Business
Know the Business
DAQO New Energy is a pure-play polysilicon commodity producer selling a single product — high-purity polysilicon — into the solar PV supply chain. The stock's value is determined almost entirely by polysilicon spot prices, which collapsed 80%+ from the 2022 peak as Chinese capacity tripled. The market is pricing DAQO at 0.45× book value, implying permanent impairment of its $3.4B manufacturing base — the key question is whether that impairment is real or whether the cycle turns before the cash pile runs out.
How This Business Actually Works
DAQO converts metallurgical-grade silicon and electricity into ultra-pure polysilicon using the modified Siemens process with hydrochlorination (closed-loop recycling). The economics are brutally simple:
Revenue = Volume (MT) × ASP ($/kg). There is one product line, one geography (China), and prices are set at prevailing spot when orders ship. Top 3 customers account for 64% of revenue.
Cost = Electricity + MG-Si + Depreciation + Overhead. Electricity is the dominant variable cost. DAQO's facilities in Xinjiang and Inner Mongolia exploit cheap coal-based power — historically $0.04-0.05/kWh versus $0.08-0.10+ in coastal China or $0.10-0.15 in Germany. This gave DAQO a structural cost advantage, with all-in production costs as low as $6.38/kg in late 2019.
The capacity trap: DAQO expanded from 35,000 MT (2019) to 305,000 MT (2024) — a 9× increase. The industry did the same. Chinese polysilicon capacity surged from roughly 600,000 MT in 2020 to over 2,000,000 MT by 2024. Demand grew fast (China installed 59.7 GW in Q1 2025 alone, +30.5% YoY), but supply grew faster. Polysilicon ASPs fell from $30+/kg (Q1 2022) to below $5/kg — well below most producers' cash costs.
Result: DAQO ran at only 40% utilization in FY2025 (123,652 MT produced vs. 305,000 MT capacity) and reported negative gross margins for six consecutive quarters (Q2 2024 through Q1 2026).
Parser Error: syntax error at or near "union"
LINE 2: ... revenue, 98 as gross_profit, 38 as net_income order by year union all
^
LINE 2: ... revenue, 98 as gross_profit, 38 as net_income order by year union all
^Revenue 15×'d from FY2018 to FY2022 on volume growth and polysilicon pricing, then collapsed 86% to FY2025 as prices cratered. The company swung from $1.8B net income (FY2022) to -$171M net loss (FY2025) — a $2B earnings swing driven almost entirely by a single commodity price.
The Playing Field
Parser Error: syntax error at or near "union"
LINE 2: ..., 1.9 as cash_b, 'Modified Siemens' as tech order by ticker union all
^
LINE 2: ..., 1.9 as cash_b, 'Modified Siemens' as tech order by ticker union all
^The polysilicon market is an oligopoly becoming a glut. Tongwei is the 800-pound gorilla — 3× DAQO's capacity, vertically integrated into solar cells, and trading at 1.2× book because the market assumes it survives. GCL Technology uses a fundamentally different granular polysilicon process (FBR) that is cheaper per kg but historically lower purity. Xinte (TBEA subsidiary) is DAQO's closest peer — similar Xinjiang location, similar cost structure, similar pain.
The western producers (Wacker, OCI) are higher-cost but more diversified. Wacker generates only ~30% of revenue from polysilicon; the rest is silicones and polymers. This diversification provides a floor that pure-play producers lack.
What the peer set reveals: In a commodity downcycle, diversification and vertical integration determine survival. DAQO has neither — but it has the largest cash buffer relative to its market cap of any pure-play peer. At $1.9B in cash and zero debt against a $2.0B market cap, the stock is essentially priced at the value of its cash minus expected future losses, with the $3.4B manufacturing base valued at zero.
Is This Business Cyclical?
