Samsung Electronics Strike vs Memory Supercycle — Can Higher Prices Really Offset the Walkout? The Real Question Is 'Who Captures the Excess Profit'

Fri, 15 May 2026 00:00:00 +0000

📚 Samsung Electronics series Previous: Samsung Electronics and Citi’s KRW 460,000 target — starting from what HBM actually is

Samsung Electronics’ union has announced an 18-day general strike from May 21 — up to 50,000 participants, the largest in the company’s history. At the same time, memory prices are exploding: 2Q DRAM contract +58–63%, NAND +70–75%. JP Morgan calls it “price gains offset the strike.” That logic is only half right. The real fight is not “will the fab stop?” but “in this memory supercycle, who captures the excess profit — labor or shareholders?”

Key takeaways

Where we stand: Samsung Electronics union has notified a general strike for May 21 – June 7. Core demands: a fixed payout of 15% of DS (semiconductor) division operating profit as performance bonus, plus removal of the 50%-of-base-salary cap. Defenses still in play: May 16 NLRC mediation, court injunctions, and the government’s emergency arbitration power.
Loss estimates: JP Morgan KRW 21–35 trillion; industry estimates cited by Yonhap range up to KRW 100 trillion. The wide spread reflects unknowns in participation, line disruption, and restart time.
Price-offset logic: with DRAM +58–63% and NAND +70–75% in 2Q, a 1–2% production hit is mathematically offsettable. JP Morgan keeps its KRW 350,000 target on this basis.
But the offset is not complete: (1) structural labor cost from a bonus formula can’t be papered over by price; (2) if customers shift orders to SK hynix or Micron, Samsung captures less of the price rally.
The real question: not “strike or no strike,” but “how the profit-sharing formula changes — that matters more for medium- and long-term valuation.”

1. What is happening right now

1.1 Scale of the strike

The union notified an 18-day general strike from May 21 to June 7. Expected participation up to 50,000 employees — the largest strike risk Samsung Electronics has ever faced.

NLRC mediation sessions on May 11–12 collapsed in the early hours of May 13. The union’s position: “without concrete institutionalization, there is no reason to talk.”

1.2 What the union actually wants — not a simple pay hike

Core union demands:

1. A FIXED bonus = 15% of DS (semiconductor) operating profit
 → Current: bonus set at management discretion (EVA-based OPI)
 → Demand: 15% of operating profit codified as a formula

2. Remove the 50%-of-base-salary cap
 → Current: total bonus cannot exceed 50% of annual base salary
 → Demand: no cap. If profit is large, bonus scales accordingly.

3. Expand OPI stock-based compensation

Why this matters:
→ Samsung 1Q DS operating profit was KRW 53.7 trillion
→ 15% = \~KRW 8.1 trillion of bonus pool
→ Even 10% = \~KRW 5.4 trillion
→ The gap vs. the existing OPI scheme is measured in trillions

1.3 Why this issue blew up

First, the comp gap with SK hynix. Following its HBM success, SK hynix has materially expanded employee bonuses. Inside the same memory industry, Samsung employees feel relatively underpaid.

Second, the profit pool got too large to ignore. 1Q DS operating profit alone was KRW 53.7T at a 65.7% margin. When excess returns are this large, the question “who gets it” becomes politically unavoidable.

Third, fab production is uniquely fragile. A semiconductor fab runs 24/7 continuous flow. Stops mean scrapped wafers, lower yields, and long restart times. Even the threat of a strike already affects operations. Press reports indicate Samsung has begun a “warm-down” — cutting new wafer starts in anticipation.

2. How big is the loss — and why the spread is so wide

2.1 The estimates

Source	Estimate	What it captures
JP Morgan	KRW 21–35T	Labor cost + production disruption
JP Morgan revenue opportunity loss	~KRW 4.5T	Direct revenue loss only
Industry estimates (via Yonhap)	KRW 40T direct, up to KRW 100T including indirect	Worst case

2.2 Why the range is so wide

Unknown variables:
1. Real participation rate — of the 50k, how many actually walk out
2. Critical line staff — do DS process engineers join
3. Essential personnel — does the court injunction carve out safety
 and critical-line workers
4. Duration — full 18 days, or settled in a few days
5. Restart time — once a line stops, how long to bring it back

Quick sanity check:
Yonhap's reference: "about KRW 2.2T loss per day"
= KRW 40T direct ÷ 18 days = KRW 2.2T / day

2018 Pyeongtaek power outage: \~KRW 50B for 28 minutes
→ Day-rate equivalent \~KRW 2.6T
The two numbers are the same order of magnitude — a useful cross-check,
though the 2018 outage and a 2026 strike are structurally different,
so don't apply it mechanically.

