Financial Shenanigans

Figures converted from TWD at historical FX rates — see data/company.json.fx_rates for the rate table. Ratios, margins, and multiples are unitless and unchanged.

The Forensic Verdict

Forensic Risk Score: 48 / 100 — Elevated. Nextronics is not a shenanigans case in the classic sense — no restatement, no auditor change, five consecutive unqualified PwC opinions, and two-year cumulative operating cash flow that exceeds net income by 79%. The risk grade is "Elevated" because the breeding ground is more concerning than the numbers: an independent director who was a signing audit partner on the same engagement four years earlier, two related-party shareholders sitting on the board (one of them also a top-three customer disclosed as "non-related"), and a missing FY2024 annual report six weeks past the customary publication window. Layered on top, FY2023 working-capital quality deteriorated sharply (DSO +18%, DIO +27% in a down-revenue year) and FY2022 net income was 43% flattered by FX gains. The one data point that would most change the grade: filing of the FY2024 annual report with a transparent related-party note that quantifies sales/purchases with Hongyi Precision, Sinbon Electronics, and the Singapore holding entity.

Shenanigans scorecard

Breeding Ground

The board has a quiet auditor-independence problem and concentrated promoter control that earnings quality alone cannot price. Three signals stand out and they compound rather than cancel.

The auditor-independence signal is the single most actionable forensic flag in this report. Hsueh Shou-Hung (薛守宏) is named as a signing PwC partner on the FY2019 (108年) Nextronics audit in the company's own five-year auditor-history table (mda.txt p.84). He returned to the board as an independent director on 2023-06-14, which is the board chair of the audit committee role for a small TPEx issuer. Under both US (SEC Rule 2-01) and IFRS-aligned IESBA Code norms, a former audit partner should observe a multi-year cooling-off period before joining the audited entity's board; the gap here is roughly four years between rotation off the audit (2020) and board appointment (2023), which sits at the edge of the cooling-off norm and creates an apparent independence shadow. This does not, by itself, imply earnings manipulation — but it is a textbook breeding-ground condition that should be priced.

The promoter-control structure is the other persistent risk. Hongyi Precision (鴻乙精密) is the Hsu-Chen family vehicle: chairman Hsu Chi-Lin holds 36.82% and CEO Chen Yen-Cheng holds 41.50% of Hongyi, which itself owns 5.54% of Nextronics and holds a board seat through VP Lai Chi-Hsien. Combined with direct holdings, the Hsu-Chen camp controls roughly 12% of votes plus the chairman and CEO seats — a structure that minimizes external accountability without crossing thresholds that trigger more aggressive related-party disclosure.

Earnings Quality

Reported earnings are durable in cash terms but have been periodically goosed by FX gains, and FY2023 working capital quietly bloated in a down-revenue year. The forensic test that matters is the gap between operating performance and non-operating items.

In FY2022, $1.47M of the $1.52M non-operating income line was FX gains, equal to 41% of operating income and 29% of pretax profit. That collapsed to $0.31M in FY2023 and $0.17M in Q1-2024 — meaning roughly $1.15M of FY2022's $4.26M net income (27%) was a transitory FX windfall that will not recur at that scale. Any model that takes FY2022 EPS of NT$3.80 as a baseline for normalized earnings is implicitly capitalizing FX volatility as core earnings power.

The working-capital story is where the FY2023 income statement starts to look better than the underlying business. Revenue fell 11% but receivable days extended 18% and inventory days extended 27% — both moving against the direction the business was supposedly slowing.

DSO at 105 days in FY2023 is back at the FY2019 baseline of 107 days — not a unique high, but a meaningful reversal of the FY2021 efficiency gain (86d). The combination — DSO up, DIO up, gross margin held flat at 38.9% — is consistent with one of two stories: (a) genuine end-of-cycle channel buildup with no shenanigans, or (b) revenue pulled forward at quarter ends through extended payment terms with inventory shipped to distributor warehouses. The Q1-2024 partial reversion (DSO back to 97d) is mild evidence that (a) is closer to truth, but the full FY2024 receivable and inventory disclosure will be the cleanest test.

Gross margin expanded 160bps in FY2023 (37.3% → 38.9%) while revenue fell 11%. That is unusual; most contract manufacturers see margin compression on volume declines because of fixed manufacturing cost absorption. The likely explanation is mix shift — the connector unit volume fell 35% (23.5M → 15.2M units) while value fell only 10%, meaning ASP rose 40% as the company sold richer high-speed I/O parts. That is plausible and reads as economically real, but the margin elasticity will be tested again in FY2024 if revenue snaps back: real mix shift sticks; reserve-release inflation does not.

Cash Flow Quality

Cash conversion is healthier than a quick read suggests — but the dividend is currently being paid out of the balance sheet, not the year's free cash flow. Operating cash flow has tracked above net income for both years where it is disclosed, and there is no evidence of factoring, receivable sales, or supplier finance.

