SoftBank scrutinises startup bookkeeping after WeWork struggle
Chinese artificial intelligence company, SenseTime Group Ltd.‘s developers, travelled to Tokyo in early 2018 to see SoftBank’s Masayoshi Son, a billionaire investor. Chief Executive Xu Li wanted to convince the president of SoftBank Group Corp. to spend $200 million in his three-year-old start-up when they reached the offices.
Son stopped a third of the way into the briefing to claim he wanted to put in $1bn. Son proposed $2 billion a few minutes later. Turning to the roomy bosses of SoftBank, Son said that this was the sort of AI agency he was searching for. “Why do you just question me about them now?”After one person in the room, he questioned.
Eventually, SoftBank has spent $1.2 billion to help turn SenseTime into the most successful AI start-up in the world. This year’s value of the young company reached $7.5 billion.
Since Son, 62, raised the capital in the office-sharing company WeWork Cos, the investment paradigm is now under threat. Inc. to see it crumble as shareholders baulked at huge losses and alarming leadership. Nonetheless, along with other investors, SoftBank has engaged in dozens of crowdfunding which, according to Bloomberg figures, have contributed a cumulative of over $150 billion to the equity of private companies. The most significant two companies in the world ByteDance Inc., estimated at $75 billion, and Didi Chuxing Inc., valued at about $56 billion, are among its transactions. In some instances, the participation of SoftBank for several funding rounds helped drive valuations that culminated in the company’s paper earnings.
The debacle at WeWork raises questions regarding statistics like this. The value of the coworking company crested with SoftBank’s acquisition this year at $47 billion but plunged to $7.8 billion in a Son-engineered rescue. WeWork slashes workers and slows down activities.
“WeWork is not just an error, it is a sign of weakness in the entire model,” said Aswath Damodaran, a finance professor at the Stern School of Business at New York University who has published four books on business valuation. “What about all the other firms in your portfolio because you screwed up their value too badly?”SoftBank said that WeWork is an exception rather than a sign of wider issues, and learned from experience.
Son has become the most successful tech entrepreneur on the globe since it launched its $100 billion Vision Fund in 2016, pouring money into more than 80 companies. According to research firm CB Insights, it helped to create a bumper crop of unicorns, more than 300 companies valued at $1 billion or more.
What is not so well known is Son’s incentive to keep valuations on the rise. Son says he’s made a profit when SoftBank sells stock in a company and then reinvests at a higher cost. Under accounting standards, this is legitimate, but SoftBank does not receive any money. The only shift is that SoftBank has increased the value of its initial investment by expanding the startup’s valuation from, say, $1 billion to $2 billion.
At least some of the extra $1 billion can be viewed as gain in SoftBank’s income statements and dividend figures. “You manipulate valuations to get higher returns and look good to investors,” says Eric Schiffer, Patriarch Organization’s Chief Executive Officer, a private equity fund headquartered in Los Angeles. “These crowdfunding tools are simply unicorn videos,” SoftBank said its reporting follows all requirements and is aligned with widely accepted procedures.
As far as start-up valuations are concerned, it said it does not assess them on its own and partners with established companies such as Sequoia Capital and Temasek Holdings Pte. “Our valuations have been confirmed by over 120 sophisticated investors who have invested with and after us,” said Navneet Govil, Chief Financial Officer of SB Investment Advisers, the Vision Fund manager, in a release.
SoftBank said it has a robust internal assessment system, and only after taking into account future cash flows and public market correlations, as well as private market borrowing costs, it reports the benefit on any cost increase. SoftBank’s auditors reviewed these numbers at Deloitte & Touche, and the limited partners of the Vision Fund have their auditors, including Duff & Phelps workers and Ernst & Young who are reviewing the final estimates.
“Our assessment process is rigorous, and external auditors evaluate it every quarter,” Govil said. “We agree that we have a strong performance. Vision Fund 1 has already had seven IPOs in just 21⁄2 years, $4.7 billion in realized gains, $11.4 billion of net investment gains, and distributed $9.9 billion to our limited partners. “Today’s accounting rules may be inadequate for a period of exponential unicorn trading.
