Standard Fiberglass Momentum Could Last Longer, with High-End AI Electronic Fabric Taking Over Growth in 2027

Overview

This report summarizes the key messages from China Jushi's business briefing, focusing on supply and demand for standard fiberglass and 7628 electronic fabric, the growth path for AI electronic fabric, the pace of capacity expansion, and valuation. Morgan Stanley's core view is that the pricing upcycle for existing products could continue through 2027—2028, while high-end AI electronic fabric will provide the next phase of earnings catalysts beginning in 2027.

Core views

The strength of the standard fiberglass cycle may persist longer than the market initially expected. Prices have more than doubled year-to-date, but customers are currently competing for supply rather than resisting price increases, so Morgan Stanley still expects room for further price increases in 2H26. Supply of 7628 electronic fabric is particularly tight: AI demand has significantly exceeded expectations, inventories across electronic fabric, copper-clad laminates, and PCBs are all extremely low, and China Jushi itself has only 3.9 days of inventory and can fulfill only about 80% of orders. The report therefore concludes that high prices have not yet eliminated the near-term supply-demand gap. The supply response is expected to remain slow in 2027—2028. Constraints include not only furnaces but also Toyota looms, downstream processing capacity, skilled workers, and customer certification; delivery lead times for some equipment have already extended to 2029. Domestic looms remain unsuitable for producing high-end or thin electronic fabric due to lower yields. These constraints mean that even if higher prices stimulate capacity expansion, effective new supply will be difficult to release quickly, supporting a longer upcycle. High-end AI electronic fabric is viewed as the next growth driver. Low-Dk G2, low-CTE, and Q-fabric products remain at an early stage, but China Jushi expects substantive progress in 2027. For these products, yield is more important than simply pursuing low costs, making equipment capabilities, accumulated process expertise, and certification barriers more important than for standard products. As AI servers and PCBs demand higher performance, an increasing proportion of high-end products could become a new earnings catalyst. The company's expansion strategy remains disciplined, with the focus shifting from increasing total scale to raising the proportion of high-end products. China Jushi prefers to expand capacity gradually to prolong the current upcycle; new capacity in Huai'an is expected to ramp progressively during 2028—2030. Given the high entry barriers, new competitors may struggle to establish meaningful supply before 2030. Morgan Stanley therefore remains constructive on the industry's outlook for 2027—2028: limited new supply, rising material intensity in AI and PCBs, and the still-early stage of data center construction in China should all support demand; traditional roving prices may also improve in 2027. The financial forecasts reflect the above volume, pricing, and product-mix assumptions. The report forecasts earnings per share of Rmb0.82, Rmb1.85, Rmb2.95, and Rmb3.33 for 2025—2028, respectively, versus Refinitiv consensus estimates of Rmb0.84, Rmb1.55, Rmb1.96, and Rmb2.43. Over the same period, net revenue is forecast at Rmb18,881mn, Rmb26,736mn, Rmb33,810mn, and Rmb36,315mn; EBITDA at Rmb6,951mn, Rmb11,534mn, Rmb17,183mn, and Rmb19,222mn; and ModelWare net income at Rmb3,285mn, Rmb7,421mn, Rmb11,804mn, and Rmb13,317mn. ROE is forecast at 10.9%, 23.9%, 33.5%, and 31.8%, respectively, while RNOA is forecast at 9.3%, 19.4%, 28.8%, and 31.4%. On valuation, Morgan Stanley views China Jushi as an AI-related materials stock and uses a relative valuation methodology, applying a 30x P/E multiple to its 2027 earnings forecast; this multiple is below those of relevant peers. Applying the multiple to 2027 EPS of Rmb2.95 yields a target price of Rmb88.50, implying 111% upside from the closing price of Rmb42.04 on August 21, 2026. The report lists P/E multiples of 20.8x, 22.7x, 14.3x, and 12.6x for 2025—2028, respectively; P/BV multiples of 2.2x, 4.8x, 4.0x, and 3.4x; EV/EBITDA multiples of 11.3x, 15.3x, 10.0x, and 8.6x; and dividend yields of 2.1%, 1.9%, 3.1%, and 3.5%.

Analysis framework

The report first assesses current supply-demand conditions based on customer ordering behavior, prices, inventory, and order fulfillment rates. It then analyzes capacity expansion bottlenecks individually, including furnaces, looms, processing, labor, and certification, to evaluate how long the tight market could persist. It subsequently combines progress in AI electronic fabric products, yield requirements, and the company's capacity expansion plans to project the growth and supply trajectory for 2027—2030. Finally, it incorporates its operational assumptions into ModelWare financial forecasts and derives the target price using 2027 earnings and a 30x P/E multiple.

Methodology notes

• Industry/Sector Analysis FrameworkSupply-demand framework

  Inventory, Order Fulfillment Rate, and Supply Bottleneck Analysis

  Using indicators such as 3.9 days of inventory, an order fulfillment rate of about 80%, customers competing for supply, and equipment delivery lead times, the report concludes that demand exceeds effective supply and accordingly infers that the price upcycle could continue.

