Context This post follows our work on Korea market liquidity and narrow leadership, quality names to revisit in Q3, Korean Semiconductor Top 50 two-month alpha candidates and target-price gap screens. It belongs in the Exclusive Analysis Hub.
TL;DR
KRW weakness is supportive for Korean exporters into Q2 earnings. But the trade is not “buy all exporters.” The better screen is narrower: high dollar revenue, meaningful won-denominated costs, improving Q2 volume or pricing, and estimates that have not fully caught up.
Using FRED DEXKOUS daily observations, the 2026 Q2-to-date USD/KRW average through June 12 is 1,493.7. That is 1.9% above the Q1 average of 1,465.6 and 6.7% above the 2025 Q2 average of 1,399.8. The June 1-12 average is even higher at 1,526.2.
The final ranking from our screen is: Kolmar Korea, PSK Holdings, Wonik QnC, FADU, LIG Nex1, HD Hyundai Heavy Industries and Pearl Abyss. The cleanest mix is Kolmar Korea. Within semis, PSK Holdings and Wonik QnC look more comfortable than the already-crowded mega-cap and substrate names.
1. The FX Setup Is Real
We recalculated the USD/KRW averages directly from FRED DEXKOUS daily observations. FRED is business-day data, so the averages use available observations.
| Period | Observations | Average USD/KRW | Window |
|---|---|---|---|
| 2026 Q1 | 61 | 1,465.6 | Jan 2-Mar 31, 2026 |
| 2026 Q2 to date | 52 | 1,493.7 | Apr 1-Jun 12, 2026 |
| 2025 Q2 | 63 | 1,399.8 | Apr 1-Jun 30, 2025 |
| Jun 1-12, 2026 | 10 | 1,526.2 | Jun 1-Jun 12, 2026 |
The math is straightforward: Q2-to-date is 1.9% above Q1 and 6.7% above the prior-year Q2. For exporters with dollar revenue and won costs, this creates a translation and margin tailwind, although actual P&L sensitivity still depends on hedging, dollar cost, inventory, shipment timing and revenue recognition.
2. Export Data Also Supports The Setup
The FX story would not matter if volume were weak. But recent export data also points in the same direction. Korea’s May exports rose 53.2% year over year, with semiconductor exports up 169.4%. May ICT exports rose 128.9%, ICT semiconductor exports rose 169.2%, and computers and peripherals rose 259.6%. For June 1-10, total exports rose 85.9%, semiconductors 205.8%, and computer peripherals 259.4%.
| Data point | Growth | Read-through |
|---|---|---|
| May total exports | +53.2% YoY | Export demand is not weak |
| May semiconductor exports | +169.4% YoY | Semis have price, volume and mix support |
| May ICT exports | +128.9% YoY | AI, semis and peripherals lead |
| May ICT semiconductors | +169.2% YoY | Memory and system demand are strong |
| May computers/peripherals | +259.6% YoY | Server and storage read-through to check |
| Jun 1-10 total exports | +85.9% YoY | June started strong |
| Jun 1-10 semiconductors | +205.8% YoY | Semi momentum persisted into late Q2 |
| Jun 1-10 computer peripherals | +259.4% YoY | AI server/storage linkage remains relevant |
The investor question is not who benefits from FX. It is who benefits from FX without the benefit already being fully reflected in price and consensus.
3. Where The Structure Is Best
| Segment | FX effect | Earnings-surprise probability | What to check |
|---|---|---|---|
| Semis/HBM/memory | Dollar revenue, partial won costs | High, but expectations are high | Pricing, shipments, HBM mix, inventory marks |
| MLCC/substrates/AI parts | Dollar revenue and premium mix | High if price and volume improve together | AI mix, product mix, orders |
| Defense exports | Dollar/euro revenue, won costs | Medium-high | Delivery schedule, revenue recognition, project margin |
| Shipbuilding | Dollar backlog | Medium | Hedging, long-cycle recognition, steel cost, progress |
| Games/software | Overseas revenue, won labor cost | Potentially high | New title, traffic, launch timing, monetization |
| Cosmetics/food exports | Export revenue and won costs | Medium-high | US/Japan/China channels, raw materials, marketing |
The best setup has three traits: dollar revenue and won costs; improving Q2 volume or price; and recent consensus upgrades that have not fully moved the share price.
4. Priority Candidates
Source: Korea Invest Insights internal screen, as of June 19, 2026 for price, flow and consensus. Flow is trailing 20 trading days in KRW 100mn units. OP is in KRW 100mn.
