What SoftBank really wants

Some in Silicon Valley aren’t quite sure what to make of SoftBank and its massive, roughly $100 billion Vision Fund. At times, they say privately, it looks like a drunken gunslinger, firing off massive checks in quick succession.

But sources close to SoftBank say there is a method to its madness. In fact, these same sources say SoftBank’s investors believe they’ll see at least a 20 percent internal rate of return (IRR) over time from Vision Fund as it funds whole sectors being disrupted by artificial intelligence and machine learning — from pharma, to utilities, to ridesharing — and whose data SoftBank can leverage into an endless stream of opportunities.

The idea, these people say, is not to produce venture-like returns. The idea is instead to return more money to investors than private equity firms like KKR, whose first 18 private equity funds wound up delivering more than two times total capital invested on a gross basis, and produced a net IRR of 18.9 percent. Says one source close to SoftBank, “If someone is investing in [Vision Fund], he’s expecting to get better returns than with KKR and Blackstone.”

Indeed, 20 percent IRR over seven years — the time SoftBank estimates it will take most of Vision Fund’s bets to play out — is the “worst-case scenario” says one source. “Best case,” adds this person, is “investors get close to what Masa has done in the past.”

It’s a reference to the 44 percent IRR on investments that SoftBank can boast over its 18-year history, though more than one critic has noted that much of this number is rooted in SoftBank founder Masayoshi Son’s early bet on Alibaba, beginning in 2000. Son would eventually pour $58 million into the company; those holdings, which SoftBank maintains, save for a $10 billion chunk it sold to finance another purchase, are currently worth $130 billion.

Higher and higher

Doing back-of-the-napkin math on this 20 percent IRR — whether over seven years or a more traditional 10-year time frame — would translate into between $130 billion and $430 billion for SoftBank’s investors — minus its initial investments, management fees and the debt that makes up roughly $44 billion of Vision Fund’s total holdings.

That’s a whole lot of capital to generate for limited partners, so how does it do it? SoftBank thinks it can get there largely through ridesharing, say sources familiar with its thinking. More specifically, SoftBank is counting on the smooth evolution of today’s rideshare companies into vast networks of self-driving taxis.

It has already made an array of bets that underscore this theme, including on the China-based ride-hail giant Didi Chuxing; in Grab, the dominant ride-hail startup in Southeast Asia; and in Ola, India’s leading ride-hailing company, which reportedly closed on $2 billion just yesterday, including from SoftBank.

Helping grow a U.S. player is also crucial to its strategy, and SoftBank has been openly unsentimental about whether that means funding Uber or Lyft, though Uber would seem to be its strong preference. Says one source close of a meeting that’s slated to take place today, wherein Uber’s directors will vote on whether to go forward with a $10 billion stock sale to SoftBank: “Uber should be scared of SoftBank funding Lyft. They better take [the money].”

We’ll see soon enough how scared or not Uber may be of scorning SoftBank. Certainly, though, concern about the companies that SoftBank doesn’t fund is growing in Silicon Valley.

Asked onstage earlier this month about SoftBank’s impact on Silicon Valley, venture capitalist Steve Jurvetson of DFJ called SoftBank a “kingmaker of sorts that’s giving a massive infusion to some companies and not others.

“In the long run,” Jurvetson said, “that’s just noise. The better product and service should win. In the short run,” he continued, “it could [create] some interesting shifts in the outcomes of companies that would otherwise be in a normal horse race.”

Another source who invested recently in the two-year-old, indoor farming company Bowery Farming, said he was taken aback when SoftBank led a $200 million investment in a competitor, Plenty, just one month after Bowery closed its latest round. Bowery has raised $31 million from VCs. “It definitely gives you pause,” says this investor.

Down to the wire

Still, even as SoftBank barrels forward — Vision Fund has so far deployed $20 billion, including investments in U.S. chipmaker Nvidia and the co-working juggernaut WeWork — questions over its pacing and strategy remain.

Last week, at a small event hosted by this editor in San Francisco, venture capitalist Megan Quinn of Spark Capital acknowledged that there have been times when a company is talking with firms like hers about a sub-$100 million round, and SoftBank has entered the picture “and is like, here’s $200 million!”

Quinn likened the deals to “baby buyouts,” saying that SoftBank is buying more of its portfolio companies than do traditional investors, sometimes because “SoftBank has been able to convince them that something that looks like a baby buyout is actually the right round for them.”

Other times, she added, “the company [that’s fundraising] has not been able to raise from traditional Series C investors, so they’re looking for something more meaningful — someone with more ownership opportunity.”

