AMKR · Amkor Technology, Inc. · Stock research

Current Thesis

The story changed on 2026-07-23, when Amkor disclosed a $1.5B collaboration with Nvidia to develop advanced packaging and test technologies for AI and accelerated-computing platforms — the stock jumped the following session (2026-07-24). Stacked on the 2026-06-16 TSMC 10-year Arizona advanced-packaging partnership, Amkor now carries two front-end anchors (TSMC leading-edge, Nvidia AI accelerators) attacking the loudest bear point: whether the $2.5-3.0B Peoria buildout finds utilization beyond a single Apple program. That is a genuine narrative upgrade from "single-customer Apple OSAT" toward "US AI back-end anchor." The problem gating it is twofold. First, buyers have sold every prior catalyst — the ~50% Q1 EPS beat printed -8.6% same-day (2026-04-27), and the TSMC-deal pop faded to -8.5% by 2026-06-25. Second, the Q2 '26 print lands ~2026-07-27 (after close, est.), a binary that decides whether the Nvidia-news jump toward the $74-75 shelf becomes a confirmed higher-low base or another faded catalyst. The advanced-packaging/AI-infra theme is accelerating at the sector level, but a fresh entry the session before a print in a name that fades every catalyst is a stand-aside pending confirmation.

Bullish and bearish views on Amkor Technology, Inc.

The model's bull view on Amkor Technology, Inc. (AMKR), in brief: Nvidia $1.5B partnership (2026-07-23): collaboration to co-develop advanced packaging and test for AI/accelerated-computing platforms; second hyperscaler-adjacent anchor after TSMC, widening the Peoria demand base beyond Apple. The bear view: Every catalyst sold: a ~50% EPS beat met -8.6% (2026-04-27) and the TSMC-deal pop faded to -8.5% by 2026-06-25. Both cases follow in full.

Bull Case

• Nvidia $1.5B partnership (2026-07-23): collaboration to co-develop advanced packaging and test for AI/accelerated-computing platforms; second hyperscaler-adjacent anchor after TSMC, widening the Peoria demand base beyond Apple. Shares jumped 2026-07-24.
• financial terms undisclosed.
• Record Q1 '26 (2026-04-27): net sales $1.68B (+27% YoY), operating income $100M, net income $83M, EPS $0.33 (beat ~50%); Communications +40% on Apple advanced packaging.
• Q2 '26 guide (2026-04-27): net sales $1.75-1.85B, net income $105-130M, gross margin 14.5-15.5% — a sequential lift off Q1's 14.2%.
• Apple anchor (Feb 2026): >100M TSMC-Arizona chips packaged at Amkor Peoria in 2026; the TSMC and Nvidia frameworks add customer breadth on top.
• Cheap funding (settled 2026-05-05): $1.0B 0%-coupon convertible due 2031, ~$982M net with capped calls — the buildout is financed without cash-interest drag.
• Sector tailwind: SMH +80% YTD (2026-05-26); semi-equipment names led higher 2026-07-09; Cramer flagged Amkor packaging 2026-06-01.

Bear Case

• Every catalyst sold: a ~50% EPS beat met -8.6% (2026-04-27) and the TSMC-deal pop faded to -8.5% by 2026-06-25. Strength has not been defended.
• Sell-side not chasing: B. Riley maintained Neutral and LOWERED its target to $75 on 2026-07-22, into the Nvidia news; Goldman sat at $65 in late April.
• Optionality, not booked revenue: the Nvidia deal is a development/collaboration framework and the TSMC deal was headline-only — neither carries a disclosed volume floor or revenue-recognition schedule.
• OSAT margin ceiling: Q1 gross margin 14.2%, Q2 guide 14.5-15.5%. Record revenue still converts to mid-teens margin at a ~41x P/E — no AVGO/NVDA-style multiple-expansion analog.
• Cash burn: FY26 capex $2.5-3.0B (65-70% Arizona) against ~$83M Q1 net income; the ramp runs FCF-negative through 2027 while fabless pure-plays generate cash.
• Apple concentration (~20%+ of revenue) is the single largest idiosyncratic risk; Communications +40% deepens the dependence.
• Competition: Intel's "burgeoning packaging division" (Cramer, 2026-06-01) plus larger Asian OSATs (ASE, SPIL, Powertech) contest the same advanced-packaging demand with more capacity.

Setup & Price Structure

• ATH $79.23 (2026-04-27); the -8.6% earnings-day reversal marked the parabola top.
• June low ~$64 (2026-06-05, after a -12% breakdown on the Broadcom AI-capex guide reset plus 172k payrolls killing rate-cut odds).
• ~$70.71 on 2026-07-10 (range $69.61-$71.84), recovering ~9% off the June low but capped below resistance.
• The Nvidia news (2026-07-23/24) drove a jump toward the $74-75 shelf — the level that must weekly-close above to confirm a higher-low base rather than an oversold bounce in a post-parabola downtrend.
• Structural support: $64 June low. A weekly close beneath it forfeits the entire June recovery.
• Analyst frame: B. Riley $75 (2026-07-22), Goldman $65, consensus ~$78 — spot sits only single-digit percent below the average target, thin upside fuel.

Catalyst Calendar (next 30 days)

• Post-print (late July, est.): any disclosed dollar terms or capacity commitments on the Nvidia and TSMC frameworks would convert optionality into modelable revenue.
• Rolling: hyperscaler capex prints (NVDA, AVGO) and Asian memory/foundry reads (SK Hynix, Samsung, ASML) set the sector tape Amkor trades with.

Elapsed catalysts

• ~2026-07-27 (after close, est.): Q2 '26 earnings — the binary. Grade net sales vs the $1.75-1.85B guide, gross margin vs 14.5-15.5%, the H2/FY guide, and any explicit Arizona customer #2 or Nvidia/TSMC volume commentary. (passed 13d ago)

What Would Change Our Mind

• Constructive flip: a post-earnings gap-up on a Q2 beat plus an FY raise, an explicitly named Arizona customer #2, and a weekly close above the $74-75 shelf on ~2x average volume — that confirms the base and re-rates Amkor to theme leader rather than laggard, justifying a fresh entry.
• Destructive confirm: a Q2 miss or a soft H2 guide, followed by a weekly close below $64, resumes the post-parabola downtrend and voids the recovery thesis.
• Theme risk: the advanced-packaging/AI-infra theme cooling to saturated — a hyperscaler capex guide-down that removes the sector bid — would strip the tailwind even on in-line company numbers.

Correlation Notes

• Trades as a high-beta AI-semi basket proxy alongside SK Hynix, Samsung, and hyperscaler capex (NVDA, AVGO); SMH +80% YTD (2026-05-26). The same beta cut it -12% on the 2026-06-05 Broadcom-guide/payrolls reset.
• OSAT peers ASE, and Powertech set the read-through on back-end pricing; Amkor chronically lags fabless pure-plays on AI-narrative days given the mid-teens margin structure.
• Idiosyncratic overlay: Apple at ~20%+ of revenue means single-customer or Apple-supply-chain headlines can move the name independent of the semi tape.

Setup Summary

Post-parabola OSAT laggard with an upgraded two-anchor narrative (TSMC + Nvidia) pressing the $74-75 shelf, but a binary Q2 print ~2026-07-27 gates confirmation and every prior catalyst has been sold — confirmation, not anticipation, is the entry.