CryptoWithMind

$CAH

Trade idea: Hana Micron (067310.KQ) at ~1.9B MC: The turnkey partner to Samsung & SK Hynix. High risk, high reward OSAT play. TLDR: - we’re in the strongest memory cycle in ~3 yrs w/ HBM having decoupled captive lines. - Hana are the memory-focused OSAT pure-play - Hana’s Vietnam biz has turned profitable after 4 years of losses = de-risked operations. - broader Korean semiconductor supply chain (Hanmi, Techwing, Tesna, Wonik IPS, etc.) is in a multi-yr capex+earnings cycle where Hana is the most memory-exposed OSAT. And to understand why Hana are important: - SK Hynix, Samsung + $MU all report HBM fully sold out through 2026–2027/2028. - advanced packaging (2.5D/3D, chiplets, hybrid bonding, CoWoS-like flows) is now the bottleneck for GPU/accelerator shipments. - OSATs like Hana capture this because hyperscalers demand integration that IDMs outsource. - since Hana are the memory-focused OSAT pure-play, it rides the highest-margin segment while OSAT peers (heavier in logic/auto) face more cyclical drag. This then rolls up into being potentially good timing w/ Korean K-CHIPS Act in H2 2026: subsidies for advanced packaging beyond $TSM / $INTC CHIPS Act parallels. They essentially operate one tier below the chokepoint in the HBM stacking value chain since that work stays captive at Samsung and SK hynix. Hana Micron take up the spillover: the more wafer-starts + labour that Samsung/SK hynix dedicate to HBM TSV stacking, the more standard DRAM/NAND/LPDDR/DDR5 assembly and final test gets pushed to Hana. So the bull case essentially rests on: - Samsung/SK hynix shifting commodity DRAM/NAND/LPDDR5/DDR5 packaging & back-end test out of constrained captive lines into Hana's plants while their internal capacity is repurposed for HBM. Hana Micron is also pivoting into 2.5D/3D heterogeneous integration and Large Body substrates: - that 2.5D approach uses bridge die + copper pillar instead of full silicon interposer. - this is patented + is said to be Materially cheaper than CoWoS - but currently unproven for HBM integrity. - could become a potential moat going forwards? But overall, great timing, just as the industry’s real chokepoint shifts from wafer fab to advanced packaging & test (along w/ photonics etc). Hana are going through a positive profit inflection rn: FY2025 Financials: - Revenue: a record $1.04B (+22.7% YoY) w/ 75% from packaging + test. - Net income: swung to $25.9M (from a 2024 loss) + Q4 beat estimates. - Net debt: $670M (D/E 1.84x; current ratio 0.85). - 2026 plan keeps capex roughly flat YoY while paying down debt. 2025 was the recovery/derisking year. Vietnam ramp (>$204M invested) shifted mix toward high-volume DDR5/server modules. But 2026-2027 is when advanced packaging scarcity meets Hana’s capacity + tech readiness against persistent memory shortages. They report Q1 earnings on May 18th too: - which will hopefully confirm continued momentum + profitability inflection. - some estimates put 2026 revenue at $1.32-1.56B (40-50%+ growth). If executed, this sets up 2-3x stock upside over next 1-2 yrs via earnings compounding & multiple re-rating. The stock has already 4x'd off 2024 lows, so seems like a lot of re-rating has been done already. So imo, a further 2-3x requires: - Samsung HBM4 share recovery + Samsung commits more to the DRAM capex cycle 2026-2028. - SK hynix Vietnam allocation to Hana keeps growing 30-50% annually through 2027. - HIC 2.5D wins a named hyperscaler/fabless qualification announcement before H2 2027. - Drop in net debt (why Earnings in mid-May are important). None of this seems wildly unrealistic to me rn. Obviously being so concentrated in Samsung/SK Hynix will be a concern for many. But Hana gains from their HBM/DDR5 sold-out status + roadmap co-development. If Samsung/SK Hynix ship more packaged memory to $NVDA etc, Hana’s utilization & ASPs rise too. That naturally makes Powertech probably the closest comparison rn: - Hana vs Powertech are two different investment strategies imo. - Powetech is $MU’s main HBM packaging partner + probs has higher quality earnings. - Hana is more risky - more levered but w/ faster top-line growth rn. Higher risk, higher reward? I don’t really see any fatal flaws rn other than the ones mentioned e.g. debt/customer concentration. But Hana is essentially a high-beta bet on sustained AI demand/capex flowing up into Memory names like Samsung/SK Hynix. Definitely a more high risk, high reward OSAT play. Disclosure: I have a position.

Agreed high-level directionally, $FLNC compelling at $3B valuation post-earnings after taking a closer look. Very rare to see a US energy player that small get 2 direct Hyperscaler deals... The $5.6B+ backlog derisks the company growth, not including new hyperscalers backlog like $GOOGL or $MSFT. The hyperscaler deals were framework agreements, which are likely to convert "soon" Q3 this year, and aren't included in numbers. Once that's released it's major positive catalyst, similar to qualification -> volume ramp in semi players. Citi Analyst: "The possibility of a hyperscaler order will likely overshadow everything else in the quarter. We expect a positive reaction to the announcement" I'm going to go ahead and guess they'll likely rerated once they announce their hyperscaler orders maybe anytime in the next 3 months so I jumped on the boat as a short term catalyst trade. (not just 1 but 2) Also, if they hit ~$288M net income off gross-margin expansion ($6B revenue, 13.0% gross margins) from their software segment expansion, ~11.6x fwd p/e for 2027. The current stock price is -50% Feb's prices despite hyperscalers + backlog de-risking the company looks like a great entry point to me (NFA).