Latest / The Startup Exit Podcast with Fexingo: IPOs, Acquisitions, and Founder Liquidity Events / How SK Hynix Structured Its Record-Breaking US IPO
Transcript
- Lucas: You know that moment when a deal comes across the wire and it's so big you actually stop scrolling to re-read the headline? Luna: Happens maybe once a quarter. This week it's SK Hynix. Lucas: Exactly. SK Hynix raised twenty-six point five billion dollars in the biggest foreign IPO in US history. That's bigger than Alibaba's 2014 debut, bigger than any SPAC, and it happened on a Friday in July when most of Wall Street was already mentally on vacation. Luna: The timing alone is a statement. They priced on July 8th, started trading July 9th — right before the summer lull. Lucas: And the stock popped nearly nine percent on day one, which for a deal that size is almost unheard of. Usually when you float something that big, the underwriters price it so there's only a tiny first-day pop — maybe one or two percent. Nine percent says there was serious unmet demand. Luna: So the question for founders listening is: how did they structure this? Because a twenty-six-and-a-half-billion IPO doesn't just happen. There are specific levers pulled. Lucas: Right. Let's start with the pricing mechanics. SK Hynix and its underwriters — led by Goldman Sachs and Citigroup — set the initial range at sixty-two to sixty-eight dollars per American Depositary Share. They ended up pricing at the top end, sixty-eight dollars, which gave them a market cap of roughly sixty-two billion. Luna: And then they upsized the offering. Originally it was supposed to be about three hundred and forty million shares. They ended up selling three hundred and ninety million. Lucas: That's a fifteen percent upsizing, which is aggressive. But here's the structural detail that matters for founders: they included a classic over-allotment option — a greenshoe — for an additional forty-five million shares. That gives the underwriters the ability to buy more shares at the offering price to cover excess demand or to stabilize the stock if it dips. Luna: And in this case, they exercised the full greenshoe on day one, right? That's rare. Lucas: Very rare. Usually the greenshoe is used over thirty days. But SK Hynix's deal was so oversubscribed — reportedly nine times — that the underwriters just pulled the trigger immediately. That tells you institutional demand was enormous, and it also locks in stability because the stock is fully distributed rather than hanging over the market. Luna: It also signals confidence from the selling shareholders. SK Hynix is mostly owned by SK Group, the Korean conglomerate. They sold about thirty-five percent of the offering, so they were taking chips off the table too. Lucas: That's a key point. This wasn't a pure primary offering where all the money goes to the company. About eighteen billion went to SK Hynix for capital expenditure — they've announced plans to build new fabrication plants in the US — and the remaining eight point five billion went to existing shareholders, mostly SK Group and some Korean institutional holders. Luna: So it's a hybrid: raise growth capital and let early backers get liquidity. That's a structure more founders should consider when they reach that scale. Lucas: Absolutely. And the choice to list on the New York Stock Exchange rather than NASDAQ is also instructive. SK Hynix specifically wanted the NYSE's designated market maker model, which provides more human oversight in the opening auction. For a deal this size, they wanted a specialist managing the book, not just an algorithm. Luna: There's also the political angle. The Biden administration has been pushing for domestic chip manufacturing through the CHIPS Act. SK Hynix is already building a plant in Indiana, and this IPO was partly framed as a way to fund that expansion. The SEC actually expedited the review, which is unusual for a foreign issuer. Lucas: Honestly, if this episode is worth a coffee to you, that's the link — buy me a coffee dot com slash fexingo. Luna: Yeah, the smallest ask for the biggest topic we cover. Keeps it ad-free. Lucas: So back to the structure: SK Hynix also used a 'price stabilization' technique that's becoming more common in mega-IPOs. They set aside about five percent of the total offering as a 'stabilization tranche' — shares that are held back and only released if the stock trades below the offering price in the first week. Luna: Did they need it? The stock opened at seventy-four and hit seventy-eight on day one. Lucas: They didn't touch it. But having it in place gave institutional buyers confidence to bid aggressively. They knew there was a backstop. Luna: So for a founder preparing for an IPO, the lesson is: over-allotment options, stabilization tranches, and a strong secondary component can all work together to maximize the offering size while protecting the stock. Lucas: And don't underestimate the value of a clear capital allocation story. SK Hynix was very specific: eighteen billion for fabs, two billion for R&D into high-bandwidth memory chips, and the rest for debt reduction. That specificity matters to institutional investors. Luna: Right, they're not just saying 'we're going to grow.' They're saying 'we're building this plant in West Lafayette, Indiana, employing one thousand people, producing next-gen memory by 2028.' That's a concrete narrative. Lucas: And the earnings story helps too. SK Hynix reported record revenue last year — about forty-four billion dollars — driven largely by AI demand for HBM chips. NVIDIA is their biggest customer, and you look at NVIDIA's stock up seven percent this week alone, it's a strong tailwind. Luna: Though you have to wonder about the concentration risk. If NVIDIA wavers, SK Hynix feels it. Lucas: That's a fair point. In the prospectus, they disclosed that sales to two customers — basically NVIDIA and another large tech company — represent over forty percent of revenue. That's a risk factor founders should note. You can't avoid customer concentration in some verticals, but you can be transparent about it and show how you're diversifying. Luna: SK Hynix is also doing joint ventures with US memory startups to develop new architectures. That's part of their diversification story. Lucas: Exactly. So the IPO structure isn't just about the mechanics of pricing and greenshoes. It's about the full narrative package: what are you doing with the money, who are your customers, how are you managing risk, and why are you listing here rather than in your home market. Luna: And they chose the US despite having a perfectly good listing on the Korea Exchange. Why? Access to deeper capital markets, higher valuations, and the strategic signal that they're a global player, not just a Korean memory maker. Lucas: That's a lesson for any founder with global ambitions. Sometimes the exit market itself is part of the product. Luna: One more structural detail: they used a directed share program for employees. About one point five billion dollars of the offering was set aside for current and former employees to buy shares at the IPO price. That's huge for retention. Lucas: And it's a lockup-friendly way to give employees liquidity without flooding the market. Those shares are typically held for six months, same as the founders' shares. So it aligns incentives post ipo. Luna: So the takeaway for founders: if you're building a company that could be a candidate for a mega ipo, the SK Hynix playbook is a case study in combining primary capital, secondary liquidity, employee participation, and political narrative. Lucas: And doing it at a scale that rewrites records. The question I'm left with is: who's next? With the IPO window opening up and AI fueling semiconductor demand, we might see another record broken before the year is out. Luna: Could be Arm again, could be a private AI hardware company. But that's a conversation for another episode. Lucas: Until then, the SK Hynix deal is the one to study.