Why does it even matter if SoftBank’s an investment holding company?

{"text":[[{"start":7.84,"text":"SoftBank Group Corp is readying itself to launch a ton of US dollar bonds. Bloomberg reckons it could amount to as much as $20bn. That’s chunky, even in the hyperscaler era, and will cause a big splash in the high-yield bond market."}],[{"start":23.38,"text":"The PowerPoint legends who manage the SoftBank Vision Fund already have €8.3bn of euro-denominated bonds and almost ¥6tn ($38.2bn) of yen bonds outstanding. Given Softbank’s junk rating, this makes it the largest single index constituent in the ICE BofA Euro high-yield index. We can’t find an equivalent yen index, but if there was one, SoftBank bonds look to us like they’d populate over four-fifths of it."}],[{"start":51.36,"text":"But where would SoftBank place in the OG market for junk bonds? To take the top spot in the US high-yield index its total bond outstandings would need to eclipse Charter Communications’ $25.9bn. Alphaville’s Bloomberg terminal informs us that SoftBank has already issued $12.6bn of US dollar bonds, so maybe SoftBank can now grab the top spot in the US index too?"}],[{"start":75.12,"text":"Running our finger down the largest dollar issuers we’d expected SoftBank to come in at number six, just below Nissan Motor Co. But it doesn’t. SoftBank’s weighting is zero/nada/zilch. Huh?"}],[{"start":null,"text":"

"}],[{"start":86.96,"text":"Almost all of the US dollar bonds SoftBank has ever issued* have been so-called Reg S securities, meaning that they are sold only outside the US and only to non-US persons. Meanwhile, US domestic high-yield bond indices consist solely of . . . US domestic bonds."}],[{"start":104.52,"text":"Alphaville understands that this lack of issuance into the domestic market is partly because demand for dollar SoftBank paper has come overwhelmingly from yield-hungry Asian private banks. So why test pricing in the US domestic market?"}],[{"start":117.68,"text":"But we understand it’s also mostly because SoftBank Group Corp is a foreign investment holding company, or at least it has been since maybe 2017.** And as such, issuing domestic bonds is not entirely straightforward because of safeguards against shenanigans implemented following the Great Depression."}],[{"start":133.84,"text":"In the 1920s, a bunch of fund sponsors took out loans against the value of the assets they were managing to build their businesses, personally enrich themselves, embezzle, that sort of thing. When the market went pop in 1929, loads of people were wiped out."}],[{"start":150.56,"text":"In the fallout, the Securities and Exchange Commission was set up in 1934. And one of its first tasks was to run a monster, Congress-mandated study of the investment industry. The recommendations that poured out of this ultimately culminated in the Investment Company Act of 1940, along with much of the legal architecture that governs today’s fund management sector. Importantly, as then-SEC Chair Mary Jo White set out in her speech appraising the history of the Act back in 2015, “the Act limits funds’ issuance of debt and other senior securities”."}],[{"start":183.84,"text":"Tl;dr: investment holding companies have it tough if they want to launch bonds into the US market. But having it tough doesn’t mean they’re completely prohibited from doing so."}],[{"start":193.16,"text":"What it does mean, we think, is that any investment company hoping to sell bonds will need to rely on Section 3(c)(7) of the Act. This requires not only that buyers are “qualified institutional buyers” (ie, own at least $100mn of stuff), just like regular domestically registered 144A bonds, but also requires every holder to satisfy a “qualified purchaser” requirement."}],[{"start":216.6,"text":"As far as Alphaville can work out, there's a lower asset threshold to qualify as a qualified purchasers versus being a QIB, so we cannot understand quite why this is a biggie. But then we’re not lawyers, and smarter folk closer to the fixed-income coalface mutter about additional transfer restrictions. Basically, it’s all sufficiently complicated that it’s easiest to steer clear of in normal times."}],[{"start":239.28,"text":"But SoftBank needs money. Lots of it."}],[{"start":240.88,"text":"Earlier this year it extended its massive punt on OpenAI, and entered into a bridge loan for $40bn earlier to do so. The bridge loan comes due in March 2027. At the time, OpenAI looked set to IPO in 2026, meaning it might reasonably have expected to be able to refinance the bridge into a jumbo margin loan at relatively cheap rates using new public stock. But with the OpenAI IPO now looking distinctly like a post-2026 event, it’s time for plan B."}],[{"start":271.605,"text":"And it looks like plan B includes this entry to the US domestic bond market, as well as maybe a ¥1tn Japanese retail bond ($6.25bn) and a new $10bn margin loan against OpenAI private stock."}],[{"start":286.44,"text":"When we asked SoftBank Global Corp for details, a spokesperson told Alphaville that they were “considering various refinancing options, but no decision has been made on the USD bond issuance, including the amount and timing”."}],[{"start":299.42,"text":"We don’t know where the new SoftBank dollar debt will price. Heck, we don’t even know what tenor, size or regulatory status any new issuance will have. Maybe the company will structure it all ‘innovatively’. But we can have a guess, based on where existing Reg S SoftBank dollar bonds trade."}],[{"start":null,"text":""}],[{"start":315.48,"text":"With rinky-dink qualified purchaser requirements and a healthy new issuance premium, existing 10-year bonds trading around 400 bps higher than US Treasuries, and the 10-year US Treasury homing in on 5 per cent, it doesn’t seem wild to speculate that any new 10-year SoftBank dollar benchmark could come with a 9 per cent coupon."}],[{"start":335.64,"text":"Given the size of the US high-yield market, it looks unlikely that the new issuance could take it all the way from index-zero to index-hero. But it appears at least that SoftBank will — at last — join the top table of index constituents."}],[{"start":349.16,"text":"* SoftBank issued a couple of 144A securities, once in 2011 through Ymobile, and then again in 2013 — bonds that look like they are private placements from a legal perspective, but can basically be bought by any investor satisfying the qualified institutional buyer test."}],[{"start":365.44,"text":"**Back in April 2013, you could still maybe squint and at least pretend that SoftBank Global Corp wasn’t actually an investment holding company. One TMT credit investor we spoke to told us it was absolutely not an investment company back in the day, and was instead mostly defined by its 100 per cent holding of the (confusingly named) SoftBank Corp, a Japanese telco (which directly guaranteed the 2013 bonds). So issuing 144A securities wasn’t such a big deal. However, with its disparate financial stakes growing in value, cosplaying as anything but an investment holding company became increasingly untenable. And after IPOing a large chunk of SoftBank Corp in December 2018, no one could really say with a straight face that SoftBank Global Corp wasn’t clearly an investment holding company."}],[{"start":365.94,"text":"Further reading:"}],[{"start":417.78,"text":"— The multiplying risks of financing data centres (MainFT)"}],[{"start":421.54,"text":"— Untethered Goose Game (FTAV)"}],[{"start":424.46,"text":"— The emerging reality of the OpenAI-SoftBank grand plan for data centres (MainFT)"}],[{"start":430.08,"text":"— All you never wanted to know about corporate bond market issuance (FTAV)"}],[{"start":438.48,"text":""}]],"url":"https://audio.ftcn.net.cn/album/a_1788436385_5826.mp3"}