The CXDMO Capex Cycle: Decoding Q3 Spend and the Capacity-Expansion Inflection

Hidden in the Q3 reports is a harder signal than revenue - capital expenditure. Six CDMOs' 2026 capex totals over RMB 20 bn and is still climbing into 2027. This piece decodes what they are building, where the inflection lies, and the risks.

Revenue is the result; capex is the bet. Lining up the six CXDMOs' 2026 capital expenditure reveals a harder signal than revenue growth: this round of expansion totals over RMB 20 bn and is still climbing into 2027. What they are building, where the inflection lies, and what the risks are - that is the key to reading the next CXDMO cycle.

1. The capex water level: six names in lockstep, still rising through 2027

WuXi AppTec guides full-year capex to RMB 8.0-8.5 bn (raised from 8.5 bn ceiling) and pre-started its Changzhou greenfield; WuXi Biologics posted ~RMB 2.8 bn in H1, guides ~RMB 7.1 bn for the full year (including RMB 1.5 bn deferred from 2025), and ~RMB 8.0 bn in 2027; Pharmaron invested RMB 2.207 bn in 9M (+16.1%) and guides ~RMB 3.0 bn for the year; Asymchem guides ~RMB 2.1 bn (70%-80% to chemical/biologics-large-molecule capacity); Samsung Biologics launched a KRW 3 trillion rights offering (for the PolyPeptide acquisition and future capacity); Lonza's H1 capex was CHF 0.5 bn (15.7% of sales) with organic capex expected in the mid-to-high teens percentage.

Key call: this is not the 'expectation-only' expansion of the 2021 financing boom, but an order-backed one. WuXi AppTec's backlog of RMB 76.74 bn (+36.4%), WuXi Biologics' $25.1 bn backlog, Pharmaron's +30% new orders, and Asymchem's $1.39 bn backlog (+32%) provide a buffer for capacity fill - not reliance on distant stories.

2. Where the money goes: from 'traditional' to 'new modalities + local-for-local'

Direction one - new-modality capacity. Asymchem's peptide solid-phase synthesis capacity rises from 45,000L to 69,000L by year-end, oligonucleotides from 120mol to 180mol by 1H26, plus a RMB 2bn Tianjin large-chemistry base; Pharmaron's Shaoxing project includes 200 t/yr of intermediates and API (peptides, oligonucleotides) with a peptide GMP workshop built in 2026; WuXi XDC (WuXi Biologics) expands XDC capacity in China. This directly maps to T3's new-modality realisation - capacity is systematically tilting toward TIDES/ADC/CGT.

Direction two - overseas localisation (geopolitical hedge). WuXi Biologics pushes its Singapore integrated CRDMO center and US plant; Samsung Biologics' March-acquired Rockville site is already contributing; Lonza's Vacaville (acquired from Roche in 2024, its largest Integrated Biologics base) and Stein ADC commercial capacity expand. Under the Section 232 framework (see T4), US local capacity is shifting from a 'cost burden' to an 'access asset.'

Direction three - steady traditional small-molecule/antibody expansion. WuXi AppTec's Changzhou and WuXi Biologics' global planned capacity (>580 kL across CN/IE/US/DE/SG post-2026) hold steady investment. The marginal direction of incremental capex is clearly the twin threads of 'new modalities + overseas localisation.'

3. The inflection: late-2026 to 2027 is the concentration-release window

Asymchem's peptide capacity reaches 69,000L by year-end; at the industry level, Porton's Chongqing 46 t/yr and Cheerland's 1,500 kg/yr of ADC/peptide capacity release in late-2026 to 2027. Fill rate is the watershed: if orders ramp faster than capacity, scale effects release and per-unit depreciation thins; if fill lags, depreciation erodes margins.

Positive signal: new orders broadly surging (Pharmaron +30%, Asymchem double-digit, WuXi AppTec backlog +36.4%) points to comfortable fill. Negative signal: capex-to-revenue is elevated (Asymchem ~39%, WuXi Biologics H1 ~30%, Pharmaron ~18%, WuXi AppTec ~18%), and some long-cycle projects only come online after 2028 (Pharmaron Shaoxing phase I 2028 / phase II 2030; Hangzhou 2029), contributing little near-term revenue while already generating depreciation.

4. Risks: depreciation, the geopolitical double-edged sword, cash conversion

(1) Depreciation and utilisation: if fill underperforms, depreciation pressure appears, especially for post-2028 long-cycle projects. (2) Geopolitical double-edge: overseas capacity hedges Section 232 access risk but US labor and compliance premiums raise cost; meanwhile, Chinese stock capacity serving US patent-drug orders faces the disentanglement risk described in T4. (3) Diverging cash conversion: WuXi AppTec's adjusted FCF of RMB 14.5-15.5 bn comfortably covers capex; WuXi Biologics' free cash flow is positive but RMB 7.1 bn capex needs balancing; smaller players with high capex share and weak cash cover are more sensitive to order cadence.

The biggest difference from 2021 is that this round of capex stands behind real orders, not a financing pitch deck. The inflection leans positive, but the post-2028 long-cycle projects are the real touchstone.

5. Conclusion

This is an 'order-backed' capex round with a positive-leaning inflection. New modalities plus overseas localisation are the main threads; late-2026 to 2027 is the release window. Risks concentrate in long-cycle project fill and geopolitical hedge cost.

For investors: watch the 'scissors gap' between capex-to-revenue and capacity utilisation - its convergence confirms the inflection. For the industry: in the capacity arms race, players that win new-modality orders and hold overseas localisation will lead the next cycle; those with capacity but no orders or geopolitical elasticity will taste depreciation first.

Series closer

This is the closing piece (T5) of the CXDMO 2026 Q3 in-depth series. The full set: the Q3 panorama (T1), upstream equipment & consumables localization (T2), new-modality TIDES/ADC/CGT realisation (T3), global vs China CDMO and the 232 tariff (T4), and the capex and capacity-expansion cycle (T5).