The decline in AI trading combined with heated product competition. SK Hynix (SKHY.US) leveraged ETF landed in US stocks and was liquidated in March

The Zhitong Finance App learned that the two US leveraged ETFs linked to SK Hynix (SKHY.US) — the Tradr 2X Long SK Hynix Daily ETF and the Tradr 2X Short SK Hynix Daily ETF — will be liquidated around November 12, only about three months until the launch on July 28. In a statement, issuer Tradr attributed the delisting to “extremely intense competition among related ETFs” and stressed that it did not see any problems with SK Hynix as an underlying asset. However, the short shelf life of the product and the drama of shrinking assets itself has formed a market footnote that is more honest than any statement: when the AI narrative shifts from “buying everything” to “settlement time,” those leveraging tools created to amplify intraday fluctuations are the first to lose their place of existence.

The “rapid extinction” of leveraged instruments: behind the liquidation is crowding, but also a decline

The immediate trigger for the liquidation was certainly “crowding” at the commercial level. After SK Hynix ADR was launched in July, publishers such as ProShares, Leverage Shares, and Direxion quickly poured into the same circuit, and Tradr's first-mover advantage was diluted within a few weeks. The combined net assets of the two products are only about 1.3 million US dollars, down more than 60% from the peak. Liquidation is almost an inevitable choice under economic rationality.

But what is really worth examining is what is happening in the macro undertones attached to this “product surplus.” The Philadelphia Semiconductor Index has retreated more than 26% from its June high and has entered a technical bear market. AI stocks and non-AI stocks have formed a rare “seesaw” split within US stocks. Leveraged ETFs were manufactured in batches during periods of market acceleration, and were the first to be cleared when volatility increased and capital was withdrawn — their life cycle itself is a yardstick for AI trading congestion.

Valuation rift before the financial report test: profits are setting records, and stock prices are “voting with your feet”

There is a sharp divergence between SK Hynix's fundamentals and stock price, which is rare in recent years. The company's revenue for the second quarter of 2026 reached a record high of 79.3 trillion won, operating profit reached 60.5 trillion won, and the EBIT margin reached 76.3%, even surpassing TSMC. HBM4 began mass production and delivery to key customers in the second quarter, and the yield is close to the mature HBM3E level. However, the stock price of SK Hynix, which is listed in Seoul, has fallen 52% from its high in late June, and Samsung Electronics also fell 36% during the same period.

On October 28, the company will release results for the third quarter. The market generally expects revenue of about 98 trillion won and operating profit of about 77 trillion won. At the same time, new shareholder return measures will also be disclosed. Market rumors may involve a special dividend of about 12 billion US dollars. This financial report will be a key point in testing a core issue: when the financing costs of AI infrastructure begin to be seriously examined, can the “supercycle” profits of memory chips support the valuation multiples that have already been reduced to single digits. The answer to the former is still likely to be positive, but the revaluation of the latter — how much the market is willing to pay for these profits — is clearly far from over.