biggest in the History of Tech Investing. Has Venture Capital Changed, or Have the Companies?

Masayoshi Son made tens of billions of dollars investing in Alibaba. What has he set his sights on now?

August 11, 2017 will be SoftBank CEO Masayoshi Son's 60th birthday.

"Retire at 60" is part of a life plan Son says in interviews he wrote down at 19. In 2015 he picked his successor, a former Google executive of Indian descent.

Last June, Son abruptly announced at an earnings call that he would not retire yet: "I can't hold on to power forever, but I'd like to stay president for close to another 10 years."

SoftBank's stock jumped on the news. SoftBank has been a Yahoo shareholder since 1996. Alibaba, Yahoo, Shanda and Supercell are among the companies Son has backed over the years, for returns of more than $100 billion.

That one remark at the shareholders' meeting soon turned into a very large move: SoftBank would raise a $100 billion fund over the next five years.

Last Sunday, SoftBank announced that the fund had closed its first round of fundraising. Son made the announcement in Saudi Arabia, during President Trump's visit to the country.

The fund, called the SoftBank Vision Fund, is the largest private equity fund ever raised — four Silver Lakes, or 15 Sequoia Capitals.

Investment firms are the ones writing the checks, but capital does not come first and new companies second. The opposite is closer to the truth: over the past forty years, the size and shape of tech venture capital have changed as the way tech companies grow has changed.

This new $100 billion pool is no exception.

More money on hand, and bigger checks than anyone else

At a shareholders' meeting this past February, Son said the giant fund exists to catch the next explosion in technology: "We foresaw the big bang in personal computing. We foresaw the big bang in the internet. I believe the next big bang will be more violent still, and to be ready for it we need a platform — and that platform is the SoftBank Vision Fund."

SoftBank went public in Japan in 1994. In 1999, SoftBank Group became a holding company built around the internet business.

As Son tells it, SoftBank saw the internet coming; its investment in Alibaba returned tens of billions of dollars; and its mobile carrier business, then third in Japan, passed its rivals by being first to bring the iPhone to the Japanese market.

Since announcing the Vision Fund last October, Son has been busy in the oil states raising money. The first close has now reached $93 billion. Son says the rest will be raised within six months.

• $45 billion from Saudi Arabia's sovereign wealth fund;
• $28 billion from SoftBank;
• $15 billion from Mubadala, Abu Dhabi's sovereign wealth fund;
• $5 billion from five parties at $1 billion each — Apple, Oracle founder Larry Ellison, Qualcomm, Foxconn and Sharp (now owned by Foxconn).

To run the $100 billion fund, Son pulled in SoftBank CFO Rajeev Misra as its head. Before joining SoftBank, Misra ran global credit trading at Deutsche Bank.

SoftBank has also put together an 80-person team that includes SoftBank International CFO Alok Sama; board member Ronald Fisher, who was deeply involved in SoftBank's purchase of Sprint; and Deep Nishar, a former LinkedIn executive.

Son himself will play an unusual role. People familiar with the matter told Bloomberg that Son will be the fund's only "key man."

By way of comparison: Preqin's 2017 Global Private Equity & Venture Capital Report puts total global venture capital deal value for all of 2016 at just $134 billion.

Son told Bloomberg that the Vision Fund will not behave like a traditional venture firm: it will seek control of the companies it backs, with most investments taking 20% to 40% of a company. "We will be the largest shareholder, have a board seat and discuss strategy with the founding team," he said.

According to The Wall Street Journal, the Vision Fund already has more than a dozen deals in progress, each above $1 billion, with tens of billions of dollars more to come. From now on, any SoftBank investment over $100 million will be made through the Vision Fund.

