Samsung's unionized semiconductor workers wrapped up their vote on May 27. After 18 days of strike threat, government-mediated negotiations, and an internal power struggle that nearly derailed the process, the deal is almost certain to pass. Turnout topped 86% as of May 25, and the voting trajectory left little room for a rejection.
But the stock story was already written before the ballots closed. The question isn't whether the union deal is expensive. It's whether Samsung's earnings trajectory - the real driver of this stock's surge from $39 to $140 this year - still supports a stock that trades at roughly 6 times forward earnings despite sitting at its 52-week high.
The deal, stripped of the drama
The terms are straightforward. Samsung agreed to abolish the long-standing 50% cap on bonuses (meaning bonuses can now exceed half of annual salary) and to formally link bonus pools to operating profit. Semiconductor employees - the most lucrative segment - get a new bonus pool worth 10.5% of operating profit, paid in stock. That last detail matters: management aligned compensation with the same currency the market rewards, which softens the cash cost.
SK Hynix, Samsung's South Korean memory rival, had already set aside 10% of operating profit for bonuses in 2025. Samsung workers rejected a $340,000 one-time bonus offer earlier this month and explicitly demanded annual recurring payouts like their SK Hynix counterparts. The 10.5% figure was Samsung's concession to close that gap.
Against Q1 2026 operating profit of 57 trillion won - up 185% quarter-over-quarter and more than the company's entire full-year 2025 operating profit - 10.5% of the memory division's annual take is roughly $4–5 billion at current profit levels. Analysts project full-year 2026 operating profit of approximately 166 trillion won ($100 billion), up 282% year-over-year. The bonus pool grows with profits, which is the union's victory, but it also means the cost scales proportionally rather than eating into margin at a fixed rate.
That's the arithmetic the market had already processed. Samsung shares jumped roughly 6% when the tentative deal was announced on May 20. Not a dramatic rerating - more like a risk-premium unwind. The strike itself threatened billions in production losses across Samsung's chip fabs. The deal removes that binary risk. What's left is the earnings story.
The compressed multiple
Here's what the factor stack actually says. Samsung's ADR is trading at approximately $140, right at its 52-week high, after a 250% run from the September lows. The TTM P/E sits around 16.8x. But the forward P/E - the multiple that matters when earnings are accelerating this fast - is roughly 6x. For comparison, SK Hynix trades at a similar forward multiple, and both sit well below TSMC, which carries a significantly higher valuation.
A 6x forward P/E on a company whose Q1 operating margin in the Device Solutions division exceeded 70% is not a number that reflects the market's full confidence in durability. It reflects a memory-cycle discount. The market knows memory is cyclical. It knows the AI-driven demand spike in HBM and high-bandwidth memory could cool. It prices accordingly.
But the question isn't whether the cycle will turn eventually. It's whether the current inflection justifies the position. Samsung is spending $73 billion on capital expenditures this year - a record investment to rebuild its HBM and foundry capacity. That's the counterweight: massive reinvestment that eats free cash flow even as operating profit soars. The stock isn't cheap because the business is weak. It's cheap because the market sees the capital intensity as a hedge against earnings durability.
The non-obvious risk
The union deal isn't the headline risk anymore. The real fracture line is internal. The non-semiconductor union division - workers outside the chip business - asked a Korean court on May 26 to block the vote, arguing the deal unfairly concentrated gains in the ultra-profitable chip segment. Roughly 12,000 members were reportedly angered by the compromise, and the bargaining representative revoked voting rights from hardline opponents to push the tally over the finish line.
For an investor, that's a structural reminder: Samsung's labor model is now profit-proportional, not flat. As chip margins compress in a downcycle, bonuses compress with them. When margins expand, bonuses expand. The union got its linkage, but it also accepted cyclical exposure. That's not a red flag for 2026 or 2027 - it's something that could resurface when the memory cycle turns.
So what does the stack say to do?
The strike overhang is gone. The bonus obligation is priced at a manageable percentage of record profits. The stock at 6x forward earnings on 282% profit growth is not an expensive number, even at a 52-week high. But it's not a blind buy either - the $73 billion capex cycle means free cash flow won't match operating profit, and memory cyclicality hasn't disappeared; it's just on the upswing.
If you already hold Samsung, there's no factor-driven reason to exit. The labor risk was always a tail event, and the market has been discounting cyclicality since the first quarter. If you're adding, size the position to the cycle risk, not the headline growth. The 6x forward multiple gives you room, but room and certainty aren't the same thing. Watch the Q2 guidance when it arrives - that's the next data point that tells you whether the AI memory demand story is accelerating or plateauing. The union deal is settled. The earnings cycle isn't.
The best outcome for this stock isn't a strike-free environment. It's a strike-free environment where the 70% DS division margin holds and the $73 billion investment pays off in HBM market share. The multiple compresses that scenario into 6x. Whether that's too low or just right depends on what happens next quarter.