Financials

Financials — What the Numbers Say

Figures in HK$ unless labelled otherwise. ASMPT also publishes USD figures because most operating subsidiaries' functional currency is USD.

1. Financials in One Page

ASMPT is a cyclical, mid-scale back-end semiconductor equipment maker (HK$13.7B FY2025 revenue, ~US$1.77B) coming off its second-worst trough of the last decade. Gross margin held at 37.8% but operating leverage worked in reverse: net margin fell to 6.6% (FY2025) vs 14.4% at the FY2021 peak. Balance sheet: HK$3.3B net cash, HK$5.7B liquidity, 3.6x current ratio. FCF turned negative in FY2025 (-HK$139M) as the order book rebuilt working capital; management is guiding to a sharp AI-driven recovery (Q1 2026 net profit +204% YoY, Q2 2026 revenue guidance +37% YoY at midpoint). At HK$176/share the HK$72B market cap equals ~80x trough FY2025 earnings and ~5.2x trough revenue — a recovery multiple. The metric that matters most: gross margin progression as TCB volumes scale, because operating leverage on 14% R&D intensity decides whether the recovery echoes the FY2021 peak.

2. Revenue, Margins, and Earnings Power

Revenue is cyclical, gross margin holds in a 37–40% band, and operating leverage on ~HK$4.6B of fixed SG&A and R&D swings net margins from low single digits at trough to mid-teens at peak. The ten-year history shows two cycles: the 2016–2018 memory + smartphone wave, the 2020–2021 COVID surge, then a two-year trough through FY2024 from mature-node SMT weakness and a pause in advanced-packaging investment ahead of HBM.

Gross margins are resilient (40% in FY2024 at decade-low revenue; 37.8% in FY2025) — product-mix discipline, and TCB/high-end wire bonding customers paying for capability not capacity. The drop from 14.4% net margin (FY2021) to 2.8% (FY2024) is operating-leverage driven, not pricing. Research spend rose through the trough — HK$1.92B in FY2024 vs HK$1.65B at the FY2021 peak — as the company doubled down on TCB for HBM and co-packaged optics. Management is treating the trough as an investment year, defending technology share in advanced packaging.

Q2 is ASMPT's only standalone disclosed quarter (HKEX requires annual + interim only). Q2 FY2025 revenue was 35% below Q2 FY2021 but already inflecting — H1 FY2025 bookings +12.4% YoY and Q1 FY2026 guidance points to a step-change.

3. Cash Flow and Earnings Quality

FY2025 conversion is uncomfortable: HK$902M of reported profit produced HK$243M of operating cash flow and -HK$139M of FCF after capex.

Three factors explain the gap:

1. Working-capital build of -HK$491M. Inventory +HK$433M and receivables +HK$1.05B as the company stocked for the HBM/TCB ramp; payables and advance customer payments offset +HK$989M. Classic equipment-maker recovery cost — financing customers before they pay.
2. Restructuring cash outflow of -HK$349M — utilisation of FY2023–FY2024 provisions. The bill for the trough, not a recurring cost.
3. One-off non-cash gain of +HK$1.11B from JV disposal sits in net income but not operating cash. Strip it out and underlying earnings were ~HK$0M.

Verdict: FY2025 net income overstates recurring cash power by the HK$1.11B JV gain; the negative FCF understates cyclical cash power because of one-off restructuring outflows and the deliberate WC build. Underlying cash earnings sit near breakeven with capex tracking R&D. Cash conversion should normalise if H2 FY2026 revenue runs at the guided pace.

4. Balance Sheet and Financial Resilience

Net cash of HK$3.28B (~24% of trough revenue), no covenant pressure, and a 3.6x current ratio give ASMPT optionality to invest through the cycle. R&D intensity rose and the trough did not force a dividend cut to zero (HK$241M paid in FY2025).

Net cash grew through the trough (HK$2.1B FY2021 to HK$3.3B FY2025) as FY2024 WC release freed HK$1.4B and the JV disposal added HK$862M. HK$2.4B of bank borrowings is mostly long-term; HK$2.0B of lease liabilities is right-of-use property accounting, not financial leverage.

One yellow flag in FY2025: HK$240M goodwill + HK$45M intangibles + HK$24M PPE impairments, all tied to the NEXX wafer-level packaging business being divested to Applied Materials. A strategic exit, not a balance-sheet alarm.

