Related series AI Back-End 11-Stock Data Comparison · Jeju Semiconductor 1Q26 Deep Dive · May 15 Flow Accumulation Screen · Korea Semiconductor Value-Chain Hub

Both stocks exploded in 1Q26. Hana Micron delivered operating profit of KRW 72.0bn; Jeju Semiconductor delivered KRW 67.1bn. On the first trading day after each announcement, the stocks gained +18.6% and +8.9% respectively. On the surface both look like “AI-era memory back-end beneficiaries.” But the two surprises are structurally different. Hana Micron is closer to a business-model improvement story — repriced cost-pass-through at the Vina entity, margin step-up at the Brazil subsidiary. Jeju Semiconductor is closer to a cyclical windfall — LPDDR4X supply scarcity amplified by tariff-driven front-loading. The same label, “earnings surprise,” masks very different durability profiles. That distinction is the whole point of this note.

Key Takeaways

• Both sets of numbers were strong. Hana Micron: 1Q26 revenue KRW 507.7bn, OP KRW 72.0bn, OPM 14.2%. Jeju Semiconductor: revenue KRW 180.5bn, OP KRW 67.1bn, OPM 37.2%.
• The nature of each surprise is different. Hana Micron’s upside was driven by a repriced cost-pass-through structure at the Vina entity and a margin step-up at the Brazil subsidiary. Jeju Semiconductor’s upside was driven by LPDDR4X supply scarcity and tariff-driven pre-orders.
• Durability favors Hana Micron. The 1Q margin cannot all be annualized on repeat, but Hana Micron may have established a structurally higher margin floor. Jeju Semiconductor’s 37.2% OPM is difficult to call a new normal.
• Multiple illusion is real for Jeju. Simply annualizing 1Q net income implies a 9x PER — but that requires four identical quarters. Adjusting for normalization pushes the expected-value multiple to 16–17x.
• Both are clearly overbought near-term. Hana Micron: +58.6% over 20 days, RSI ~78. Jeju Semiconductor: +94.3% over 20 days, RSI ~81.5, and +574% from the 52-week low.
• Entry priority: Hana Micron > Jeju Semiconductor. But chasing either immediately after the print is inefficient. Real alpha lies in confirming that the surprise is not one-off — i.e., 2Q26 results — and buying into any post-confirmation weakness in tranches.

1. The Numbers Alone Were Explosive for Both

Taken at face value, both companies qualify as “earnings surprises.” Hana Micron beat consensus operating profit by roughly 30%. Jeju Semiconductor grew operating profit nearly 18-fold year-over-year.

Looking only at this table, Jeju Semiconductor appears cheaper — higher OPM and a lower annualized PER. But that arithmetic requires 1Q26 to repeat identically in Q2, Q3, and Q4. In memory cycles, that assumption is among the most dangerous you can make. A quarter driven by supply scarcity and pre-ordering can reverse within one reporting period.

The real question, then, is not “who had the better quarter?” It is: how repeatable is that profit?

2. Hana Micron’s Surprise Is Closer to Structural Improvement

The two pillars behind Hana Micron’s KRW 72.0bn in 1Q26 operating profit are (1) a repriced cost-pass-through mechanism at the Vina entity and (2) a margin step-up at the Brazil subsidiary.

At Vina, the company appears to have established a more stable arrangement under which raw-material cost increases are reflected in the unit prices charged to SK hynix. Traditional OSAT contracts often require the packaging house to absorb input cost volatility within a fixed unit price — so when substrate or wire prices rise, revenue grows but margins compress. If raw-material escalations are now passed through more reliably, the dynamic shifts: Hana Micron begins to share price volatility with its anchor customer rather than bearing it alone. That is a business-model change, not a one-quarter fluke.

The Brazil subsidiary is equally important. 1Q26 Brazil revenue is estimated at roughly KRW 105.6bn, with OPM in the high-teens — well above the mid-to-high single digits typical for mainstream OSAT. Regional positioning, contract structure, favorable currency translation, and a higher mix of value-added packaging appear to have combined to produce this level.

Caveats are real. Non-operating FX gains of approximately KRW 27.2bn are unlikely to repeat, and incremental operating leverage diminishes as utilization approaches its ceiling. But the more important question is where the margin floor has moved. If Hana Micron’s prior cycle average OPM was 6–8%, and this cycle can sustain 11–13%, then earnings power has structurally shifted — and the stock is rerating in response to that shift, not just to a single quarter’s number.

3. Jeju Semiconductor’s Surprise Is Closer to a Cyclical Windfall

Jeju Semiconductor’s 37.2% OPM in 1Q26 is extraordinary even by fabless memory standards. The cause, however, looks more like a supply-demand shock than a structural improvement.

The first driver is LPDDR4X scarcity. Samsung, SK hynix, and Micron have been migrating production capacity toward HBM, DDR5, and LPDDR5/5X — higher-margin, AI-driven products. The natural consequence is a shrinking supply of the mature LPDDR4X node. But IoT devices, automotive applications, and a tail of industrial and mobile platforms still require LPDDR4X in volume. Supply falls; demand persists; prices rise. Jeju Semiconductor, as a focused legacy-memory fabless house, was the most direct beneficiary of that dislocation.

