The Bank of Korea raised its policy rate to 3.00% on 27 August, up from 2.75%, having already gone from 2.50% to 2.75% in July. Back-to-back increases are unusual for this central bank. In the same session it lifted its 2026 growth forecast to 3.3% from 2.6%, the highest it has projected in five years, and Yonhap's account of the decision puts the semiconductor cycle at the centre of the reasoning: exports and facility investment running hot enough that the bank chose to move ahead of inflation rather than behind it.
For a US reader following Samsung Electronics Co., Ltd. (KRX:005930), the thing to register is the direction of causation. This is not a macro event happening to Samsung. Samsung's memory boom is a large part of why it happened.
And it has a consequence that shows up in the accounts.
The Bank of Korea's rate decisions are made by a seven-member Monetary Policy Board that meets eight times a year. Its mandate is price stability. What the 27 August statement describes is an economy where the impulse to tighten is coming from a single export sector, and where the bank is also watching house prices and the exchange rate. The governor characterised the consecutive moves as acting early with a small tool rather than late with a large one — a Korean proverb about fixing things before they grow — and as a deliberate signal to markets.
Korea has been here before. When one industry is large enough to set the national growth rate, monetary policy calibrated for the whole economy ends up calibrated mostly for that industry. That is fine while the cycle is up. It becomes uncomfortable on the way down, because the tightening is still in the system when the sector rolls over.
None of that is a forecast. It's a structural observation, and it's the reason a memory investor should care what the Monetary Policy Board does.
Samsung prices memory in dollars. It reports under K-IFRS in won. Every quarterly figure a foreign investor reads about this company has already been through a currency conversion.
For most of the first half of 2026 that conversion was helping. The won was weak, and the half-year cash flow statement carries the fingerprint: a positive ₩3.9tn effect from exchange rate changes on cash and cash equivalents, against a negative ₩1.6tn in the same six months of 2025. I want to be precise about what that line is and isn't. It is the revaluation of foreign-currency cash balances, not an earnings item, and at ₩3.9tn against ₩146.7tn of half-year operating income it's a rounding error in profit terms. It's useful as a direction indicator, not as a magnitude.
The direction has now changed. On 24 August the won traded into the ₩1,370s against the dollar intraday, its strongest level in thirteen months, and it closed at ₩1,386.1 on 25 August after eight consecutive sessions of appreciation. Reporting attributes the move to exporters converting dollar receipts back into won, which in the current Korean economy means chip receipts, plus foreign investors reducing dollar positions.
So the sequence is: memory prices rise, Korean export dollars flood in, the won strengthens, and the same memory revenue converts into fewer won. The company's own success is the mechanism that trims its reported result.
I can't size this, and I'd rather say so than pretend. Samsung does not publish a headline currency sensitivity in the half-year summary statements, and the hedging position and currency mix of its receivables aren't visible at the level I can read. What I can say is that a roughly 2% move in the won over eight sessions applies to a revenue base of ₩305.4tn per half. Even a fraction of that passing through is a larger number than most of the things analysts argue about.
The interest rate channel and the currency channel don't point the same way here, which is the part that makes this genuinely uncertain rather than simply negative.
Samsung paid ₩240.7bn of interest in the first half and received ₩2.06tn. That's more than eight times as much coming in as going out. Short-term borrowings fell ₩4.0tn over the period, and long-term borrowing activity was trivial in both directions. This is not a company that gets hurt by a higher policy rate through its funding cost, because it barely has one.
It is a company sitting on ₩92.9tn of cash and, as of June, a very large and recently expanded pile of financial instruments. Higher Korean short rates raise the return on all of that.
Which effect is bigger? I don't know, and the disclosure doesn't let me work it out, because the currency composition of those balances isn't broken out. A won-denominated deposit book benefits from a 3.00% policy rate. A dollar-denominated one doesn't. My guess is the currency channel dominates, simply because it applies to the revenue line rather than to a treasury balance, but that is a guess and I'm flagging it as one.
The first objection is that ₩1,386 is not a strong won by any historical standard, and thirteen months is a short window. That's true. If the recent move is a retracement inside a multi-year weak-won regime, the effect on Samsung is noise and this piece is about noise.
The second is that the whole framing may be backwards. A memory upcycle of this size swamps a few percent of currency in either direction. Memory average selling prices have moved by amounts that make a 2% won move irrelevant, and if that continues, nobody will ever mention the exchange rate.
The third is more specific: Samsung hedges. Large exporters run natural hedges through foreign-currency costs and financial hedges on top. The net exposure could be far smaller than the gross revenue figure suggests. I have no visibility into the hedge book and neither does anyone reading the half-year summary.
There's also a reasonable view that the Bank of Korea has now done what it intends to do. The Yonhap coverage notes the bank signalling a shift toward pacing after two moves. If the tightening is finished, the currency impulse from the rate side is finished too.
The exchange rate effect line in the Q3 cash flow statement. In the first half it was positive ₩3.9tn. If it flips negative, the direction argued above is confirmed in the company's own filing rather than inferred from a currency chart.
The Monetary Policy Board's October and November meetings. A third hike would mean the bank thinks the chip boom is still generating more inflation than it can absorb, and it would say something about how the central bank reads the durability of this cycle — which is a view worth having next to the company's own.
And the level itself. ₩1,350 would be a materially different world from ₩1,400, and kstock's own convention of converting at roughly ₩1,400 to the dollar is already out of date by about 1%. That convention appears on this page and in these articles, and it's an approximation, not a rate. Treat every dollar figure here accordingly.