Under SoftBank’s International Financial Reporting Standards (IFRS), businesses have wide latitude when deciding how much portfolio companies they believe are worth— and hence how much income they show to shareholders. It’s uncertain whether any organisation is now using SoftBank to calculate paper revenue from tech startups. “I don’t think we’ve ever seen an effort to document the number of profits from unquoted equity investments,” New York tax expert Robert Willens said.
The bookkeeping of Son has helped him to assert his average internal rate of return well outstrip those of other shareholders. When SoftBank took a hit from WeWork in November, Son justified his commitment to investing. “Worldwide, there is 5,000 venture capital, and the median IRR is 13%,” he added.
“Our return is about twice as large as this.” Son’s confidence in his understanding led to the creation of the Vision Fund in 2017, which was more than ten times the scale of any venture capital fund at the moment. He tried to replicate the performance of his principal investment— a $20 million gamble on China’s Alibaba Group Holding Ltd. that has now transformed into a stock worth more than $120 billion.
Son started a blitzkrieg of transactions in 2017 with Abu Dhabi’s Mubadala and Saudi Arabia’s Crown Prince Mohammed bin Salman funding the Vision Fund. According to research firm Preqin, he has spent more than $35 billion in about 100 firms. Among the strongest was WeWork and Didi Chuxing’s multibillion-dollar fundings, the ride-hailing Chinese giant modelled after Uber Technologies Inc. A SoftBank-led consortium spent $9 billion in Uber in December of last year, including buying stock from existing shareholders.
During the fiscal year ending in March 2018, SoftBank started publishing financial results for the Vision Fund. Total operating income was 303 billion yen, or less than 3 billion dollars, including a separate Delta Trust. This grew to 1.26 trillion yen the following year, rendering it the most productive division of Son’s group, accounting for over half of the operating income of the parent company. SoftBank spun off the domestic telecom sector with Son’s power directed at entrepreneurs, which had made it famous and produced cash for his early investments.
“My head and my soul, virtually everything about me depends on the Vision Fund,” Son told shareholders in May. But most of the gains reported by SoftBank were on paper. Unrealized gains on portfolio valuations accounted for most of the estimated income for the Vision and Delta funds during the first fiscal year.
Unrealized investment losses contributed to 1 trillion yen in the most recent fiscal year, while perceived increases— like India’s e-commerce giant Flipkart’s sale to Walmart Inc. — reached fewer than 300 billion yen.
WeWork illustrates the risks associated with this strategy. SoftBank first took a $21 billion interest in August 2017. It then spent another $3 billion in a $45 billion appraisal in November 2018 and eventually committed to a $1.5 billion contract at $47 billion. He cited WeWork as an illustration of portfolio companies preparing to IPOs when Son reported results this May. SoftBank took a hit of 498 billion yen when the agreement failed.
SoftBank Vision Fund said by pushing it all the way up to $47 billion, it never took money from WeWork. It held the stock at only half the value on its accounts. It still had to take down that by around 75%, which resulted in the loss.
Venture capital and private equity firms are mostly anonymous, so they don’t have to disclose quarterly profits to public investors, so their limited partners usually focus on dividends as investment companies cash out through IPOs or acquisitions. For each of its portfolio companies in a third, SoftBank does not announce detailed price shifts, usually listing just a few winners or losers. Often investors who put money alongside SoftBank are partners such as Uber, Didi and Alibaba.
“If I’m an investor, I’d like to know how these numbers come up. You can’t believe any of the valuations otherwise, “said Damodaran of NYU. “The more they think to accountants, the less I can trust the numbers,” SoftBank claimed that the Vision Fund would classify increases in value as profit under IFRS standards because it invests in its primary business, and SoftBank Group must add the revenue into its accounts because the Vision Fund is a merged branch.
One person close to the organization said the corresponding income estimates were nearly meaningless, but given the current regulations, there is no better accounting process. “I’m not an accountant or an attorney, but it doesn’t make any sense to view this as profit,” said Ilya Strebulaev, a finance professor at Stanford University’s Graduate School of Business whose research suggests that the latest post-money valuations usually exceed startups by about 50 fold.
Adding to the uncertainty is that sometimes start-up shares are exchanged between the SoftBank Group and the Vision Fund, to separate investors. The cost at which the goods were transferred has an effect on both sides earnings.
Other contracts are under investigation following WeWork. In 2015, SoftBank invested in Didi Chuxing with the value of the Chinese company