what is a SoftBank?

and what can it tell us about technology and finance?

SoftBank is, by differing accounts, a Japanese mobile phone company, a vehicle for Saudi money and corporate intrigue, and an “options whale” spreading its wake through the market for tech stocks.

While that combination may seem bizarre at first glance, SoftBank isn’t interesting because it’s an aberration. It’s interesting because it is the prototypical global Big Tech company.

Before explaining why, it could be helpful to first discuss what exactly SoftBank is. Those of us who aren’t on the tech beat and were 12 years old during the 1990s dot-com boom might think of SoftBank as mainly a Japanese mobile-phone operator. But that is just one of its subsidiaries, called SoftBank Corp.

What we want to cover is SoftBank Group Corp., the company that Masayoshi Son started in 1981.

That company was founded as a personal-computer software distributor—a software bank, if you will—and soon became a publisher of tech trade magazines. Notably, one of Son’s first big wins was a Yahoo! investment that the company still credits to its mid-1990s acquisition of Ziff-Davis Publishing Co., publisher of PC WEEK. Other famous and infamous deals followed, like the stake in a young Jack Ma’s Alibaba, the 2006 acquisition of Vodafone’s Japanese business, and of course its massive stake in WeWork.

In short, SoftBank’s main business has always been to investigate and/or invest in tech companies. If we think of Amazon as a platform for stuff, Facebook as a platform for people and Google as a platform for information, we can think of SoftBank as a platform for tech, or even a platform for platforms.

So it probably shouldn’t have come as a surprise when SoftBank started investing its cash in publicly listed Big Tech firms earlier this year. Nevertheless, reports that SoftBank had made massive single-stock options trades led the company’s own shares to sell off sharply.

I must admit that selloff was a little confusing to me. Because while SoftBank was among the first to make platform building a business, it was actually the big U.S. tech companies that paved the way for financial-market investing. For years, they have hoarded massive cash piles and invested them in U.S. financial markets, from mutual funds to corporate credit to repurchase agreements.

The original cause for their cash hoarding was said to be U.S. tax law: Until recently, American companies have been able to defer taxation on foreign-derived income indefinitely until it was “repatriated,” or distributed to U.S. shareholders. (Unlike many other countries in the O.E.C.D, the U.S. still taxed income globally.)

That particular incentive was supposed to be removed with the new tax law in 2018, yet many big U.S. tech companies haven’t shrunk their cash piles.

For example, at the end of 2017 Apple had $285 billion in cash, cash equivalents and financial securities. It still had more than $190 billion as of June 27 of this year, with about 41% of that invested in corporate debt, 16% in Treasuries and agency bonds, 10% in other government bonds and 6.6% in mortgage- and asset-backed securities.

Microsoft had $133 billion in Sept. 2017, slightly less than the $137 billion it had on June 30, 2020. In 2017, the vast majority of that was invested in Treasuries, while today Microsoft says that it is invested primarily in highly liquid investment-grade securities with maturities of three months or less.

Alphabet, parent company of Google and YouTube, had $121 billion of cash and financial securities on its balance sheet as of June 30, up from $102 billion in 2017. Amazon and Facebook have also grown their cash piles in that time, with Amazon’s more than doubling to $71 billion and Facebook’s rising to $58 billion from $42 billion.

While Big Tech isn’t new to asset management, it should be noted that SoftBank isn’t either. It has been years since SoftBank closed its first round of fundraising for its Vision Fund, a Saudi-financed investment fund that is also on its balance sheet.

But that fund was created to invest in private companies, not public ones. And SoftBank appears to have built up its current cash hoard under some pressure from activist investor Elliott Management. Going by the FT’s reporting, SoftBank’s decision to build out its public-market investments was part of a shift that occurred early this year, when the company’s shares sold off during the coronavirus panic. As its shares dove, SoftBank assured investors that it would raise JPY4.5 trillion (or nearly $43 billion) in cash, and use up to JPY2 trillion (or $19 billion) on a share buyback program—reportedly the biggest in Japan’s history, and definitely the biggest in SoftBank’s.

The company’s recent divestment of U.K. chip maker Arm Holdings, which it sold to Nvidia for $40 billion, should bring its cash balance above its target, analysts say. So it is on track to fulfill its recent promises to investors.

Analysts and investors still seem nervous, however, largely because there are a couple of other ways that SoftBank’s internal asset manager differs from those run by the U.S. Big Tech companies.

First, Wall Street has been speculating that there could be some type of big end game for the cash, specifically taking the company private.

Second, even though SoftBank had reportedly made money on its options bets as of the start of this month, its bets do seem riskier than the typical corporate bond portfolio on the balance sheet of one of its U.S. Big Tech peers. For example, on Friday, there was at least one outstanding call spread in the red that was large enough to have potentially been SoftBank’s. It was a large bet that shares of Alphabet would be trading between $1,650 and $1,950 on or before Oct. 16. Alphabet’s stock was trading at $1,444.96 on Friday, leaving that trade—or more realistically, one leg of a broader trade—down nearly $73 million. To be fair, the trade was probably hedged, and we can’t know the full details even with the helpful info we got from Henry Schwartz (thanks Henry!), because the options trades could have been offset by positions in the underlying tech companies’ shares that weren’t linked to specific options.

And ultimately, any profit (or lack thereof) that SoftBank makes on its options trades isn’t the real news—it’s an exciting distraction, but a distraction nevertheless. The real story is that, in the span of six months, Softbank raised a pile of cash that was massive enough to buy back shares, pay down debt, appease an activist investor and make “whale”-sized bets in the options market.

In fact, Big Tech’s balance sheets have grown so large that even economists at the Bank for International Settlements—the global central bank for central banks—are starting to pay attention.

In a Sept. 22 paper, a team of researchers pointed out that while much of the “fintech” attention gets directed towards peer-to-peer lending platforms and blockchain technology, U.S. companies such as Amazon and Apple have been providing credit directly to individuals and companies. Apparently they have only provided about $8 billion of credit, in the BIS’s view.

Clearly that sum doesn’t include tech companies’ large bond portfolios. On one hand, that could make sense, because owning a bond isn’t “providing credit” in the same way that a bank provides it—the BIS’s researchers did also include fintech in this paper, so they may have focused more on non-bank debt underwriters like Greensill. (By the way, for folks who are interested in supply-chain finance, my colleagues in London at Financial News wrote a really nice profile of that firm’s founder last year.)

On the other, the BIS appears to consider bank-partnered credit cards, like the Apple-Goldman Sachs venture, to be “credit provision.” And that’s where the line gets fuzzier. What is the difference between buying a bond that was underwritten by a bank and partnering with a bank to provide a credit card? And barring any elaborate credit-insurance arrangements, isn’t Apple exposed to most or all of the credit risk of any given bond in its portfolio?

And if so, shouldn’t tech companies’ billion-dollar bond portfolios count as credit provision as well?