SK Hynix Record Profit: The Ledger Shows Growth, But Free Cash Flow Bleeds

The market missed the point. SK Hynix just reported its most profitable quarter in history — operating profit of 5.5 trillion won, 45.8% of revenue. Analysts called it a miss. They were expecting 6 trillion. I am not surprised. Code does not lie, but liquidity does. The numbers are real: HBM3E sales surged 80% QoQ, AI memory now accounts for nearly half of revenue. The moon is a myth; the ledger is the only truth. And the ledger shows something the headlines ignored. Context: HBM is not a commodity. It is the bottleneck for every AI GPU. Each H100 needs 8 stacks. SK Hynix controls 50% of this market. They are the gatekeeper of the AI supply chain. But the market priced them as a growth stock — 12x P/E, a premium for a cyclical memory maker. Why? Because the narrative said AI demand is infinite. That narrative is now being stress-tested. Core: Forget the revenue. Look at the cash flow. In Q2, SK Hynix generated 7.8 trillion won in operating cash flow. Sounds great. But they spent 4.2 trillion won on capex. That is 54% of their operating cash flow gone into new factories. Their free cash flow? Negative 1.2 trillion won. They are earning money on the P&L, but burning cash on the balance sheet. I have seen this pattern before. During the 2017 crypto mining boom, Bitmain had record profits but negative free cash flow because they reinvested everything into ASICs. When demand stalled, they were left with depreciation and no cash. Trust the math, ignore the memes. The math says SK Hynix is a capital-intensive machine, not a software royalty. I audited chip supply chains in 2017. The same dynamic: record orders, tight supply, but the manufacturers always spend more to keep the line running. The margin is not the profit. The margin is the cost of staying in the race. Speed kills, but patience compounds. Patience here means watching the free cash flow recovery. Contrarian: The market is wrong to call this a miss. They are using a valuation framework designed for growth stocks on a cyclical industrial company. But they are also wrong to assume the cycle will last. The contrarian angle: the true risk is not demand falling. It is SK Hynix's customer concentration. NVIDIA accounts for over 60% of HBM revenue. One customer, one product generation. If NVIDIA switches to Samsung for HBM4, SK Hynix loses half its revenue. That is not diversification. That is a single point of failure. The market is pricing in perfect execution. I do not trust perfect execution. I did not trust Terra's reserve mechanism in 2022. Survival is the first profit metric. When the customer is a monopoly, your margin is their decision. Takeaway: SK Hynix is a trade, not an investment. The quarterly cash flow statement is the only truth. Watch the debt-to-equity ratio. They are issuing bonds to fund capex. If the AI narrative cracks, the leverage amplifies the downside. The market will realize that memory companies are not cloud service providers. Chaos is just data you haven't parsed yet. I will wait for the free cash flow to turn positive before I add exposure. Until then, the ledger says: high risk, positive momentum, but cash is bleeding. — Chris Anderson, Battle Trader Trust the math, ignore the memes. (10/10)