| Rank | Name | Core logic | Q2 OP | FY26 P/E | 20D return | 20D flow | View |
|---|---|---|---|---|---|---|---|
| 1 | Kolmar Korea | Cosmetics exports, won cost base, lower valuation burden | 923 | 12.2x | -4.4% | Foreign +482 / Inst. +133 | Buy candidate |
| 2 | PSK Holdings | HBM/back-end equipment, high margin, institutional flow | 183 | 20.3x | +3.4% | Inst. +346 | Buy candidate |
| 3 | Wonik QnC | Semi consumables, utilization recovery, less crowded | 280 | 15.5x | +4.9% | Inst. +240 | Buy candidate |
| 4 | FADU | SSD/storage re-rating, strong foreign/program flow | 93 | 113.4x | -3.6% | Foreign +4,120 / Inst. -2,866 | High-beta candidate |
| 5 | LIG Nex1 | Defense exports, strong institutional and real-money flow | 1,052 | 53.3x | +2.6% | Foreign +340 / Inst. +1,172 | Valuation gated |
| 6 | HD Hyundai Heavy Industries | Shipbuilding, defense and nuclear option, KRW tailwind | 9,933 | 23.8x | +14.4% | Foreign -1,440 / Inst. +1,746 | Pullback candidate |
| 7 | Pearl Abyss | Overseas revenue, won labor cost and FX leverage | 1,450 | 6.8x | -17.0% | Foreign +79 / Inst. -173 | Event candidate |
5. Name-By-Name Judgment
Kolmar Korea is the cleanest mix. It has export-linked beauty manufacturing exposure, a won cost base, a reasonable FY26 P/E of 12.2x, and a -4.4% 20-day price move despite positive foreign and institutional flow. The main diligence items are whether overseas demand flows through to margin and whether raw materials or marketing costs offset the FX benefit.
PSK Holdings and Wonik QnC are the preferred semi names for this specific FX-alpha screen. They are not as crowded as Samsung Electronics, SK Hynix, Samsung Electro-Mechanics, Daeduck Electronics, Simmtech or Korea Circuit. PSK Holdings adds HBM/back-end equipment exposure and institutional buying. Wonik QnC adds semi consumables and utilization recovery with a lower valuation burden.
FADU has the biggest right-tail risk but also the largest risk. The foreign buying is powerful, but the institution selling is equally important. This is not a clean long-only accumulation signal. It is a high-beta storage re-rating option that needs SSD controller revenue and customer diversification to show up.
LIG Nex1 has a good defense-export and FX structure, but valuation is the gating item. At 53.3x FY26 P/E, the company needs more than FX. It needs export orders, project margin and evidence that the European/NATO air-defense channel can become a durable pipeline.
HD Hyundai Heavy Industries benefits from a dollar backlog, but shipbuilding FX effects are slower because of hedging, progress accounting, contract duration, steel cost and labor. Pearl Abyss has clear overseas-revenue and won-cost leverage, but the stock is really a new-title event story, not a pure FX story.
6. Lower-Priority And Excluded Names
| Name | Reason |
|---|---|
| iFamilySC | Low P/E and export-consumer logic, but flow size is small |
| VT | Oversold export-consumer candidate, but flow confirmation is needed |
| Silicon2 | Export platform structure is attractive, but recent flow is weak |
| ST Pharm | Possible dollar-revenue structure, but biotech flow needs to recover |
| TLB | SOCAMM/substrate candidate, but foreign flow is weak |
| HD Hyundai | Undervalued shipbuilding holding company, more NAV/beta than direct exporter |
| Name | Reason for exclusion or chase discipline |
|---|---|
| Samsung Electronics / SK Hynix | Q2 surprise probability is high, but expectations and price already reflect a lot |
| Samsung Electro-Mechanics | Strong thesis, but price and target upside are less comfortable |
| Daeduck Electronics | Good company, but 20D +24.8% and limited target upside |
| Simmtech | SOCAMM expectation reflected, 20D +29.2% |
| Korea Circuit | 20D +68.3%, weaker new-entry expected value |
| PSK / Eugene Tech / TES | Fundamentals may be good, but short-term price moves are already hot |
7. Final Ranking
| Rank | Name | Action | Reason |
|---|---|---|---|
| 1 | Kolmar Korea | Buy candidate | Best balance of FX, exports, valuation, pullback and flow |
| 2 | PSK Holdings | Buy candidate | HBM back-end exposure, institutional flow and lower chase risk |
| 3 | Wonik QnC | Buy candidate | Semi consumables recovery, lower valuation and less crowded setup |
| 4 | FADU | High-beta candidate | Foreign buying and storage re-rating, but high P/E and institution selling |
| 5 | LIG Nex1 | Valuation gated | Defense-export quality is strong, valuation is the hurdle |
| 6 | HD Hyundai Heavy Industries | Pullback candidate | Good strategic exposure, but FX P&L recognition is slower |
| 7 | Pearl Abyss | Event candidate | FX structure is good, but the real driver is the new-title event |
The portfolio-manager summary: Q2 KRW weakness is supportive for Korean exporters, but generic exporter exposure is too obvious. The better candidates are companies where FX works with volume, price or margin and where consensus has not fully caught up. Kolmar Korea is the cleanest current setup. In semis, PSK Holdings and Wonik QnC are more comfortable than crowded mega-cap or substrate names. FADU has convexity but needs limited sizing. LIG Nex1 needs valuation discipline.
Missing Evidence
This is a screen, not a final trade ticket. The next layer should check company-level hedge ratios, exact dollar revenue and dollar cost exposure, the broker notes behind Q2 consensus, trailing five-day detailed investor flow, and actual entry/exit levels.
Sources And Data
- FRED DEXKOUS daily observations through June 12, 2026. FRED DEXKOUS
- Korea May export data: total exports +53.2%, semiconductors +169.4%. KDI Economic Information Center
- Korea May ICT exports: ICT +128.9%, ICT semiconductors +169.2%, computers/peripherals +259.6%. MOTIR
- Korea June 1-10 export data: total exports +85.9%, semiconductors +205.8%, computers/peripherals +259.4%. KDI Economic Information Center
- Stock-level price, flow and consensus: Korea Invest Insights internal screen, as of June 19, 2026. No internal filesystem path is exposed in this public post.