The information in this Prospectus is not complete and may be changed. We may not sell these securities until the registration statement filed with the Securities and Exchange Commission is effective. This Prospectus is not an offer to sell these securities and is not soliciting an offer to buy these securities in any state where the offer or sale is permitted.
SK Hynix (KRX: 000660), one of Korea's industrial titans, is expected to list American Depositary Receipts (ADRs) in the United States in what may be the largest listing in history, eclipsing the ADR of Alibaba in 2014.
Read ArticleSK hynix leads the market for high-bandwidth memory (HBM) right now, the stacked DRAM that sits next to an AI accelerator and feeds it data fast enough to keep it working. Standard memory can't move data at those speeds. As AI models get bigger, that bandwidth becomes the bottleneck, which is why HBM has turned into a core part of every new accelerator. The lead isn't permanent, and that's part of the story.
Its HBM ships inside the accelerators doing the heavy lifting today, including Nvidia's H100, H200, and Blackwell chips and AMD's MI300 series. SK hynix works with foundry partners like TSMC (NYSE: TSM) to turn raw inputs into finished memory. In effect, every accelerator that ships with HBM is HBM sold, so SK hynix results track the AI hardware cycle closely.
Two things. AI spending keeps pulling HBM demand higher, and SK hynix is preparing a US listing of American Depositary Receipts, seeking to raise up to $29 billion by issuing 17.79 million shares. That would give US investors brokerage access to a name they currently trade on the Korea Exchange (KRX: 000660). The other side is real too: Samsung and Micron (NASDAQ: MU) are spending to close the HBM gap as the HBM4 standard arrives, and memory has always been a cyclical market. Both the momentum and the risks are live.
SK hynix HBM goes into the AI accelerators companies are buying right now, including Nvidia’s H100, H200, and Blackwell chips and AMD’s MI300 series. When those accelerators sell, the HBM inside them sells too. That’s why SK hynix results tend to move with the AI hardware cycle.
Making HBM runs on a chain of specialized equipment and materials. Lithography, deposition, and etch tools from vendors like ASML (NASDAQ: ASML), Applied Materials (NASDAQ: AMAT), and Lam Research (NASDAQ: LRCX) sit upstream of any advanced memory maker. When memory demand runs hot, that demand flows back up the chain to the companies that build the tools.
SK hynix turns specialized inputs into HBM, working with foundry and packaging partners like TSMC (NYSE: TSM). That HBM then gets built into the accelerators Nvidia and AMD sell. Sitting at that pivot point lets SK hynix capture a slice of demand from across the AI buildout.
SK hynix leads HBM today, but it doesn’t own the market. Samsung (KRX: 005930) and Micron (NASDAQ: MU) are both spending to close the gap, and the next standard, HBM4, is coming. In past memory transitions, whoever matched performance with volume first tended to take the lead, so the picture can shift.
SK hynix sits upstream of Nvidia. For traders who want leveraged single-stock exposure downstream, Direxion offers the Direxion Daily NVDA Bull 2X ETF (NVDU) and the Direxion Daily NVDA Bear 1X ETF (NVDD). For broader semiconductor exposure, there’s the Direxion Daily Semiconductor Bull 3X ETF (SOXL) and Bear 3X ETF (SOXS). All are leveraged or inverse funds that seek daily objectives and are built for short holding periods.
To hold its exposure at the target multiple, the fund adjusts positions daily as the underlying’s price moves. A 2X fund that gains has to add exposure to stay at 2X; one that loses has to cut it. This daily rebalancing is the mechanism behind the reset, and it’s standard for leveraged and inverse ETFs.
At the end of each trading day, the fund resets its exposure to line up with the next day’s objective. This is what keeps it targeting its multiple day after day. It also means each day’s result compounds on the last, so returns over several days can differ, sometimes by a lot, from the stated multiple times the underlying’s move over that same stretch.