China-HQ green FDI readiness / Method appendix
Back to the story

Battery case notes

Purpose. Per-cell observations for the five battery country-technology pairs in the narrative batch, organized by supply-chain position family [1].

Data sources. Per-cell capability and trade outcomes from data/processed/country_tech_quadrants.csv. Project lists from data/processed/china_outward_fdi.json. HS6 decomposition from data/processed/case_card_hs6_decomposition.csv. Manufacturing-only BACI cut by default; the stricter factory-stage cut excludes Processed Material and Raw Material.

Windows. Capability window 2015–2019. China-HQ FDI window 2020–2024. Trade pre-window 2017–2019; trade post-window 2022–2024.

1. Hungary / Battery

An operating Korean-built export platform attracted major China-HQ additions; the measured 2022–2024 export jump is mostly the pre-existing platform.

Hungary entered the FDI window with high battery PC (0.675) and broad RCA footprint near parity (0.921). Battery manufacturing exports were already $2.50B in 2017–2019 at RCA 1.13. The China-HQ FDI signal is large and disclosed: nine projects, $11.83B in announced capex, led by CATL Debrecen ($7.50B, commenced 2022), Huayou Cobalt ($1.50B, commenced 2023), EVE Energy Debrecen ($1.40B, commenced 2023), and Yunnan Energy New Material Debrecen ($825M, commenced 2020). The base also includes major non-China-HQ investors (Samsung SDI's Göd plant and SK On's Komárom and Iváncsa complexes) that are outside the panel by construction.

Manufacturing exports rose to $9.98B (+299%) and RCA rose to 2.98. HS 850760 (lithium-ion accumulators) accounts for 94.1% of the $7.48B manufacturing-basket increase. The timing test rules out a CATL-Debrecen causal attribution: Hungary already exported $8.29B in 2022, far above the 2017–2019 baseline, while CATL's own update placed first cell production in 2025.

The pattern is consistent with capability selection: pre-existing export capability, EU market access, government incentive packages, and an operating Korean-built base made Hungary credible as a host for major Chinese-headquartered additions.

PC pre 0.675 · FDI $11.83B / 9 projects · finished share 94.1% · factory RCA 1.36 → 3.81 · signal strong_growth

2. Indonesia / Battery

Manufacturing-basket RCA crossed 1.0, but the specialization is midstream nickel and cathode precursors; factory-stage RCA does not cross.

Indonesia entered with battery PC 0.028, below the 0.10 capability floor. Manufacturing exports were $2.35B at RCA 0.64. The China-HQ FDI signal is large: seven projects and $6.86B in disclosed capex, anchored by CATL Karawang ($6.00B, commenced 2022) and Hailiang Gresik ($850M, announced 2023), with smaller completed projects from SAIC/CATL, SAIC-GM-Wuling, Gotion, Greenway, and Phylion.

Manufacturing exports rose to $10.57B (+351%) and RCA crossed to 1.58. Under the stricter factory-stage cut, exports rise from $550M to $1.30B but RCA moves only 0.21 → 0.26. Finished-cell HS codes (850760 + 850780) explain just 2.2% of the $8.23B manufacturing-basket increase. The largest contributors are nickel mattes (HS 750110, $3.02B, 36.7%), nickel/manganese ores used in the CAM chain (HS 750120, $2.93B, 35.6%), alumina current collectors (HS 281820, $1.02B, 12.4%), electrolyte (HS 382499, $473M), and nickel sulfate (HS 283324, $260M).

The pattern is consistent with conversion of mineral endowment and processing capacity into midstream battery-material exports while China-HQ capital was being deployed. It is not consistent with finished-cell export entry.

PC pre 0.028 · FDI $6.86B / 7 projects · finished share 2.2% · factory RCA 0.21 → 0.26 · signal strong_growth

3. United States / Battery

Domestic-market platform: cells, electrolyte, and equipment all grow while factory-stage export RCA falls. The United States is the largest net finished-cell importer in the panel, so the RCA dip reflects home absorption and faster global export growth, not a failed buildout. Export RCA is the wrong yardstick for a domestic-market recipient.

The United States entered with low direct battery PC (0.029, below the 0.10 floor) but large manufacturing-basket exports already in place ($29.86B at RCA 1.06) and a stronger broad RCA footprint (0.803). The panel lists nine projects and $4.38B in disclosed capex: Gotion Manteno ($2.00B, commenced 2023), Yunnan Energy New Material Sidney ($916M, announced 2022), Canadian Solar Shelbyville ($712M, commenced 2024; Canada-HQ via the e-STORAGE subsidiary, see [2]), EVE Energy Marshall County ($250M), Hithium Mesquite ($200M), Capchem ($175M, announced 2023), and Tinci Houston ($102M). Under documented exclusions (Canadian Solar removed), disclosed capex drops to $3.66B; the cell stays above the $100m FDI-First threshold.

