The whole equity case rests on one variable — does the 9M-2025 operating-margin lift from 7.4% to 11.2% hold or fade?

• Gross margin is proven, operating margin is the swing. FY23 gross margin of 38.9% already beats Amphenol (36.9%) and TE Connectivity (35.2%) — pricing power exists. But 7.4% operating margin trails the broad-line peer band of 18–25% because a 10-country, 480-person footprint was built for a business 2–3× this revenue base.
• 9M-2025 cracked the door open. Revenue +30% YoY, net income +73%, operating margin lifting to 11.2% with a Q3 standalone of 10.5% — the incremental-margin signature of fixed-cost absorption, not a single-quarter mix accident.
• Q1 2026 already slipped 80bps. First print of the new fiscal year landed at 8.1% net margin on revenue +26% — inside the bear's reversion path, not above it. Q2 2026 around 2026-08-12 is the resolving event.

Six years of mix migration lifted gross margin from 32.7% to 40% — but FCF has not yet caught up with the dividend.

The mix migration from D-Sub legacy work into AI HSIO, medical, and aerospace lifted gross margin 720bps over five years and operating margin 13 points off the 2019 trough. The catch sits one line down: FY23 free cash flow of $2.2M failed to cover the $3.1M dividend, share count is up 20% via stock dividends from end-FY23 to 9M-2025, and at 60× trailing earnings the multiple has already paid for operating-leverage proof that the next four quarters now have to deliver.

The tape is paying Lotes / Amphenol multiples for Bizlink / FIT economics.

• The wrong peer band. Operating margin of 8–11% and revenue of roughly $40M put Nextronics squarely inside the Bizlink (8% op margin, 22× P/E) and FIT Hon Teng (7.3% op margin, ~55× LTM P/E) band. The 60× LTM multiple it actually trades at sits closer to Lotes (34% op margin, 31×) and Amphenol (25% op margin, 40×).
• Premium without the profitability. A ~50% P/E premium to Amphenol and ~70% premium to TE Connectivity, despite earning a third of their operating margin and less than 1% of their revenue. The premium is paid for growth optionality, not realized quality.
• Bizlink-band math implies $2.70. 22× a normalized FY26 EPS of $0.12 lands at $2.70 — well below the $6.40 spot — and the path between the two runs through one quarterly print and one annual-report filing, both inside the next 90 days.

Half the revenue, an undisclosed name, and four of six failure modes touch the same pocket.

• 48.2% of revenue from communications / cloud. That share was already half the mix in FY2023 — before the AI HSIO surge. After 9M-25's +30% growth the pocket is likely well over half of revenue, the largest single customer category sits inside it, and the customer is not named in any filing.
• FY2023 already showed the asymmetry. Revenue fell only 10.8% in the medical destock; operating income fell 24%. The same single-customer pocket now carries 9M-25's operating leverage — a hyperscaler-program cancellation in 2026 would force the symmetrical air-pocket on the other side.
• The overdue FY2024 AR is the file that resolves it. Six weeks past the customary TPEx publication window. A top-1 customer under 20% with broadening top-10 confirms the diversification path; 25%+ with AI-cluster dependency confirms the bear pocket and breaks the multiple.

The cleanest investor-friendly tell on file sits inside a cluster of unresolved governance flags.

The clean tell. On 2026-05-14 the board withdrew the AGM-authorised domestic convertible bond without issuing — unusually rare for a TPEx small-cap that had already cleared the dilution path. FY23 cash dividend tripled to $3.1M on declining earnings; cash sits at $20M against essentially zero interest-bearing debt; no SBC, no PSU, no option grants. Founders monetise via dividends and a 13% combined family stake.

The cluster. Independent director Hsueh Shou-Hung was the signing PwC partner on the FY2019 Nextronics audit and was appointed to the audit committee in 2023 — inside the SEC-style cooling-off norm. The FY2024 annual report is six weeks overdue. Sinbon Electronics holds 8.35% plus a board seat and was a 10% customer in FY2022, disclosed as non-related. FY23 dividend of $3.1M exceeded free cash flow of $2.2M.

Today. Each signal is small alone; the cluster is what raises the discount rate at 60× LTM. Forensic risk score sits at 48 / 100 — Elevated, not High — because the earnings are real, two-year cumulative operating cash flow exceeded net income by 79%, and PwC has signed clean for five years. The cluster argues for a smaller position, not a refusal to own.