In China, solar manufacturing in the first half of 2026 was a price war with almost no winners. Daqo sold polysilicon at ¥30.63 a kilogram against a cash cost of ¥34.75 (2026 interim report, p. 10), and the four largest module makers together made about 0.4% gross margin, as our earlier article on where China's solar chain still makes money showed. The response from the industry's leaders is to move production closer to the customers who still pay. The catch, visible in the same interim reports, is that trade barriers are moving with them, country by country.
The plants going up abroad
- JA Solar runs a base in Vietnam and is bringing a 6 GW cell and 3 GW module project in Oman into production in phases with partners (2026 interim report, p. 12). Overseas sales were 70.53% of its revenue in the half (p. 16).
- Jinko Solar has factories in Southeast Asia and the US and is setting up a joint venture plant in Saudi Arabia. Overseas business took more than 75% of its revenue (2026 interim report, p. 19).
- Xinyi Solar started its first solar glass line in Indonesia in January 2026, with a second due within the year and a second phase of 2,300 tonnes a day being planned, while some idle capacity in China has been shut (2026 interim results, p. 54). Overseas markets took 38.8% of its solar glass sales, up from 31.6% a year earlier (p. 46).
- Flat Glass has a subsidiary in Vietnam and a plant in Indonesia, with an investment of US$290m, in preparation (2026 interim report, p. 13).
- Drinda, the cell specialist, holds a stake in a cell plant in Turkey that has started production and is ramping up, aiming at local production for local delivery (2026 interim report, p. 12).
- TCL Zhonghuan is discussing joint ventures abroad to make back-contact modules locally with regional partners, after its overseas wafer shipments rose 3.7 times and it won new customers in India and Turkey (2026 interim report, p. 12).
The pattern is consistent: glass and cells, the stages that ship in bulk, go to Indonesia, Turkey and the Middle East; modules follow through joint ventures that give the local partner a stake. Nobody is building a replica of the whole Chinese chain abroad. Each plant covers one stage, in one market, with a local partner.
Why they are going: buyers abroad still pay
The filings give a simple reason. In the first half, Jinko earned about 5.6% gross margin abroad against about -3.3% in China, and JA Solar earned 3.02% abroad against -2.86% at home (our calculations from Jinko p. 165 and JA p. 16). Sungrow took 73.40% of its revenue from outside mainland China (Sungrow p. 35). Xinyi says its overseas plants help it avoid trade barriers and tariffs, cut freight costs and earn higher prices (2026 interim results, p. 42).
Export data shows the same pull, with a warning attached. Citing China Photovoltaic Industry Association figures, Jinko says China's solar product exports rose 24.3% in value to about US$17.18bn in the first half, but almost all of the increase came from a rush before China ended the export VAT rebate on solar products on 1 April 2026; in May, wafer, cell and module export volumes fell 33%, 35% and 26.7% respectively from April (Jinko p. 9). Cells were the exception over the half as a whole, with export volumes up 36.8%, as other countries build module assembly faster than cell capacity (Drinda p. 11). That is the gap Drinda's Turkey plant and JA's Oman cells are aimed at: sell the cell where the module is assembled, rather than shipping the finished module from China.
The barriers are moving too
The Southeast Asia route already shows how fast a refuge closes. The US set final duties in April 2025 on cells and modules from Cambodia, Malaysia, Thailand and Vietnam, with anti-dumping rates of 0% to 271.28% and countervailing rates of 14.64% to 3,403.96% (JA p. 30). In August 2025 the US Court of International Trade ruled that a 2022 tariff exemption for Southeast Asian solar products was unlawful, opening the way to retroactive duties on imports from April 2022 to June 2024; in February 2026 the US Justice Department withdrew its appeal against that ruling (JA p. 30).
The next round is already open. In February 2026 the US Commerce Department announced preliminary countervailing duties of 80.67% to 143.30% on crystalline silicon cells from India, Indonesia and Laos, followed in April 2026 by preliminary anti-dumping duties of 22.46% to 123.04% (JA p. 30). Indonesia is on that list at the same time as Xinyi and Flat Glass are building glass plants there. JA also notes rising local manufacturing requirements in India, Europe and the US, including India's approved list of models and manufacturers and its production-linked incentive scheme (JA p. 30). A plant abroad now buys market access that a shipment from China no longer has, but only for as long as that country stays off the duty list.
The cases that show the limits
Drinda's Oman project is the cautionary case. Its planned 5 GW cell plant in Oman has been held back by trade policy and Middle East uncertainty, and the company says it is looking for other investors and partners (Drinda p. 12). Its overseas revenue fell 32.71% in the half, to 40.20% of the total, even as China's cell exports rose 36.8% (Drinda p. 14). Being a cell specialist did not on its own capture the export boom, and an announced overseas plant is not yet a working one.
Jinko shows the other discipline: it says it cut shipments to low-price, loss-making markets and sent about 70% of its shipments to overseas markets with higher margins (Jinko p. 10). Moving abroad is being paired with choosing where not to sell, which is a change from an industry that spent a decade selling everywhere at whatever price cleared the factory gate.
What to watch
Our view: the overseas move is real, but it is a relocation of the price war's front line, not an escape from it. The figures to track are whether JA's Oman cells and modules reach full phased production, whether Xinyi's second Indonesia line and Flat Glass's Indonesia plant are completed on schedule, whether TCL Zhonghuan's back-contact joint ventures are signed, and whether the US duty cases extend to Oman, Turkey or Saudi Arabia next. If they do, the premium that justified building abroad narrows to the freight and tariff saving on a single stage, and the industry is back to competing on cost, just from more countries.
Follow the chain on our solar PV industry map, from polysilicon and solar wafers to solar cells, solar glass and solar modules.