The Second-Largest Listing Will Beat the Largest

The forecast in one line

The second-largest US share sale will outperform the largest one next week.

SK hynix is expected to begin trading on Nasdaq on 10 July. Its American depositary receipts should finish their first session 10 to 15% above the offering price. SpaceX will enter the Nasdaq-100 three days earlier, but its shares should fall from Thursday’s US$162 close toward US$145 to US$150 by Friday.

That produces an expected relative spread of roughly 20 to 25 percentage points in only four trading days.

Security	2 July reference	Base case by 10 July	Expected return	Positioning
SK hynix ADR	Offering price set next week	10–15% above offer	+10% to +15%	Participate near the offer; do not chase above +15%
SpaceX	US$162.00	US$145–150	−7% to −11%	Avoid the inclusion-day bid; bearish below US$160
Relative trade	—	SK hynix over SpaceX	+20–25pp	Highest-conviction expression

Both companies sit inside the artificial-intelligence capital cycle. Both will arrive at a major Nasdaq milestone. Both can attract billions of dollars in demand.

The distinction is the source of that demand. SK hynix will offer exposure to scarce memory capacity and rapidly growing earnings. SpaceX will receive a large but finite mechanical purchase from funds that have no discretion. The first bid can persist; the second expires when the index rebalance is complete.

The trade is not old AI against new AI. It is earnings-backed scarcity against demand that has already been scheduled.

Sneak peek: the securities that will define the week

SK hynix: buy the offer, not the opening excitement

SK hynix will give US investors direct access to one of the most consequential bottlenecks in the AI buildout. The company supplies high-bandwidth memory, or HBM, that sits beside advanced accelerators and allows them to move enormous datasets quickly enough to train and serve large models.

The addressable demand is expanding in two directions. Nvidia’s Rubin platform is expected to use 288GB of HBM4 when it rolls out commercially in the second half of 2026. Rubin Ultra could use 1TB in 2027. Custom accelerators from Google and Amazon are also increasing the memory content attached to each processor.

That matters because the supply response is slow. HBM requires advanced packaging, high manufacturing yields and capacity that could otherwise produce conventional DRAM. Customers are responding with longer agreements, prepayments and price bands designed to secure supply rather than merely obtain the lowest spot price.

SK hynix therefore arrives in New York with three advantages:

1. The product is scarce. Incremental AI accelerators require an increasing amount of high-value memory.
2. The earnings are visible. Long-term agreements reduce the amount of volume exposed to short-term memory pricing.
3. The listing removes friction. US-dollar trading, US settlement and normal brokerage access expand the potential investor base.

The valuation still appears modest relative to that position. Mirae Asset estimated on 9 June that SK hynix traded at approximately 5.8 times 2026 earnings, compared with a 12.4 times average for memory peers. Its later ADR analysis used a 2026 forward multiple of approximately 6.7 times, still at a discount to Micron and Kioxia.

Those estimates are unusually optimistic and should not be treated as settled fact. Memory remains cyclical, current margins are unlikely to persist indefinitely, and the Korean-listed shares have already produced an exceptional return. But the discount leaves room for a US listing premium before the valuation begins to resemble the more expensive parts of the AI complex.

Our first-day framework is deliberately tactical:

SK hynix debut outcome	Probability	Definition
Scarcity premium	60%	Closes 10–15% above the offering price
Orderly price discovery	25%	Closes between the offer and +10%
Oversupply or risk-off	15%	Closes below the offer

The preferred execution is an allocation near the offering price or a limit order no more than 10% above it. A first print already 15% higher would consume most of the expected one-week return. The thesis is strongest before enthusiasm becomes the price.

SpaceX: index inclusion will become exit liquidity

SpaceX will join the Nasdaq-100 before trading begins on Tuesday, only weeks after completing the largest initial public offering in US history. Nasdaq says more than 200 products with over US$800 billion in assets track the index. The inclusion will therefore require meaningful purchases by passive funds.