The second driver is tariff-driven front-loading. As uncertainty over U.S. semiconductor tariffs escalated, buyers moved to pre-build inventory. One quarter’s revenue can balloon when customers pull forward purchases — but the following quarter then works through that stockpile rather than placing new orders. Separating genuine demand growth from demand-borrowed-from-the-future is essential when reading any 1Q26 memory print.

This is not a criticism of the company. Jeju Semiconductor has real capabilities: LPDDR4X and MCP productization experience, customer qualifications, and an ability to serve niches that tier-1 suppliers have deprioritized. When those tier-1 players shift capacity away from legacy nodes, Jeju can capture pricing power quickly. The problem is that if the primary earnings driver is supply scarcity, margins fall rapidly when supply normalizes or when pre-ordered inventory is digested.

Consequently, the 37.2% OPM in 1Q26 is better read as a cycle-peak excess margin than as a new steady state. A normalized OPM in the 15–25% range over a 2–3 year horizon is a more realistic anchor for valuation.

4. Same Label, Different Durability

The sharpest summary of the contrast is this: for Hana Micron, the question is whether the pricing structure has permanently changed. For Jeju Semiconductor, the question is how long the supply shortage lasts.

This distinction matters for portfolio construction. Hana Micron is a question of whether the company’s earnings power has permanently risen. Jeju Semiconductor is a question of how long a supply-demand shock sustains. Both stories are live. The risk profiles are not the same.

5. Multiples: Where Illusion and Reality Diverge

Hana Micron’s market cap as of May 15 is approximately KRW 3.52tn. If 2026 full-year operating profit is modeled at around KRW 299.5bn, the Mkt cap/OP ratio is roughly 11.7x and the PER is approximately 19x. Not cheap on an absolute basis, but not demanding if structural improvement persists into 2027.

Scenario analysis for 2027 produces an expected value above the current price:

Weighting Bear/Base/Bull at 20%/50%/30% gives an expected value of approximately KRW 63,945 — roughly 21% above the current price. Not a screaming bargain, but still investable under a structural-improvement assumption.

Jeju Semiconductor is harder to value. Annualizing 1Q net income of KRW 78.1bn yields KRW 312.4bn in annual net income. Dividing the ~KRW 2.83tn market cap produces a 9.1x PER. That looks cheap. But it requires four identical quarters — an aggressive assumption.

If 2Q and beyond decelerate, the numbers shift materially. Modeling 2026 full-year OP at around KRW 200bn and converting to after-tax income pushes the implied PER to 16–17x. That is no longer inexpensive.

Weighting Bear/Base/Bull at 35%/40%/25% yields an expected value of approximately KRW 67,056 — below the current price. Adding the roughly 7.7% potential dilution from outstanding CBs and BWs compresses the Bull-case fair value further. Jeju Semiconductor is not a flawed company; it is a company whose current price already embeds significant optimism.

6. Near-Term Overheating Is Clear for Both

The danger of chasing immediately after a strong print is simple: good news is already aggressively priced in.

Hana Micron is up +58.6% over 20 trading days with an RSI of approximately 78. Jeju Semiconductor is up +94.3% over 20 trading days, RSI approximately 81.5, and is +574% from its 52-week low — enough to trigger an investment-warning designation. Even with genuine fundamental improvement, a new buyer entering here begins in the middle of peak volatility.

Near-term flows are clearly constructive. On May 15, during a broader market selloff, foreigners net-bought Hana Micron shares worth approximately KRW 91.8bn while retail sold KRW 93.8bn. Jeju Semiconductor also saw coordinated foreign and institutional buying. But “flows are strong” and “now is a good entry” are different statements. Strong flows signal that the name deserves attention; the right entry price is a separate question.

7. Positioning Within the Broader Semiconductor Back-End Universe

As covered in the AI Back-End 11-Stock Data Comparison, even within “AI back-end,” substrates, test sockets, memory packaging, and legacy memory are entirely distinct businesses with different margin structures, customer dynamics, and cycle exposures.

Hana Micron is primarily an OSAT story. As demand for high-value memory packaging — HBM, DDR5, eSSD — grows, both volume and unit pricing move in Hana Micron’s favor. Jeju Semiconductor is a memory fabless company, but the current tailwind is specifically “ordinary memory made scarce by AI capex” rather than AI-driven demand for advanced products directly. Both sit on the outer ring of the AI cycle. One is a back-end structural improvement; the other is a legacy-memory supply shock.

In this peer table, Hana Micron sits at a reasonable middle-ground valuation for comparable back-end names. Jeju Semiconductor appears cheap on the 1Q run-rate but requires a normalization scenario to assess properly — making it closer to a scenario bet than a straightforward value hold.