A few of the investments are already public, and they are all big:

• Chipmaker ARM. The architecture behind the main processor in every phone on the market today comes from ARM. Last July, SoftBank bought it outright for $32.4 billion. In March 2017, SoftBank sold $8 billion of ARM shares to the Vision Fund, giving the new fund 20% of the company — counted as part of the $25 billion SoftBank is putting into the Vision Fund.
• Satellite maker OneWeb. Its goal is a satellite communications network. Last December, SoftBank announced a $1 billion investment in OneWeb; in February, it said it would put in $1.7 billion to merge OneWeb with another satellite company, Intelsat. Both deals will go into the new fund.
• WeWork. Shared office space. SoftBank invested $3 billion at a $17 billion valuation — $300 million from SoftBank itself and the remaining $2.7 billion from the Vision Fund.
• Didi. China's largest ride-hailing service, now preparing to expand worldwide. Didi recently raised $5.5 billion, $5 billion of it from SoftBank, invested through the Vision Fund. How much of the resulting stake belongs to SoftBank and how much to the Vision Fund has not been settled.
• Nvidia. The world's largest maker of graphics processors, now moving into artificial intelligence. SoftBank already holds 4.9% of Nvidia, worth about $4 billion, making it the fourth-largest shareholder. The Vision Fund may take a quarter of that stake.

Of the deals disclosed so far, ARM and Nvidia are public and profitable; Didi and WeWork are large but still burning cash by the fistful; OneWeb is relatively early.

What they obviously have in common is that all of them want to be the infrastructure of the future.

Son has come back to the ARM purchase again and again in his talks. He is bullish on the internet of things, and expects 80% of IoT chips to come from ARM. "A trillion IoT chips are coming, and every one of them needs a network connection," he said. Nvidia is the same story: as artificial intelligence advances, Nvidia's graphics chips supply the machine intelligence underneath one field after another.

Didi and WeWork are aimed at transportation and at office space. OneWeb is the communications technology that Google and Facebook are both betting on — a way to skip laying traditional fiber and put the internet into places like Africa quickly.

Tech companies did not always lean on venture capital the way they do now

This $100 billion is meant for tech companies. The eight largest listed tech companies in the world today are Microsoft, Apple, Amazon, Google and Facebook in the United States, and Alibaba, Tencent and Baidu in China.

They come from different eras. The oldest, Microsoft and Apple, were both born in the mid-1970s and drove the spread of the personal computer directly. Amazon, Alibaba, Google and the rest arrived with the birth of the internet. Facebook came later, after the internet bubble burst in 2000.

How they raised their money tracks how the growth path of a tech company has changed over forty years.

• Microsoft and Apple had almost nothing to do with venture capital

The relationship between tech companies and money was not always what it is now, where you need investment simply to start.

Bill Gates and Paul Allen wrote a BASIC compiler and earned their first revenue from it. In 1977, Gates dropped out of Harvard, took a job writing programs, and ran Microsoft on the side. Beyond paying people, Microsoft had no real expenses.

From its founding to its 1985 IPO, the company that would go on to own the PC operating-system market for decades never took a cent of venture capital.

Apple, founded in 1976, started its business with the Apple I, built by hand in a garage with help from the neighbors. A year later Apple was formally incorporated and brought in former Intel executive Mike Markkula to help write the plan, find investors and get the Apple II into production.

Markkula's own money plus a few investment firms came to $600,000. In China today, that is roughly the startup capital of any number of companies that deliver fruit to your door.

Markkula later said those firms could have put in another $5 million to $8 million, but Apple did not need it: the Apple II sold $2.7 million worth in its first year on the market. Three years later Apple went public and raised $100 million — an unremarkable number now, but for a time the largest public offering since Ford's in the 1950s.

• The giants born in the 1990s leaned on venture capital while they found a business model

Microsoft and Apple spread the personal computer and paved the road for the internet boom of the mid-1990s.

Amazon (1994), Google (1998), Alibaba (1998), Tencent (1998) and Baidu (2000) were all founded in that window.

Most of them, you can see, kept the lights on with investments of a few million to a few tens of millions of dollars while they worked out how to make money.

