Management is betting three times the company's capital base on a packaging node that misses the current HBM cycle.
Thesis: Powertech is tripling its capital base to build capacity for overflow AI demand it cannot currently serve. They are skipping the HBM3E cycle to target 2027 production. By then, TSMC will have scaled CoWoS capacity to meet demand. PTI is building a bridge to a market that will be oversupplied by the time they arrive. The dividend commitment is a trap designed to keep investors through a massive dilution of return on invested capital.
Verdict: SHORT — Conviction: HIGH
Catalyst: A 2026 guidance cut when the first tier customer brings packaging in-house or TSMC CoWoS capacity expansion exceeds expectations.
Key Risk: Shortage of ABF substrates persisting through 2027 which would maintain artificial scarcity for second-tier packaging providers.
The Tell: The Chairman's blunt admission that we lost the CoWoS-S business. This confirms PTI is not a design-win partner but an overflow assembly house catching the scraps TSMC cannot process.
Friction Level: HIGH_FRICTION — The fundamental disagreement centers on whether massive capex for Fanout PLP is a strategic growth engine or a defensive survival play to capture TSMC overflow demand. One side sees execution discipline. The other sees a stranded asset risk as the company skips the current HBM3E upgrade cycle entirely.
Report not found
The report data is no longer available. Please return to the archive.