That fact is not a secret. It has been known since 26 June.

Replicating funds must be positioned for the new index composition. Arbitrageurs can buy in advance and sell shares into the closing rebalance. Existing holders know the date on which price-insensitive buyers must appear. Every participant is looking at the same scheduled demand.

This is why inclusion is not automatically bullish after the effective date. It transfers shares from investors who chose to own them to funds that are required to own them. Once that transfer is complete, the marginal buyer must again make an active decision at a valuation above US$2 trillion.

SpaceX has genuine strategic assets. Starlink supplies recurring revenue, reusable rockets provide an extraordinary launch-cost advantage, and the company’s infrastructure could support communications, defence and AI workloads. The bearish call does not require any of those businesses to deteriorate next week.

It requires only three narrower conditions:

• expected passive demand has already encouraged pre-positioning;
• the free float remains small enough to amplify both buying and selling;
• no new fundamental disclosure will reset the valuation during the forecast window.

SpaceX closed at US$162 on 2 July, 20% above its US$135 offering price but well below its US$225.64 intraday high from 16 June. That path suggests price discovery is not complete. A rally into Monday’s close should be treated as the final opportunity for index arbitrageurs to distribute shares, not evidence that the inclusion creates a permanent floor.

The forecast is a close between US$145 and US$150 by Friday. It will be invalidated by two consecutive closes above US$168 after the inclusion, accompanied by volume that shows discretionary buyers—not merely the rebalance—are absorbing supply.

The broad market will rise, but most risk will remain hidden

The paired trade sits inside a market that is rotating faster than its headline index suggests.

On 2 July, more than two-thirds of S&P 500 constituents advanced. Yet the index was virtually unchanged because semiconductor and other AI-linked shares fell. The Dow gained 1.1% to a record while the Nasdaq Composite dropped 0.8%.

That session demonstrated the reverse of the concentration investors have become accustomed to seeing: most stocks could rise while the cap-weighted index failed to participate. Next week should flip that relationship again. AI infrastructure shares should recover enough to lift the S&P 500 and Nasdaq even if the Dow and smaller companies lose momentum.

Index	Base-case weekly move	Interpretation
Nasdaq Composite	+1.0% to +2.0%	AI and semiconductor leadership returns
S&P 500	+0.5% to +1.5%	Cap-weighted support, narrow participation
Dow Jones	−1.0% to +0.3%	Oil and rate sensitivity offset record-level momentum
Russell 2000	−1.0% to +0.3%	Higher financing sensitivity prevents confirmation

The counterintuitive result should be a positive S&P 500 week that does not feel broadly risk-on. Nasdaq outperformance will coexist with weakness in its newest large constituent. The index can rise because established AI infrastructure companies recover even while SpaceX falls.

A midweek oil shock will interrupt, not end, the technology bid

The largest threat to this forecast is not the listing calendar. It is the interaction between the conflict in the Middle East, oil and interest rates.

The temporary US–Iran framework remains fragile. Any sign that negotiations are failing could push crude higher, lift near-term inflation expectations and pressure equities. The Dow, transports, consumer cyclicals and small caps would be most exposed. Semiconductor stocks could initially sell off as investors reduce risk indiscriminately.

Our base case nevertheless assumes that a midweek oil shock will be partially reversed before Friday. The market has repeatedly separated temporary geopolitical escalation from the longer AI capital-spending cycle. A crude move that remains below the earlier wartime extremes would create volatility without necessarily changing 2026 earnings expectations for memory and computing suppliers.

This produces an expected sequence rather than a smooth weekly gain:

1. AI shares rebound as the market reopens after the holiday.
2. SpaceX weakens when Nasdaq-100 inclusion converts anticipated demand into completed demand.
3. Oil and rates create a midweek risk-off interruption.
4. Semiconductor and memory shares recover into the SK hynix debut.

