摩根士丹利称 KLA 财报近端偏弱但 2027 年仍有工艺控制增长支撑

KLA Corp: Mixed Bag (again)
What happened? KLA’s SepQ guide came in a touch below MSe (KLA Corp: Long-term positive, near-term cautious (9 Jul 2026)). On the outlook the company revised its 2026 WFE forecast, including AP, from $140bn+ to low-$150bn (25%+), and talked toward ensuring that KLA themselves can prepare to supply and support the more bullish 2027 WFE views out there ($190bn+). 2026 advanced packaging revenue was revised up from $1bn to $1.1bn (+70% y/y), with growth rate 2x the market (KLA’s estimate).
Mixed bag (again). Similar to last quarter, we think this print offered something for both bulls and bears, but with the negatives concentrated in the near term, and vice versa. The bull case (primarily CY27): KLA’s confidence on outgrowing WFE in 2027 is based on backlog, as the company spoke of $12.5bn in RPO (vs FY25-end $7.86bn). We feel more confident in our thesis that process control intensity will improve in 2027 led by 1D and 1.4nm and were also encouraged by the company’s commentary around capturing share in advanced packaging on the back of hybrid bonding in logic and SoIC. The bear case (primarily CY26): Near-term numbers for KLA were uninspiring as the company wasn’t able to beat and raise vs intra-quarter commentary. Given KLA’s lead-times (12-months+), numbers in-line with intra-quarter commentary aren’t a surprise, but earnings does serve as a reminder that mix has been a significant headwind for process control intensity in 2026 and KLA is a share donor this year.
Where does this leave us? KLA’s earnings have been underwhelming vs etch/dep peers who have continued to show significant near-term upside. While there is a longer term debate around whether AI is a tailwind or not for process control (more N-1/2 wafers vs larger die/broader customer base), we are strong believers that process control intensity will be a tailwind for KLA in 2027. We stand by our analysis of process control intensity at TSMC by node, as well as industry DRAM wafers, which suggest that KLA’s leading-edge logic revenue at TSMC should accelerate in 2027, while DRAM growth rates should sustain despite a deceleration in DRAM WFE growth rates. These tailwinds should complement the broadening of leading-edge spend across Intel, Rapidus, and TeraFab, supporting our view that KLA will outgrow WFE in 2027, with growth of 36% versus WFE at 31%.
Changes to our estimates. We fine tune our CY27 forecast from $20.2bn/$6.12 to $20.4bn/$6.23 as we marginally revise up our system growth forecast from 34.7% to 35.7%. Our CY28 forecast of $23.4bn/$7.22 is largely unchanged.
Thoughts on the stock: We understand the bear thesis around KLA (2026 underperformance) and we agree that KLA will under grow WFE this year. We remain OW KLA; however, we’re anticipating the quarter when KLA materially beats street numbers and CY27/28 estimates come up. In the mean time, we nudge down our target multiple for KLA from 38x to 36x given derating across the cohort, which brings down our PT from $274 to $253. We shifted our valuation base year to 2028 for most US SPE names in Semiconductor Capital Equipment: Rolling over PTs to 2028 (6 Jul 2026). In the near-term we favor other OWs (AEIS, MKSI) given the magnitude of CY27/28 estimate revisions.