SK Hynix reported record-breaking Q2 2026 results on July 29, 2026. Revenue hit 79.3187 trillion won (roughly $54 billion USD) and operating profit reached 60.5426 trillion won, fueled by surging demand for AI memory chips. Net profit for the quarter came in at 93.9226 trillion won (SK Hynix).

The year-on-year growth rates are extraordinary. Operating profit jumped 557% from the same period a year earlier, equivalent to about $41.62 billion USD (Reuters). Revenue more than tripled (CNBC). The operating margin — the share of revenue left after covering the costs of production — landed at roughly 76% (Investing.com).

Yet the market's reaction was sharply negative. SK Hynix shares fell about 10% after the earnings release. The selloff happened because the record profit figure still missed Wall Street forecasts, despite the massive year-on-year jump (Reuters).

The balance sheet stands on its own. SK Hynix's net cash position — cash on hand minus debt — reached 88 trillion won at the end of June 2026, giving the company substantial capacity to keep investing in HBM (High Bandwidth Memory) capacity and advanced chip manufacturing (Reuters).

The broader context here is about what those numbers mean and why expectations matter more than headline growth. A 76% operating margin is remarkable for a semiconductor company. Memory chip makers have historically posted strong margins when supply is tight, but a quarterly margin at this level, if it holds, would put SK Hynix in territory usually occupied by software companies — businesses that don't require multi-billion-dollar factories. The fact that shares dropped on a forecast miss, not a demand contraction, tells you that analysts had baked in even more aggressive assumptions about HBM pricing or product mix than the company actually delivered.

A 557% year-on-year profit increase that still disappoints the market says something about how expectations were set. When the comparison period is a trough — a low point in the business cycle — triple-digit growth rates lose analytical meaning. The more useful comparison is sequential: how the company's actual selling prices and shipment volumes tracked against its own guidance. The revenue figure of 79.3 trillion won against an operating profit of 60.5 trillion won implies a cost structure where the economics of HBM3E and next-generation HBM products are generating profit margins that would have seemed implausible two years ago.

There's also the gap between net profit and operating profit. Net profit of 93.9 trillion won exceeding operating profit of 60.5 trillion won points to substantial non-operating gains — likely from the company's investments in other firms and foreign exchange effects, though the reported facts don't break out the composition. For a company sitting on 88 trillion won in net cash, interest income alone could contribute meaningfully to that gap in a high-rate environment.

The 10% share price drop on a record quarter deserves scrutiny. It fits a pattern where semiconductor stocks trade on what investors expect will happen next, not on what just happened. If the forecast miss reflected even slightly softer HBM pricing assumptions or conservative volume guidance for the second half of the year, the market may be pricing in a peak-margin story rather than a fundamental deterioration. That distinction matters: a cyclical peak followed by gradual normalization is very different from a structural demand collapse, and the reported facts don't provide enough detail to tell which dynamic is at play.

The 88 trillion won net cash position is a strategic factor in its own right. In a memory market where competitors face varying degrees of constraint on how much they can invest, SK Hynix's balance sheet lets it keep pouring money into HBM capacity expansion and yield improvement without the financing costs that typically limit supply growth during tight markets. That ability to invest through the cycle could shorten the duration of the current boom if supply growth outpaces demand, though that is forward-looking speculation beyond the reported facts.

For those tracking the AI memory supply chain, the data points to watch going forward are selling price trends for HBM3E and HBM4, shipment growth versus capacity additions, and any commentary on how much inventory big cloud customers are holding. The forecast miss, whatever its exact composition, suggests that consensus estimates may need recalibrating — the growth assumptions had become extrapolative, projecting recent gains forward without enough regard for what could shift.