8. Practical Checkpoints

For Hana Micron, the critical variable in 2Q26 is OPM. If 1Q’s 14.2% was genuinely one-off, the margin will fall sharply in Q2. If OPM holds at roughly 13%, that is evidence that the Vina pricing structure and Brazil subsidiary margin are real, durable changes to the company’s earnings power.

Four checkpoints to monitor:

• 2Q26 OPM ≥ 13%
• Continued Vina revenue growth
• Brazil subsidiary sustaining high-teens OPM
• Rising share of DDR5/eSSD packaging within the revenue mix

For Jeju Semiconductor, the critical variable is the magnitude of 2Q deceleration. If Q2 operating profit holds above KRW 50bn with OPM above 30%, 1Q was not simply a one-time peak. If Q2 drops below KRW 40bn, the front-loading and scarcity effects are unwinding faster than hoped.

Five checkpoints to monitor:

• 2Q26 OP ≥ KRW 50bn
• OPM ≥ 30%
• Inventory growth decelerating
• Accounts receivable growth decelerating
• LPDDR4X spot pricing trend sustaining

On positioning, waiting for a pullback makes more sense than chasing for both names. For Hana Micron, the KRW 49,000–51,000 support zone is the first level to watch. For Jeju Semiconductor, the KRW 68,000–72,000 range or the post-2Q26 results window is a more realistic entry consideration. These are not buy-recommendation price targets — they are observation levels to gauge whether the current overheating has normalized.

9. If Forced to Choose One

For a 12-month-plus holding period with emphasis on earnings durability, Hana Micron has the edge. If the Vina pricing structure and Brazil subsidiary margins hold, 2027 earnings estimates have upside. Even in a down-cycle, the structural changes may produce a higher margin floor than in prior cycles.

For a 3-month high-volatility trade, Jeju Semiconductor offers more explosive potential movement. But that trade requires correctly calling a cycle peak. The Bull case has meaningful upside; the Bear case has -50% drawdown risk from current levels. Factor in CB/BW dilution and the risk/reward tightens further.

If holding both, a core-satellite structure is more natural: Hana Micron as the core (60–70%) and Jeju Semiconductor as the satellite (30–40%). A concentrated single-name position in Jeju Semiconductor at current prices carries substantial volatility risk.

10. One-Line Summary

Hana Micron and Jeju Semiconductor both delivered strong 1Q26 results. They are not the same kind of surprise.

Hana Micron is a structural improvement story. The Vina cost-pass-through repricing, the Brazil high-teens OPM, and growing SK hynix-related packaging volume moved together. The 1Q margin cannot be fully normalized to an annualized run-rate, but the evidence suggests a higher earnings floor than in prior cycles.

Jeju Semiconductor is closer to a cycle peak. The 37.2% OPM was produced by LPDDR4X supply scarcity and tariff-driven front-loading. The company’s product capabilities and niche-market positioning are genuine, but treating 37.2% OPM as a sustainable steady-state is the wrong frame.

On multiples, Hana Micron looks more attractive. Jeju Semiconductor’s 1Q run-rate PER of 9x is an illusion — it requires four identical quarters that are unlikely to materialize. Adjusting for normalization pushes the expected-value PER to 16–17x. Hana Micron, under a structural-improvement scenario through 2027, still offers expected-value upside from the current price.

Neither stock is a good chase right now. The most dangerous move after a strong earnings print is buying immediately because the numbers look good. Real alpha comes from distinguishing the cause of the surprise, confirming durability in 2Q26, and entering in tranches after the post-announcement heat fades.

The one-line version: Hana Micron is a structural-improvement confirmation trade; Jeju Semiconductor is a cycle-peak timing call. New money priority favors Hana Micron, but chasing either before 2Q26 results is inefficient.

This article is for research and commentary purposes only and does not constitute investment advice. Hana Micron 1Q26 results (revenue KRW 507.7bn, OP KRW 72.0bn, OPM 14.2%) are sourced from company disclosures and Korea Investment & Securities/Meritz Securities research reports. Jeju Semiconductor 1Q26 results (revenue KRW 180.5bn, OP KRW 67.1bn, net income KRW 78.1bn, OPM 37.2%) are sourced from company disclosures. The +29.7% consensus beat figure is based on Korea Investment & Securities data. Jeju Semiconductor 2026E/2027E official consensus is not available in public sources; all scenario estimates in this article are analyst projections based on 1Q results, LPDDR4X supply-demand dynamics, and AI-edge transition potential. Vina cost-pass-through structure and Brazil subsidiary high-teens OPM are interpretations based on brokerage research materials. CB/BW outstanding of approximately KRW 117.0bn and potential dilution of approximately 7.7% are based on company disclosures and media reports. Scenario fair values and probability weights are subjective analyst estimates and may differ materially from actual outcomes. The timing of LPDDR4X supply normalization, success of AI-edge memory conversion, and 2Q26 earnings durability are all uncertain. This analysis may be wrong. Data as of May 17, 2026 KST.