Parser Error: syntax error at or near "union"
LINE 2: ... year, 32.5 as gross_margin, 12.6 as net_margin order by year union all
^
LINE 2: ... year, 32.5 as gross_margin, 12.6 as net_margin order by year union all
^This is one of the most violently cyclical businesses in public markets. Gross margins swung from +74% (FY2022) to -21% (FY2024-2025) — a 95-percentage-point swing in two years.
Where the cycle hits: Polysilicon is a commodity with high fixed costs (depreciation, electricity minimums, overhead) and a price set by supply/demand. When supply overshoots, prices crash below variable cost, forcing producers to idle capacity or operate at losses. The current downcycle echoes the 2011-2013 bust that sent DQ's stock below $1 and bankrupted dozens of producers.
Critical difference from the last bust: In 2012, DAQO had $280M in debt and $52M in cash. Today it has zero debt and $1.9B in cash. The balance sheet buys time — but not indefinitely. At the FY2025 cash burn rate (~$160M operating cash consumed), the cash pile would last roughly 12 years. However, Q1 FY2026 revenue collapsed to just $27M, suggesting the burn rate could accelerate if conditions worsen.
What triggers the turn: The Chinese government has declared "anti-involution" a national policy priority and is drafting mandatory energy consumption limits for polysilicon production. This regulatory push could force high-cost capacity offline. Industry utilization at ~50% suggests 500,000+ MT of capacity needs to exit before pricing normalizes.
The Metrics That Actually Matter
Parser Error: syntax error at or near "union"
LINE 2: ...#1 — below cash cost = losses' as why_it_matters order by metric union all
^
LINE 2: ...#1 — below cash cost = losses' as why_it_matters order by metric union all
^Polysilicon ASP — the single variable that determines everything. Revenue, margins, FCF, and survival all move in lockstep with this number. At $5/kg, everyone loses money. At $10/kg, DAQO breaks even. At $15/kg, it prints cash.
Cash cost per kg — DAQO's historical cost advantage ($6.38/kg in 2019) came from cheap Xinjiang electricity. But at 40% utilization, fixed costs per kg balloon. Cost leadership only matters when you can produce volume.
Utilization rate — At 305,000 MT capacity running at 40%, DAQO produced 123,652 MT in FY2025. Each percentage point of utilization improvement drops costs per kg and improves operating leverage.
Cash balance minus expected burn — $1.9B with zero debt is the moat in a downcycle. But DAQO burned ~$160M in operating cash in FY2025 and faces continued losses. Monitor quarterly.
Industry capacity exits — The cycle turns when enough high-cost capacity shuts down. Watch for bankruptcy filings, mothballing announcements, and the effect of China's new energy consumption standards on marginal producers.
What I'd Tell a Young Analyst
This is not an earnings story — it is a survival and optionality story. The market is pricing DAQO's 305,000 MT manufacturing base at roughly zero and paying $2B for $1.9B in cash plus whatever option value exists on the cycle turning. The equity is essentially a levered call on polysilicon prices.
What to watch: Monthly Chinese polysilicon spot prices (published by the Silicon Industry branch of China Nonferrous Metals Industry Association) and quarterly production/inventory reports from major producers. The first sign of a turn will be sustained production cuts from high-cost Chinese producers, followed by spot price stabilization above $8-10/kg.
What the market may be missing: China's anti-involution policy is real and has teeth — the proposed mandatory energy consumption standard would disproportionately hurt smaller, less efficient producers while DAQO's newer Inner Mongolia facilities meet the threshold. If enforced, this is a regulatory moat that accelerates supply rationalization. Conversely, the PFIC classification for US holders and ongoing US-China regulatory uncertainty create structural selling pressure on the ADSs that may keep the stock cheap even as fundamentals improve.
The risk the bull ignores: DAQO's $1.9B cash pile sits primarily in China, trapped by capital controls and dividend restrictions. The Cayman holding company depends on dividends from Chinese subsidiaries, which are subject to statutory reserve requirements and regulatory approval. The cash is real, but not all of it is accessible.