Two-year cumulative numbers tell the most honest story:

• Net income: $7.17M
• Operating cash flow: $12.85M (1.79x NI)
• Capex: $4.03M
• Free cash flow: $8.82M (1.23x NI)
• Dividends paid: $4.26M (48% of FCF)

Those numbers are clean for an industrial-component manufacturer. The deterioration story sits inside that average: FY2023 alone produced FCF of $2.20M against a dividend payout of $3.08M, with the gap funded by drawing the cash balance. Cash flow adequacy ratio fell from 88.8% (FY2023, 5-yr basis) toward dilution as the company also issued shares (capital surplus +17% YoY) and watched non-current liabilities go to zero as convertibles reclassified into current.

The honest reading: FY2023 OCF ($4.68M) is not propped up by working-capital releases. Receivables grew, inventory grew, and OCF was supported by $2.20M of depreciation and amortization plus the underlying $2.91M of earnings, with a slight working-capital drag. Compare that to FY2022, where $1.87M of working-capital favorability boosted reported OCF by about 30%. So if anything, FY2022 OCF flattered the run-rate and FY2023 OCF is the more representative number, which is the opposite of what a quick "OCF declined 43%" headline would suggest.

The one cash-flow signal that does warrant attention: the external Simply Wall St risk feed flags a "high level of non-cash earnings" and a dividend payout ratio of 98-99% with poor cash coverage. That aligns with the FY2023 dividend exceeding FCF picture and means the FY2024 dividend (if maintained at FY23 levels) will require the cash balance to fund the gap, especially with the planned Xizhi plant purchase ahead.

Metric Hygiene

This is the cleanest part of the file. Management does not publish a non-GAAP earnings number, an adjusted EBITDA, an adjusted EPS, an "organic growth" metric, or a same-store-style comparable. There is no metric to manipulate because there is no metric to highlight beyond statutory IFRS.

The absence of non-GAAP is genuinely positive — the company does not have a track record of explaining away "one-time" charges, recurring restructuring, or stock-comp adjustments. The headline that "AI revenue was 30% of FY2025 sales" (per the CommonWealth March 2026 article) and "gross margin of 40%" is a marketing line that has not been formalized into a periodically disclosed KPI, and that is the right answer.

The disclosure gaps that matter sit in three places: (a) no quarterly related-party transaction breakdown, (b) no segment margin reporting, and (c) no backlog disclosure. None of these are unusual for a small TPEx issuer, but they collectively limit independent verification.

What to Underwrite Next

The accounting risk here is a position-sizing limiter, not a thesis breaker. The earnings are real, the cash is mostly real, and the auditor has signed clean for five years. But the governance breeding-ground means a reasonable margin of safety should be required before sizing up.

The five items to track before the next print:

1. FY2024 annual report publication. Track whether (a) it lands on the IR site, (b) it is unqualified, (c) the related-party note quantifies Hongyi Precision, Hongyi-affiliated entities, and Sinbon transactions, and (d) the strategic-investment line in "other assets" is itemized. This is the single highest-value test in the next 90 days.

2. FY2024 dividend coverage. Compute FY2024 FCF (CFO − capex) net of the Xizhi building purchase. If dividend exceeds FCF for the second consecutive year while the company is also expanding plant footprint, watch for either a dividend cut or fresh equity issuance — both would be confirming signals.

3. DSO and DIO in FY2024 and Q1-2025. If revenue accelerated 30%+ as 9M-2025 data suggests, DSO and DIO should compress from the FY2023 readings. A persistent 100+ day DSO during an upturn would mean the FY2023 bloat was not cyclical — it was customer-mix or revenue-recognition aggression.

4. Sinbon Electronics relationship. If FY2024 still discloses Sinbon as a top customer AND as a board-represented 8%+ shareholder, AND IFRS 24 disclosure shows no related-party sales, the gap between substance and form widens. Demand triangulation against Sinbon's own RP disclosures (Sinbon is a larger listed connector group).

5. Convertible bond cancellation rationale. The 2026-05-14 board reversal of the AGM-approved private placement deserves a published explanation — either market timing (defensible) or anchor-investor withdrawal (much more concerning).

What would upgrade the grade toward "Watch" (sub-40):

• FY2024 AR published with quantified RP notes
• Sinbon disclosed as related party
• Hsueh Shou-Hung either steps off the board or assumes a non-audit-committee role
• DSO back below 95 days during the FY2025 upturn

What would downgrade the grade toward "High" (above 60):

• Auditor change in FY2025 cycle (especially mid-year)
• FY2024 AR delayed past September 2026
• Discovery of undisclosed related-party sales above 5% of revenue
• Material weakness disclosure
• Inventory write-down concentrated in a single quarter that masks a margin reset

The bottom line for the file: Nextronics earnings are not fictional, but they are partially flattered by FX in good years and partially propped by inventory/receivable bloat in bad years, and the governance superstructure does not provide the institutional checks that would normally make those small distortions immaterial. An investor underwriting the AI-connector growth story should haircut the FY2025 reported margin by 50-100bps for the residual accounting risk, require visible related-party disclosure in the next AR, and size the position such that an unexpected FY2024 AR delay or inventory write-down is survivable. This is an Elevated-risk forensic profile that does not contradict the bull case; it just argues for buying a smaller piece.