3. Can price gains really offset the strike — JP Morgan’s logic

3.1 What JP Morgan said

In its May 13 note, JP Morgan (Jay Kwon’s team) maintained Overweight, target KRW 350,000. Core argument: “memory prices are stronger than expected and can absorb much of the strike loss.”

Item	JP Morgan
2Q DRAM contract price	+58–63%
2Q NAND contract price	+70–75%
Prior expectation	+40–50%
Production disruption (% of DS revenue)	1–2%

3.2 Does the math actually work

Memory revenue = shipments × average selling price (ASP)

If the strike reduces shipments,
revenue is defended whenever ASP gains exceed the volume loss.

Offset condition:
(1 + ASP growth) × (1 - volume loss) ≥ 1

If production loss is 1–2%:
→ ASP growth needed to defend revenue ≈ 1–2%
→ Actual ASP change: DRAM +58–63%, NAND +70–75%
→ Price gains dwarf the needed amount by tens of times

∴ For DIRECT production loss alone, the math offsets cleanly.

The more important calculation — the “excess” price gain vs prior expectations:

Prior expectation: DRAM +45%, NAND +45% (midpoint)
Updated view: DRAM +60.5%, NAND +72.5% (midpoint)

Incremental ASP:
DRAM: 1.605 / 1.45 - 1 = +10.7%
NAND: 1.725 / 1.45 - 1 = +19.0%

→ DRAM ASP is \~11% stronger than the prior model;
 NAND ASP is \~19% stronger.
→ This delta is what offsets the strike loss.

3.3 What’s right and what’s missing

Right: direct production loss (~1–2% of DS revenue) is comfortably covered by price gains. 2Q estimates could actually move up. For short-term earnings, JP Morgan’s logic holds.

Three things it misses:

1. Structural labor cost is not a "price" problem.
 If the demand for "15% of DS operating profit as a fixed bonus"
 is codified, then every future quarter has 15% of operating profit
 bleed into compensation. That's not offset by higher prices —
 prices and labor cost rise together.

 At 1Q levels: 15% = KRW 8.1T
 → This is not "offset by price." It's a SHIFT in how profit
 is divided between labor and shareholders.

2. If customers shift orders, the price rally bypasses Samsung.
 HBM, high-capacity server DRAM, enterprise SSD — once a customer
 moves a qualified supplier slot to SK hynix or Micron, it is
 slow and expensive to move it back.

 Samsung strike → delivery uncertainty → customers re-allocate
 to SK hynix / Micron
 → Memory prices keep rising, but Samsung captures less of it

 TrendForce flagged exactly this: order shift, not direct loss,
 is the bigger risk.

3. The memory cycle eventually turns.
 A fixed-ratio bonus is fine in an upcycle — and rigid in a downcycle.
 That's a long-run cost-flexibility loss → structurally lower ROIC.

4. Will the strike actually happen — there are firebreaks

4.1 Key events over the next week

Date	Event	Significance
May 15, 10:00	Deadline for the CEO’s response to the union	Tone can pivot here
May 16	NLRC follow-up mediation reopens	Last formal negotiation window
Around May 20	Court ruling on injunction	Whether essential workers are excluded
May 21	Strike start date	Actual participation rate is the swing factor

4.2 A full shutdown is unlikely

Firebreak 1: NLRC mediation (May 16)
→ Last formal chance to settle — both sides have face-saving incentives

Firebreak 2: Court injunction (around May 20)
→ Carving out safety and essential-process staff keeps critical lines running

Firebreak 3: Government emergency arbitration
→ If invoked, forces a 30-day suspension of strike action
→ Industry minister has publicly said "if the strike happens, it's unavoidable"

Firebreak 4: Both sides' own incentives
→ Union: 18 full days off-payroll hurts members too
→ Company: a shutdown damages customer trust, the share price, and exports
→ Government: semis were 37% of April exports — national economic stakes

Base case: a full shutdown is likely to be averted. But “no shutdown” and “no risk” are different statements.

5.1 Three scenarios and price impact

Scenario	Description	Price impact
A. Settlement / suspension	Strike averted via NLRC mediation or government intervention. Bonus stays under existing OPI + special payout	Short-term bounce — risk premium removed, memory cycle back in focus
B. Partial strike	Some staff walk out; critical lines maintained; participation modest	Neutral to slightly negative — noise persists, earnings impact limited
C. Major strike	DS critical staff participate; warm-down deepens; customers re-allocate	Material drawdown — estimate cuts plus multiple de-rating

5.2 More important than the scenario — the SHAPE of the bonus deal

Even after the strike ends, the question stays:
"How was the bonus institutionalized?"

