A 5× peer premium prices the FY26 margin snap-back as a near-certainty.

• The premium. Msscorps trades at 50× trailing EV/EBITDA, 20× revenue, and 13× book against direct peer MA-Tek at 11×, 4×, and 4.5× — same Taiwan oligopoly, same down-cycle, MA-Tek delivered double the FY24 operating margin (13.9% vs 6.85%).
• The bet behind it. Capex spiked from 34% of revenue in FY22 to 85% in FY24 (NT$1.67B); management guides FY26 capex back to ~17%. Gross margin halved from 40% to ~24% under the depreciation load, and is supposed to mechanically rebuild as the FY24-vintage fleet fills.
• The decision date. Q2 FY26 earnings in early August is the first quarter where 1H operating margin tests the rebuild. Bull thesis needs OM ≥ 10% on revenue ≥ NT$600M; OM below 5% is the bear trigger for peer-multiple compression.

The recovery curve cracked again in 1Q26 — Q2 in August is the print that resolves it.

• The recovery. Quarterly gross margin rebuilt 13.2% → 25.7% → 25.3% → 28.6% across FY25 on monotonically rising revenue (NT$465M → NT$592M); operating margin clawed from −10.5% to +5.0%.
• The reversion. 1Q26 broke the curve — revenue a record NT$579M (+24.5% YoY) but gross margin back to ~13% and EPS −NT$0.61, the second consecutive quarterly net loss after Q4 FY25 (Q3 FY25 was the only profitable quarter in the last five). The stock fell from NT$975 (5 May close) to NT$746 intraday on 13 May — a six-session drawdown.
• What August settles. A Q2 print at GM ≥ 25% and OM ≥ 10% reads 1Q26 as an anomaly; another GM-13 print sustains the bear's argument that better-executing MA-Tek delivers the same playbook at one-fifth the multiple.

Cash flow held while GAAP earnings fell off a cliff — the cliff was depreciation; the watch item is the next financing event.

Operating cash flow grew NT$663M → NT$743M across FY22–FY25 even as net income swung from a NT$288M profit to a NT$37M loss — the GAAP collapse is depreciation, with a clean forensic scorecard (accruals ratio −12.4%, CFO/NI 3.28×, zero red flags). But cumulative FY24–25 capex of NT$2.85B forced a NT$471M convertible bond, a 78% dividend cut, and pushed the current ratio from 3.27 to 1.09. If FY26 capex actually drops to NT$500M and quarterly margins hold, FCF flips positive and FY27 EBITDA roughly doubles. If utilisation slips, the next financing event is plausibly an equity raise into a stock the market has bid to 13× book.

Three things the consensus quote on this name misreads.

• Wrong-cycle peer compare. MA-Tek absorbed its capex wave in FY21–22 and has been harvesting since (capex intensity ~30%, OM band 11.5–18.7%); Msscorps' capex intensity ran 34% → 55% → 85% across FY22–24 and only just turned. The bear's 5× peer-multiple anchor compares a harvester to a builder — the cleaner test is Msscorps FY27 normalised against MA-Tek FY24, decided in two prints.
• MSS HG is equipment, not a story slide. The CPO inspection platform launching end-2026 carries disclosed pricing of NT$40–100M per unit and is backed by an offensive NT$200M patent suit (Enli, March 2026) — the legal posture of equipment IP, not service IP. Even ten units in FY27 would add NT$400M–1B at semiconductor-equipment-grade margins; no analyst has put that line in a model.
• The "3.7× sole broker target" is consensus by vacuum. Masterlink's NT$204 target was set on 2 January 2026 — before the MSS HG unveil, the Enli suit, the 4Q25 GM print, and the named-customer disclosure. The right description is "no consensus exists," not "consensus is far below the tape."