• Industry/Sector Analysis FrameworkUpstream-Midstream-Downstream Value Chain Transmission

  Electronic Fabric—Copper-Clad Laminate—PCB Value Chain Inventories and AI Demand Transmission

  The report observes low inventories across electronic fabric, copper-clad laminates, and PCBs, and traces stronger AI demand through to orders, prices, and the earnings outlook for 7628 and higher-end electronic fabrics.

• Company Fundamentals and Financial Framework

  Morgan Stanley ModelWare Financial Forecasting Framework

  The report uses ModelWare to produce forecasts for revenue, EBITDA, net income, earnings per share, and returns, and compares them with Refinitiv consensus estimates.

• Valuation MethodPE/PEG valuation

  Relative P/E Valuation Based on 2027 Earnings

  Morgan Stanley applies a 30x P/E multiple to 2027 EPS of Rmb2.95 to derive a target price of Rmb88.50; because the company is viewed as an AI-related materials stock, the multiple is also described as lower than those of relevant peers.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• China Jushi (600176.SS, 600176 CG)

  Tight supply and demand for standard fiberglass and 7628 electronic fabric directly support prices and earnings, while high-end AI electronic fabric is viewed as an additional growth driver after 2027.

  Strengths

  Inventory is only 3.9 days, and the company can fulfill only about 80% of orders; it has a foundation in high-end products, processes, and certification, and is pursuing a disciplined capacity expansion strategy to increase the proportion of high-end products.

  Weaknesses

  Furnaces, imported looms, processing capacity, skilled workers, and certification all constrain the release of effective capacity, while high-end AI electronic fabric remains at an early stage.

  Comparison

  The report values the company as an AI-related materials stock but applies a 30x 2027 P/E multiple that is below those of relevant peers.

  Risks

  A slowdown in global demand, higher raw material prices, or industry overcapacity could weaken pricing and earnings; conversely, stronger-than-expected demand, lower raw material prices, or more factory maintenance shutdowns represent upside risks.

Key data

• Stock Rating and Industry ViewOverweight / AttractiveThe stock rating and Greater China materials industry view currently presented in the report
• Target Price, Closing Price, and UpsideRmb88.50 / Rmb42.04 / 111%Closing price as of August 21, 2026; the target price applies a 30x multiple to 2027 earnings
• 7628 Price ChangeMore than doubled year-to-dateCustomers are still competing for supply, and the report expects further price increases may occur in 2H26
• China Jushi Inventory3.9 daysIllustrates the tight supply of electronic fabric
• Order Fulfillment RateAbout 80%The company is currently still unable to fully meet customer orders
• Equipment Delivery Lead TimeExtending as late as 2029Bottlenecks in looms and other equipment constrain the supply response in 2027—2028
• Earnings per Share ForecastRmb0.82 / Rmb1.85 / Rmb2.95 / Rmb3.33Corresponding to 2025, the 2026 forecast, the 2027 forecast, and the 2028 forecast, respectively
• Refinitiv Consensus Earnings per ShareRmb0.84 / Rmb1.55 / Rmb1.96 / Rmb2.43Corresponding to 2025—2028, respectively
• Net Revenue ForecastRmb18,881mn / Rmb26,736mn / Rmb33,810mn / Rmb36,315mnCorresponding to 2025—2028, respectively
• EBITDA ForecastRmb6,951mn / Rmb11,534mn / Rmb17,183mn / Rmb19,222mnCorresponding to 2025—2028, respectively
• ModelWare Net Income ForecastRmb3,285mn / Rmb7,421mn / Rmb11,804mn / Rmb13,317mnCorresponding to 2025—2028, respectively
• Current Market Capitalization and Enterprise ValueRmb168,292mn / Rmb178,163mnThe current market capitalization and enterprise value presented in the report
• Diluted Shares Outstanding4,003mnThe current share count presented in the report
• Average Daily Trading ValueRmb3,359mnThe average daily trading value presented in the report
• 52-Week Price RangeRmb77.20-12.96In the original order presented in the report

Impact & implications

The report believes that supply expansion is proceeding more slowly than growth in AI and PCB demand, allowing the price and earnings upcycle for standard fiberglass and 7628 electronic fabric to potentially continue through 2027—2028. Progress in high-end AI electronic fabric could further enhance the product mix and profitability beginning in 2027. Gradual capacity expansion and high entry barriers also mean that new supply may only begin to emerge progressively around 2028—2030.

Risks

• Upside risk: Market demand is stronger than expected.
• Upside risk: Raw material prices decline.
• Upside risk: More factories than expected enter maintenance shutdowns.
• Downside risk: Global demand slows.
• Downside risk: Raw material prices rise.
• Downside risk: Industry overcapacity.

What to watch

• Monitor the magnitude of further price increases for 7628 electronic fabric and customer ordering strength in 2H26.
• Monitor whether the company's 3.9 days of inventory and approximately 80% order fulfillment rate indicate that the supply-demand gap is persisting.
• Monitor substantive progress in low-Dk G2, low-CTE, and Q-fabric in 2027.
• Monitor how equipment delivery lead times, customer certification, and domestic loom yields constrain supply releases in 2027—2028.
• Monitor the gradual ramp-up of new capacity in Huai'an during 2028—2030.
• Monitor whether traditional roving prices can improve in 2027.