5. Returns, Reinvestment, and Capital Allocation

Reinvestment economics swing between excellent (FY2017 ROE 28%, FY2021 ROE 22%) and poor (FY2023 ROE 4.5%, FY2024 ROA 1.5%). Ten-year average ROE sits in the 12–14% range — respectable for an equipment maker but below true compounders. The question is whether 14% R&D intensity generates excess returns.

Capital allocation: hold technology investment (R&D HK$1.92B/yr through the trough); maintain dividends (HK$468M FY2024 to HK$242M FY2025, roughly halved as cash conversion weakened); no buybacks (float constrained by ASM International's 25% stake); one disciplined divestiture (NEXX → Applied Materials at HK$862M cash). Share count flat at 416M; SBC at ~1% of revenue.

Management is reinvesting against future expected returns. If TCB share holds against BESI and Hanmi through the AI-packaging cycle, ROE can return to the high-teens; if share slips, R&D capital looks like deadweight.

6. Segment and Unit Economics

ASMPT discloses two segments; the mix shift matters more than the aggregate line.

FY2025 mix shifted decisively: SEMI passed SMT to 53.7% of revenue (vs 44.2% in FY2022). SEMI grew +21.8% YoY on Advanced Packaging / TCB / HBM; SMT was flat-to-down on weak consumer/industrial demand. SEMI margins (7.45%) overtook SMT (6.36%) for the first time in years — TCB is higher-ASP and richer-mix than mature wire bonders. R&D intensity inside SEMI is 15.7% vs SMT 12.1%.

Note on assets: ASMPT does not split balance-sheet items by segment, only revenue and profit. Segment ROIC is not directly computable from filings.

7. Valuation and Market Expectations

At HK$71.9B market cap and HK$176/share, the market is paying for recovery, not trough fundamentals.

Share price has decoupled from current earnings: net income fell -89% from FY2021 to FY2024 while the share price rose through FY2025 toward HK$176. Classic equipment-recovery setup — the market is pricing FY2026/FY2027 consensus (revenue +30%, EPS +258% to HK$4.01), implying forward P/E ~44x.

Peer context is harsh: BESI trades at 80x FY2025 P/E on a 29.3% operating margin and 28.9% ROIC; ASMPT trades at 80x P/E on a 5–7% operating margin and 3.6% ROA. Same multiple, roughly one-third the profitability. Bull case: mean-reversion in operating leverage. Bear case: BESI's higher steady-state margin reflects a genuinely better positioned advanced-packaging business.

At HK$176, the stock sits between the base and bull ranges — roughly 80% of the way to the bull case if those scenarios hold. On these inputs, upside and downside are no longer symmetric.

8. Peer Financial Comparison

Relevant peers: BESI (most direct TCB), KLIC (wire-bonder leader and TCB challenger), Hanmi (HBM TCB at SK hynix). Disco and TOWA are adjacent.

Two-sided comparison: ASMPT is the largest pure back-end packaging name by sales (HK$13.7B / ~US$1.77B, more than twice BESI). On profitability, BESI is 2–3x more profitable per dollar (65% gross / 22% net / 27% FCF margin vs ASMPT's 38% / 7% / -1% in trough). On valuation, ASMPT, BESI, and Hanmi all sit in the 60–80x P/E band — the market is pricing the AI/HBM TCB ramp across the cohort, not any quality discount. Should ASMPT's scale and SMT diversification matter more than BESI's purer mix and higher margins? Today's price says "close to equivalent" — that is the bet.

9. What to Watch in the Financials

The financials confirm: balance sheet supports through-cycle investment; gross-margin discipline survived the trough; SEMI mix is shifting decisively toward TCB/HBM; Q1 FY2026 has inflected. They contradict the simplest bull case: ROIC and FCF margins sit visibly below BESI and KLIC, and FY2025 net income is flattered by HK$1.1B of JV-disposal gains. Whether Q2 FY2026 gross margin prints above 40% is the next test, because the valuation premium rests on operating leverage flowing through.

Gross margin in Q2 FY2026 H1 results is the key tell — above 40% supports the leverage thesis; below 38% suggests AI-packaging revenue is more competitive than the price implies.