We put the fundraising dates and amounts of the world's eight biggest tech companies on a single chart. The money and the time a tech company needs between founding and IPO have both clearly gone up.

Google and Tencent both have stories about nearly selling themselves. In early 1999, Google's two founders tried to sell the company for $1 million, then cut the price to $750,000, and still found no buyer.

But both soon found their model. Google turned profitable on ad sales in 2001. Tencent launched QQ Games in 2003, along with a paid VIP service.

Amazon and Alibaba both used the internet to overturn traditional retail. Each scraped together its first money from family or partners, then grew on venture capital. Today the two are worth $480 billion and $300 billion.

Getting a company to profit costs more money now

The change starts with Facebook.

Facebook is the only one of them born after the bubble, at the exact moment internet investing had gone cold. Mark Zuckerberg started the company in his dorm room in 2004, and not long after took a $500,000 investment from Peter Thiel. Over the following years, the money Facebook raised came to more than the pre-IPO investment and IPO proceeds of Apple, Microsoft, Amazon and Google put together.

Before going public, it expanded as hard as it could to get more users.

On its first day of trading, Facebook was worth $104 billion. What followed was a stretch of doubt and a sagging stock that at one point fell below the offering price. Two years after the IPO, advertising revenue settled the argument, and Facebook is now one of the most valuable tech companies in the world.

Facebook's IPO was a turning point for American venture capital: after it, raising money got noticeably easier for internet companies. Facebook is regarded as the first "unicorn" — a company that does not chase profit at the start, but raises a great deal of money to capture users and lift its valuation. After it, there were more and more of them.

No investor wanted to miss the next Facebook, and all of them were willing to throw money at a startup to shorten its trip to unicorn status.

Startups that once took years to find a business model were forced to ripen early by capital. There is more cash in the industry now than at any point in the 1990s.

Uber, Didi and even Airbnb have each now raised more than Facebook raised in its whole history. In the past year alone, Uber and Didi each raised more than Facebook's lifetime total.

Facebook raised ten rounds before it went public, but each was in the tens of millions of dollars, and only the last three before the IPO passed $100 million. Uber and Didi, during the two years they fought head-on in the Chinese market, raised rounds of $1 billion and sometimes several billion dollars.

Facebook was different. Before going public it took a $550 million pre-IPO round from Goldman Sachs and DST. That money was for future growth; the company did not need investment to keep running. Facebook was already profitable, and every number — users, revenue, profit — was growing healthily. The one risk was that users were moving to phones while Facebook was still largely a PC website.

But Uber, which burned $2.8 billion in the past year, and Didi, forced to pull back in China by regulation, genuinely need to spend the billions of dollars a round brings them.

Startups' paper value grows faster

A 2013 essay by the well-known Silicon Valley investor Aileen Lee, “Welcome to the Unicorn Club,” made "unicorn" the shorthand for a billion-dollar startup: a company that starts from nothing and is worth more than $1 billion before it ever lists.

Then companies valued at $1 billion multiplied, "unicorn" stopped being useful, and Silicon Valley coined "decacorn" for private companies worth more than $10 billion.

Didi launched in 2012; by 2016, Uber and Didi each had $10 billion sitting in the bank. Apple, by comparison, raised $100 million in its IPO, and at the time that was the largest IPO in history.

A company that raises $100 million today may not be able to burn it for very long.

More money goes to companies that already have scale, not to new ones

Companies worth investing in seem to have grown scarcer over the past few years. The well-known Silicon Valley firm a16z made half as many investments in all of 2016 as it did in 2013.

CB Insights data shows that early-stage projects are finding it harder and harder to raise money, with capital tilting toward mid- and late-stage deals. By the fourth quarter of 2016, growth in the number of venture deals had slowed for four straight quarters, and the count had fallen back to 2014 levels.

