Why WeWork's Failed IPO Might Not Mean Disaster for SoftBank After All

The popular story is that Japanese investment company SoftBank Group Corp. faces a disappointing comeuppance—a worst case estimate of a $3.4 billion operating loss for the quarter that ended in September, according to Mitsubishi UFJ Morgan Stanley Securities Co., in part because of investments in Uber and WeWork.

Uber’s stock has been tanking (down 37% to just over $29 a share from the $46.38 high in July). But the biggest issue is WeWork and the cancelled IPO. The shared workspace company is struggling, and SoftBank founder and CEO Masayoshi Son is now “embarrassed” by his investing record.

But the magnitude of disaster WeWork represents for SoftBank could be grossly overstated and, under some conditions, the investment could even turn a profit.

A problem of valuation

The WeWork risk is over the slippery concept of valuation, which is completely subjective to market whims. Under standard accounting rules, companies must regularly revisit the value of assets on their balance sheets using real market transactions as proof. Otherwise, a business could artificially inflate values and lead investors astray.

Early in 2019, WeWork’s value was supposed to be $47 billion. Now? “Based on the rumored market feedback, it should be looking at a valuation of less than $10 billion,” says Dan Baker, director of equity research at Morningstar Investment Management Asia Limited.

WeWork has also needed ongoing cash infusions and still does, even with planned major cost cutting by its two new co-CEOs. Some banks might still lend money, although probably not as much as wished.

That seems a dilemma. SoftBank could buy more shares, at a lower valuation to protect itself, but that transaction would force a reassessment of previous investments in WeWork and possibly drive down the value. Yet, if it walks away, SoftBank would still be forced to write the investment down. Not good when SoftBank is raising money for a second investment fund.

But public speculation about SoftBank’s interest in WeWork turns on assumptions that it invested at top valuations. “Anything SoftBank does gets magnified,” says Nicholas Tsafos, an audit partner at EisnerAmper LLP and chairman of EisnerAmper Global. “We don’t know exactly what SoftBank invested at.” And it is those specific valuations that frame the actual risk to SoftBank.

SoftBank’s real investment

Fortune has learned that at least one analyst (who was not authorized to be directly quoted) was told by a SoftBank official that the investment giant has less exposure to WeWork than generally thought.

Atul Goyal, a technology equity research analyst at Jefferies, tells Fortune that he is “very surprised” to hear this. “We are pretty certain that it is about $10 to $11 billion that they have invested,” Goyal says. If WeWork’s value drops to $10 billion, how could SoftBank’s share represent anything but a loss?

The answer is complicated because of the opaqueness of the investment arrangements. (SoftBank did not respond to multiple requests for comment.)

Startup information site Crunchbase shows SoftBank participation in multiple WeWork funding rounds that totaled $10.4 billion. In the first and biggest round of $4.4 billion, the assumed valuation before the investment was $15.6 billion. The last investment, in January 2019, was $2 billion was assumed to be at a $47 billion valuation.

But half of that investment was a purchase of shares from employees and previous investors at a valuation of about $20 billion, according to the New York Times. As far as other investment rounds, SoftBank may have taken the common VC practice of negotiating down valuations. So long as WeWork’s valuation doesn’t drop below the valuation levels at which SoftBank obtained shares, there would be no absolute loss to the investment firm.

The opacity of the SoftBank investments further obscure the risk. “I did ask SoftBank IR for details of its WeWork investments several weeks ago but got no further detail other than that contained in the WeWork Prospectus, which unfortunately does not paint quite clear enough picture,” Baker says.

SoftBank’s potential escape vessel

According to the filed S-1 from the failed WeWork IPO, significant portions of SoftBank’s investment were in convertible preferred shares. Unlike regular common shares of stock, preferred shares have guaranteed dividend payments and act like bonds. But holders of preferred shares can also exchange them for common stock. That lets the holders either benefit from an appreciation in stock price or protect themselves from a price drop.

The valuation of preferred shares is different from that of common shares. They have the guaranteed payments that help prop up their value even if a company’s overall valuation declines. If WeWork keeps paying those promised dividends, the preferred shares help protect SoftBank’s investment, Tsafos says. And the S-1 shows that only about 20% of the shares held by “Softbank entities” were common stock meaning the other 80% were preferred shares.

In addition, the S-1 filed for the failed IPO indicates that some of SoftBank’s investments were in joint ventures of WeWork’s to operate in China, Japan, and the greater Pacific region. The joint ventures would be valued separately from WeWork and a sudden drop in the U.S. parent’s fortunes would not necessarily affect their value. Again, there is no way to know how much of SoftBank’s investments are in the joint ventures, in WeWork’s common shares, or in preferred shares that may or may not have been converted to common shares.

Then there SoftBank’s history of hedging bets. SoftBank’s Vision Fund did this with its roughly 5% investment in semiconductor company Nvidia. When shares dropped by more than 40% in 2018, SoftBank sold off its entire stake but offset almost the entire loss by complex derivatives on the Nvidia shares.

Private company shares have more limited hedging opportunities. But one that exists is a swap agreement. SoftBank would have gone to an investment bank and paid an “extremely costly” fee, according to Tsafos, for the right to sell shares to the bank at a predetermined price below the pre-IPO $47 billion valuation WeWork commonly received. It would have been a tempting deal for bank—many of which thought in 2007 and 2008 that betting against a drop in mortgage-backed securities was a sure thing—and a smart move for SoftBank.

There is no way to know from the outside whether SoftBank went this route. But given the Nvidia hedging and complexities of SoftBank’s investments in WeWork, by the time he dust clears, the WeWork investments may not be a total wash.