FORM · FormFactor, Inc. · Stock research

Current Thesis

FormFactor is the probe-card pick-and-shovel levered to HBM and advanced-packaging test intensity, and the story just took its first real momentum hit. After running +112.8% YTD into a fresh all-time high of $160.27 (2026-06-30) printed alongside 2026-06-29 Russell 1000 inclusion the name got caught in the 2026-07-02 "chip bloodbath" and fell ~17.5% in early July to about $121.42 (2026-07-09). The fundamentals held: Q1 (reported 2026-04-29) guided Q2 to record HBM and a record DRAM probe-card quarter as a second customer ramps SmartMatrix at-speed stack test. The 2026-07-29 Q2 print is the binary that decides whether this is a pullback to buy or the top of the parabola.

Bullish and bearish views on FormFactor, Inc.

The model's bull view on FormFactor, Inc. (FORM), in brief: Q2 2026 guide (given on the 2026-04-29 Q1 call): revenue $240M ±$5M, non-GAAP gross margin 49.5% ±150bps, opex $65M ±$2M, non-GAAP EPS $0.61 ±$0.04 a record-revenue guide, no deceleration signaled. The bear view: The June 30 ATH looks like a blow-off: a Russell-inclusion-day high immediately followed by a 24% peak-to-trough drop ($160.27 → ~$121) is distribution, not accumulation. Both cases follow in full.

Bull Case

• Q2 2026 guide (given on the 2026-04-29 Q1 call): revenue $240M ±$5M, non-GAAP gross margin 49.5% ±150bps, opex $65M ±$2M, non-GAAP EPS $0.61 ±$0.04 a record-revenue guide, no deceleration signaled.
• HBM demand broadening past one buyer: management flagged HBM growing to another Q2 record driven by a SECOND customer adopting SmartMatrix for at-speed stack test the first dilution of the SK hynix single-customer dependency that has capped the multiple.
• HBM4 mass production begins in 2026 with layer count up and pin count doubling versus HBM3E; the higher test intensity wears probe cards out faster and shortens replacement cycles, scaling content per stack independent of wafer-start growth.
• SK hynix FY2026 capex guided KRW 35T+, up from KRW 29T in FY2025 the lead HBM customer is still spending into the ramp.
• 2026-06-29 Russell 1000 inclusion moved FORM from small-cap to large-cap index status (~$11B cap at the June peak), adding a passive holder base under the float.
• 7-analyst Buy consensus (2026-06-11); peers Aehr, Lam and Applied Materials bounced 2026-07-09, so the test/semicap complex found a bid after the early-July flush.

Bear Case

• The June 30 ATH looks like a blow-off: a Russell-inclusion-day high immediately followed by a 24% peak-to-trough drop ($160.27 → ~$121) is distribution, not accumulation. A name that round-trips a quarter of its value in a week is carrying weak hands.
• Valuation derates hard on any imperfection the multiple compressed the moment the 2026-05-11 Investor Day 2030 targets ($1.6B revenue / 55% GM / $5.00 non-GAAP EPS) met a rich tape.
• SK hynix is reportedly slowing its HBM4 ramp to redirect wafers into conventional DRAM, where shortage-driven margins now beat HBM's a direct threat to the "HBM4 replacement-cycle acceleration" leg if the lead customer pushes qualification right.
• Concentration remains: probe cards to the top-3 HBM makers means the growth vector is SK hynix / Micron / Samsung capex timing, while foundry/logic probe-card demand has been flat-to-down for several quarters.
• The next binary the 2026-07-29 print lands into a choppy, post-selloff tape with the stock already 24% off its high; a soft HBM4-timing comment re-rates it lower fast.

Setup & Price Structure

• Last ~$121.42 (2026-07-09), down ~17.7% on the week and ~24% from the 2026-06-30 ATH of $160.27; 52-week range $26.08–$160.27, still +112.8% YTD.
• The early-July flush found buyers near $121; the post-Investor-Day recovery base sits lower, around $109–110 the shelf the +22.8% June rally launched from. A weekly close under that base is the structural line.
• Overhead supply is now the problem the June breakout wasn't: the ~$144 street-average target and the $160 ATH are the reclaim levels, and a wall of trapped late-June buyers sits between $135 and $160.
• Consensus PT ~$144.29 (2026-06-11) sits ~19% above spot, but that average carries stale marks; the June cluster (Craig-Hallum Buy $175, B. Riley target $165, Evercore target $155) predates the selloff and has not been reaffirmed.
• Character read: this is no longer a stretched-above-MA chase but an unproven higher low. Knife-catching into the print is the trap here; a reclaim of the $135 area on volume would be the first sign the base is in.

Catalyst Calendar (next 30 days)

• 2026-07-29 FORM Q2 2026 earnings (confirmed). The numbers to clear: revenue $240M ±$5M, non-GAAP EPS $0.61 ±$0.04, and the HBM/DRAM probe-card records. Fresh sizing into the print is a coin-flip, not an edge.
• ~2026-07-24 (est.) SK hynix Q2 call: the key HBM4 read-through. The line to watch is whether the HBM4 ramp is being deferred in favor of conventional DRAM that comment moves FORM ahead of its own print.
• Ongoing Micron / Samsung HBM4 qualification cadence; any push-out headline is a direct probe-card demand signal.

What Would Change Our Mind

• A weekly close below $110 forfeits the post-Investor-Day recovery base and turns the July correction into a trend break.
• A Q2 revenue miss versus the $240M ±$5M guide, or non-GAAP EPS under ~$0.57, on the 2026-07-29 print.
• SK hynix confirming an HBM4 push-out or DRAM reallocation on its late-July call the single-customer dependency cutting the other way.
• Conversely, a reclaim of $135 on volume plus an in-line-or-better Q2 print would re-establish the accelerating leg and re-open the $160 ATH and the $175 top of the target range.

Correlation Notes

• Single-customer beta to SK hynix capex and HBM4 timing is the dominant factor; Micron and Samsung are secondary probe-card demand inputs.
• Trades with the semicap/test complex (Aehr, Camtek, Onto, Lam, Applied Materials) the 2026-07-02 flush and 2026-07-09 bounce were sector-wide moves rather than FORM-specific.
• High beta to the AI-memory / HBM narrative broadly (NVDA Rubin demand → HBM4 → probe-card content); a cooling in the AI-memory bid hits FORM harder than the multiple implies given how richly it is priced.