Best case (shareholder-friendly):
→ Keep existing OPI framework + cycle-driven discretionary top-up
→ Workers get a share of excess profit, but not a fixed ratio
→ No cost rigidity in a downcycle

Worst case (shareholder-unfriendly):
→ 15% of DS operating profit codified as a fixed bonus
→ Every quarter, 15% of operating profit flows to comp before shareholders
→ Fine in an upcycle; in a downcycle, harder to defend earnings
→ Structural decline in long-run ROIC

How this formula gets decided matters more for Samsung’s medium- and long-term valuation than whether the strike happens. A strike that ends in a day still locks in lower shareholder economics if the bonus is fixed at 15%.

6. Where does the stock actually sit

6.1 Earnings are very strong

As covered in the prior post:

1Q revenue KRW 133.9T, operating profit KRW 57.2T
DS operating profit KRW 53.7T, margin 65.7%
Annualized operating profit ≈ KRW 228.8T → trailing-style PER ~9.6x at the current price

6.2 Sell-side outlook

Broker	2Q operating profit	Target price
Kiwoom Securities	KRW 100T	KRW 330,000
Yuanta Securities	KRW 86.8T	KRW 340,000
JP Morgan	—	KRW 350,000

6.3 Where the price sits

May 14 close: KRW 296,000. Implied upside to Kiwoom’s target = +11.5%.

What the current price reflects:
→ The memory supercycle (1Q margin 65.7% confirms it)
→ Strike risk — partially (the "price offsets" hope is baked in)
→ Structural change to bonus formula — NOT priced

Upside +11.5% against a real event risk:
→ The asymmetry for a new buy is weak
→ Target hit = +11%; strike escalation = -10–15% plausible
→ This is a "great company," not a "great entry price"

7. What to watch — checklist

#	Event	What to check
1	May 15, 10:00 CEO response	Union reaction — chance to reopen talks
2	May 16 NLRC mediation	Does a middle-ground bonus proposal emerge
3	Around May 20 Court ruling	Scope of essential-worker carve-out
4	May 21 Strike start	Actual participation, DS critical-line impact
5	After the strike	Wafer starts, shipment delays, customer allocation changes
6	Bonus settlement	Fixed-ratio vs discretionary — drives long-run ROIC

8. How this links to other posts

Samsung Electronics part 1 (Citi target KRW 460,000):
→ "The memory-cycle frame is wrong — AI is creating structural demand"
→ If that thesis still holds, the strike is noise.

Samsung Electro-Mechanics piece:
→ "MLCC / FC-BGA are also AI-infrastructure components in shortage"
→ A Samsung strike that cuts chip supply → DRAM / NAND price up
→ Indirectly tightens demand for SEMCO AI-server MLCCs as well

Jeju Semiconductor piece:
→ "Commodity DRAM is short because of AI"
→ A Samsung strike → further legacy DRAM supply tightness → prices up
→ Could actually help Jeju Semiconductor

US-China summit piece:
→ "We're in the regime where you have to find the least-priced exposure"
→ Samsung strike risk could translate into relative SK hynix upside

9. The one-line bottom line

The Samsung Electronics strike is not really about “does the fab stop.” It is about “who captures the excess profit of the memory supercycle — workers or shareholders.”

In the short term, JP Morgan’s logic holds: with DRAM +60% and NAND +73%, a 1–2% production loss is mathematically offset. But if 15% of DS operating profit gets locked in as a fixed bonus, that is not a one-quarter cost to “offset” — that is a change in how profit is divided. And if customers move HBM and server-DRAM allocations to SK hynix or Micron, Samsung captures less of the rally, even though prices keep rising.

A full shutdown is unlikely — NLRC mediation, court injunctions, and government emergency power all sit as firebreaks. But “no shutdown” and “no risk” are not the same thing. Holders need not overreact; new buyers are better off waiting for the May 16 mediation outcome and the May 21 strike decision.

The need today is event-risk management, not courage.

This article is research and commentary only and is not investment advice. Samsung Electronics 1Q figures (revenue KRW 133.9T, operating profit KRW 57.2T, DS operating profit KRW 53.7T) are from the company’s official newsroom disclosure. The strike notice (May 21 – June 7, 18 days) and union demands (15% of DS operating profit as a fixed bonus) are based on Yonhap and Yonhap Infomax reporting. JP Morgan’s analysis (target KRW 350,000, KRW 21–35T impact) is per Money Today and Reuters coverage; full text of the note has limited public access. DRAM +58–63% and NAND +70–75% are TrendForce estimates. Industry estimates of KRW 40T direct / up to KRW 100T including indirect are per Yonhap and reflect a worst-case scenario. Emergency arbitration authority rests with the Minister of Employment and Labor under Korea’s Trade Union Act. Strike participation rate, line disruption, and customer allocation changes are not knowable at this point. The analysis may be wrong. Data cut-off: May 15, 2026 KST.

Disclaimer: For research and information purposes only. Not investment advice. Names cited are for analytical illustration; readers should perform their own due diligence and consult licensed advisors before any investment decision.

Strike on Korea Invest Insights