Manufacturing exports rose to $36.09B (+21%) but RCA fell to 0.86. No single HS type clears the 50% threshold: finished cells (HS 850760) contribute 31.0% of the $6.23B basket delta, electrolyte (HS 382499) 29.2%, and process equipment (HS 847989, 902780, 903190, 903180) most of the residual. Factory-stage RCA fell 0.25 points (1.19 → 0.94).

The pattern is consistent with US-bound capital oriented toward domestic-market capacity rather than export-platform formation, but the trade panel alone cannot distinguish domestic-market substitution from a faster-growing global denominator.

PC pre 0.029 · FDI $4.38B panel / $3.66B documented / 9 projects · finished share 31.0% · factory RCA 1.19 → 0.94 · signal flat

4. Poland / Battery

Finished-cell export entry without any China-HQ FDI in the panel; the export-growth channel does not require Chinese capital.

Poland entered with one of the strongest battery capability signals: PC 0.847, broad RCA footprint 0.806, manufacturing exports already $4.33B at RCA 0.95. The panel records zero China-HQ battery FDI projects in the 2020–2024 window. The operating base is tied to non-China-HQ investors (most prominently LG Energy Solution's Wrocław facility) that are filtered out by construction.

Manufacturing exports rose to $14.32B (+231%) and RCA crossed to 1.85. Finished-cell HS codes (850760 + 850780) explain 82.8% of the $9.99B basket increase; adding parts (HS 850790) lifts the share to 84.9%. HS 850760 alone contributes $8.27B (82.8%).

The pattern is consistent with finished-cell export entry under EU supply-chain integration and Korean industrial capital. Public copy should report Poland as carrying no China-HQ projects within this dataset and window, not as absent of all FDI.

PC pre 0.847 · FDI absent in panel · finished share 82.8% · factory RCA 0.98 → 2.25 · signal strong_export_entry

5. Mexico / Battery

Battery process equipment and components grew into a US-absorbed demand channel; an equipment / process-tools platform without China-HQ FDI in the panel.

Mexico entered with battery PC 0.014, far below the 0.10 floor, broad RCA footprint 0.430, and manufacturing exports of $2.71B at RCA 0.32. The panel records zero China-HQ battery FDI projects in the 2020–2024 window.

Manufacturing exports rose to $4.76B (+76%) but RCA barely moved (0.32 → 0.35). Finished-cell HS codes (850760 + 850780) explain only 12.9% of the $2.05B basket increase; cells plus parts (850790) reach 19.7%. The largest contributor is HS 847989 ($536M, 26.1%), battery process equipment covering module-pack assembly and cell conditioning. Other major contributors are battery casing (HS 761699), quality-control equipment (HS 903180), electrolyte (HS 382499), separators (HS 850790, 392020), and capacitance testing equipment (HS 903089). Destination decomposition shows USA + Canada absorbed 88.4% of the basket delta (USA alone 86.5%), and the USA + Canada delta share holds 65.6% to 98.8% across all eight top growth codes [3].

The pattern is consistent with Mexico's auto-supplier base translating into battery-supply-chain output absorbed primarily by US-based battery plants in the post-IRA window. The trade panel does not identify specific plants or buyers.

PC pre 0.014 · FDI absent in panel · finished share 12.9% · factory RCA 0.37 → 0.41 · signal moderate_growth

Cross-cell observations

  1. China-HQ FDI lands in different supply-chain positions across the five battery cells: finished cells (Hungary), midstream materials (Indonesia), and a mixed cells/electrolyte/equipment portfolio with relative-share decline (United States).
  1. Manufacturing-basket RCA crossing is not factory-stage RCA crossing. Indonesia crosses on the manufacturing-only cut while finished cells explain 2.2% of the increase and factory-stage RCA remains 0.21 → 0.26.
  1. Finished-cell export specialization can emerge without China-HQ FDI in the panel. Poland reaches finished-cell entry on Korean and EU industrial capital outside the China-HQ filter.
  1. Direct PC misses supply-chain adjacency. Mexico's growth concentrates in battery process equipment, components, separators, and electrolyte; the 0.10 PC floor classifies the cell as Bypassed despite a +76% manufacturing-basket increase absorbed primarily by USA + Canada.
  1. Greenfield ramp lag means the 2022–2024 trade window does not measure the largest 2020–2024 China-HQ projects at commercial output. Hungary's 2022 baseline of $8.29B predates CATL Debrecen first cell production; the United States and Indonesia carry similar late-cycle exposures.

References

Markdown source on GitHub