The average size of each deal, meanwhile, keeps rising — a trend that started in 2010 and has held ever since.

Data from Cooley shows that in the first quarter of 2017 the number of deals fell 10% from a year earlier, while the total raised dropped 49%.

Reaching profit takes longer, too

Bill Gurley, an angel investor in Uber, has warned that the flood of capital keeps good companies and bad ones alike afloat, and that too much cash has pulled founders into unhealthy business practices.

Of the 20 most highly valued startups in the United States, seven have had their valuations cut by the firms holding their shares in the past two years. Some of them, including Palantir and WeWork, have also had layoffs or waves of departures.

For companies that have raised more than $1 billion, are valued above $10 billion, and still have no profit worth the name, going public is the only way out.

Yet startups now push their IPOs further and further back. By staying private as long as possible, they generally avoid having to operate under strict financial and operational scrutiny. They tend to take more money before listing, and their business models need more money before they can turn into sustainable businesses.

The first-quarter 2017 MoneyTree report, published jointly by CB Insights and PwC, shows that over the past two years the birth rate of new "unicorns" has not returned to the level of 2015, and that Europe produced a single unicorn in the past six quarters.

Fewer and fewer companies can make money by solving just one need

Didi and Uber are the best examples.

In the largest single round of last year, Didi took $7.3 billion at once. Didi did beat Uber in the Chinese market in the end. But under regulatory limits it will be hard for Didi to own China's ride-hailing market outright.

And by now, it is hard for Didi to stop.

Go abroad, work on driverless cars, study artificial intelligence — in short, since profit is not coming soon, it has to keep telling new stories.

A large round lets Didi support companies overseas, expand into financial services, or put money into driverless research, as its newly announced U.S. research institute will do.

The $5.5 billion Didi took from SoftBank and other investors is, roughly, for those reasons.

The main business does not make enough money. To hold up their valuations, these startups have to keep telling new stories: Didi expanding overseas, Meituan doing ride-hailing and short-term rentals, Uber building driverless cars — it is all the same logic.

Next to today's eight biggest tech companies, these startups have far more complicated business models.

Even Airbnb, in the best shape of the lot and closest to an IPO, cannot go public on one business the way Microsoft and Apple did. After planting itself firmly in short-term rentals, Airbnb announced it would do destination experiences, bought companies to build a luxury offering, built its own payments, and hired McKinsey to study the long-term rental market.

A startup that solves only one problem now has trouble going the distance.

Solving more problems takes more money.

The competition is not short of money either, so beating it is harder

These already-large tech startups are caught in a loop of competition.

After merging with Dianping, Meituan owned China's group-buying market, but that market no longer makes enough money. Through food delivery, Meituan now finds itself squared off against Alibaba and Baidu. China's food-delivery war has been running for at least three years.

Alibaba and Ant Group recently put another $1 billion into Ele.me, so Meituan's cash burn in this market is unlikely to stop.

The same thing is happening worldwide, and SoftBank's investments are no exception. On one side, Amazon put $5 billion into its Indian arm, squeezing the breath out of a string of Indian e-commerce companies SoftBank has backed. On the other, SoftBank has itself invested in ride-hailing companies in Southeast Asia, India, South America and China. In the United States, Lyft has picked up backing from carmakers and foreign capital. Together these companies keep Uber from owning any one market and reaching profit.

All of these changes point the same way: tech companies that have been at it for years and are already large on paper still need more time and more money before they can run normally and become the next Google or Tencent.

And the changes themselves cannot be separated from the push of capital.

Whether or not it can put the money to work in the end, the $100 billion fund SoftBank has set up is made for exactly this new world.

It is now hard to say whether companies decide what capital does, or capital is pushing the companies.

Charts by Feng Xiuxia

Chart data compiled by Xie Ruohan and Zhu Ruomiao

Cover image from Visual Hunt, EmreAyar via Visualhunt.com / CC BY, hdimagegallery