The sequence is less important than the closing relationships. Nasdaq should outperform the S&P 500; both should outperform the Dow and Russell 2000; SK hynix should outperform SpaceX by at least 15 percentage points.

The Federal Reserve will reinforce the split

Minutes from the Federal Reserve’s June meeting are scheduled for release on Wednesday. The June statement already described economic activity as solid, inflation as elevated and energy as an important supply shock. It also replaced the previous easing bias with a blunt commitment to deliver price stability.

The minutes are therefore more likely to reveal a discussion of holding rates high—or potentially raising them again—than an urgent desire to ease. A hawkish reading should keep the 10-year Treasury yield inside a wide 4.45% to 4.65% range and make the market less tolerant of securities whose valuation depends primarily on distant cash flows.

That does not make all technology equally vulnerable.

SK hynix is being valued on near-term memory prices, contracted volume and current earnings expansion. SpaceX is being valued partly on infrastructure and AI opportunities that may take years to mature. Higher yields increase the relative value of the first cash-flow profile and reduce the present value of the second.

The Fed catalyst consequently supports the pair trade even if it produces little net movement in the major indexes.

Why the two listings are not comparable

Calling both transactions AI listings obscures the mechanism that matters.

Dimension	SK hynix	SpaceX
Immediate catalyst	New US access and price discovery	Mandatory index ownership
Primary demand	Active institutional and retail allocation	Passive replication plus arbitrage
Earnings anchor	Current memory sales and pricing	Long-duration growth across several businesses
Supply condition	HBM capacity constrained	Tradable float constrained
What ends the trade	Excessive debut premium	Completion of forced buying
Main risk	Memory-cycle peak and ADR dilution	Valuation and future float expansion

The supply constraint also differs. SK hynix benefits because customers cannot easily obtain enough HBM. SpaceX benefits because investors cannot easily obtain enough freely tradable shares. The former can increase operating profit. The latter can temporarily increase the share price but eventually resolves when additional stock becomes eligible to trade.

This is the central thesis: scarcity in the product market is more durable than scarcity in the share register.

What would prove the forecast wrong

A prophetic forecast is only useful if it can fail clearly.

The base case will be confirmed if all three conditions occur by Friday’s close:

1. SK hynix finishes its first US session at least 10% above the offering price.
2. SpaceX closes between US$145 and US$150, or at least 7% below its 2 July close.
3. The Nasdaq outperforms the Dow while the S&P 500 finishes the week positive.

It will be partially confirmed if the relative call succeeds but both securities move in the same direction. It will be rejected if SpaceX closes above US$168 after absorbing the index rebalance and SK hynix finishes below its offer.

Three developments could cause that rejection:

• an unexpectedly dovish set of Fed minutes pushes yields sharply lower;
• a durable geopolitical agreement reduces oil and risk premiums simultaneously;
• SK hynix prices the ADR at a premium large enough to eliminate its valuation advantage before trading begins.

The last risk is particularly important. A successful company can still produce a poor listing if bankers capture all expected upside in the offer price. The forecast is conditional on an offer that preserves a visible discount to US memory peers.

The signal for 6 to 10 July

Next week will be described as another referendum on artificial intelligence. That description will be incomplete.

The market will instead test two forms of scarcity. SK hynix controls difficult-to-replace memory capacity whose price is visible in current contracts and earnings. SpaceX has a scarce public float facing a known burst of index demand. Both can create spectacular prices, but only one directly increases the cash generated by the underlying business.

The expected result is precise: SK hynix should close its debut 10 to 15% above the offer, SpaceX should fall toward US$145 to US$150, and the Nasdaq should still finish ahead of the broader market.

The second-largest listing will beat the largest because investors will distinguish between being forced to buy a security and wanting to own its earnings.

Market data are available through the 2 July 2026 close. All prices, returns and scenario probabilities for 6 to 10 July are Q/Analyst estimates prepared on 5 July 2026, not reported outcomes or investment recommendations.