Asked about these comments, one source close to the Vision Fund tells us, “If you can show [SoftBank] where it can invest $10 million and make a billion, who is going to say no? But how many opportunities are out there like that? If you think about size of the fund and you think about what’s going to move the needle, then it’s fair to say [its team] has to invest a few hundred million in something. Otherwise, it’s not worthwhile.”

Some say that no matter what SoftBank’s aim — whether to produce venture- or instead private-equity-like returns — pouring so much money into tech over a relatively short period is a strategy that’s difficult to grasp.

Speaking alongside Quinn at the same gathering, investor Jules Maltz of Institutional Venture Partners meanwhile questioned SoftBank’s math, saying it “has some challenges.

“If you think about trying to return $100 billion, you don’t just want to get the money back; you want to make a return on that,” said Maltz. “So you probably want $200 billion, meaning you want to double your money, which is good.”

Still, said Maltz, “after fees and all that, SoftBank’s investors [get less than $100 billion in profit]. When you then think that [SoftBank’s] average ownership [stake] is between 15 and 20 percent, and you think about how much liquidity they have to generate in terms of the companies sold in that fund just for them to hit the relatively modest returns of doubling the fund, you start realizing they need more than $1 trillion of market cap in their companies.”

That’s possible if “they get the next Alibaba, the next Facebook,” said Maltz. “And I think that must be what they are going for . . . but . . . I don’t think they can do it.”

Out of the fire

VC Jeff Bussgang of Flybridge Capital Partners seems to suggest it doesn’t matter — that given how fast the money is being put to work, investors should expect that SoftBank will be raising another $100 billion fund soon enough.

“It’s incredibly inflationary when that much capital flows into the system,” says Bussgang, who says SoftBank routinely pays twice the prices that even deep-pocketed mutual fund investors are willing to pay because it can.

Thanks to that 20 percent IRR that it’s targeting, “[Vision Fund] can pay what they’re paying. They’ve promised their LPs a lower return threshold, so they can underwrite to those lower returns while projecting much bigger outcomes.”

What “people don’t appreciate,” adds Bussgang, is that this is just the beginning. “Son is providing a way for sovereign wealth funds to invest $2 billion at a time into an asset class they can’t otherwise access,” he says.

“If they want to invest through a Spark or an IVP, the most [of their capital that they can invest through the funds] is, say, $50 million. Vision Fund meanwhile [accommodates] a flow of capital that hasn’t had a way of accessing this highly fragmented market.”

Unsurprisingly, those close to SoftBank’s thinking say that the fund is simply misunderstood.

Asked, for example, about the giant checks that Vision Fund is writing, and whether these could limit the upside for its portfolio companies’ employees, limit the companies’ exit options or promote the kind of reckless spending that has killed many a promising company, these loyalists argue that SoftBank is anything but careless in its check writing, noting that SoftBank typically demands board seats and certain rights that protect its investments.

“I don’t think it’s overfunding companies,” says one source. “The team is definitely holding companies’ feet to the fire to ensure they allocate their resources the right way.”

It’s also not funding as many companies as the press would have people believe, says this same source, saying specifically of the self-driving car startup Zoox that SoftBank will “not fund it in this lifetime.” (A report last month spawned many more reports that the company was in discussions with SoftBank for a round that could value the company at upwards of $4 billion.)

Zoox, which reportedly hired a group of 17 Apple auto engineers in late August, after Apple abandoned its own plans to build a vehicle, declined to comment for this story.

Watch it go, bling!

Despite the massive checks it’s writing, SoftBank has already appeared somewhat fallible at times.

Last week, for example, just one month after leading a $1.1 billion investment in a biopharma holding company called Roivant, a widely awaited drug candidate that would have been a blockbuster for one of Roivant’s subsidiaries, Axovant, was deemed ineffective.

Axovant’s formerly high-flying shares tanked on the news; Roivant remains the company’s biggest shareholder.

It was a crushing development for Roivant. Yet in what could serve as a reminder that it’s too early to judge SoftBank’s strategy, Roivant received more promising news yesterday, when a Phase 3 trial in which another of its subsidiaries is involved produced positive results.

Says one source close to SoftBank, “If mistakes have been made so far, no one will know for another 18 months.”

This person points to Nvidia, whose value has risen since SoftBank invested $4 billion into the publicly traded company in May, when the  company’s shares were trading at $137 apiece. (They now trade at $180.) “Whatever is visible,” he added, “has made money.”