Chapter 5

Waiting on Beijing

Daqo's case rests on polysilicon prices recovering above cost before the trough spends its balance sheet. The past year shows that recovery is a policy outcome, not a market one: prices cleared cost in late 2025 only after Beijing intervened, then relapsed below cost within two quarters when enforcement stalled [1]. Demand offers no independent lift. The live test is whether China's price law is enforced above cost around mid-2026.

The rebound, and the relapse

For most of 2024 and the first half of 2025 the polysilicon market did what a competitive commodity in deep oversupply does — it fell until it was below everyone's cash cost. China spot polysilicon slid from roughly RMB39–45/kg in April 2025 to RMB32–35/kg by the end of June, and Daqo ran its plants at about 34% of nameplate, scaling back sales rather than adding supply [2].

What turned it was not demand clearing the glut but the Chinese state. From late June 2025 the government made "anti-involution" — curbing ruinous below-cost competition — an explicit priority: a People's Daily article on June 29, a July 1 statement by President Xi on regulating disorderly low-price competition, and a July 2 Ministry of Industry symposium with fourteen solar manufacturers [3]. Prices responded: they surged more than 50% from the mid-2025 lows to RMB50–56/kg by year-end [4], and Daqo's annual ASP decline slowed to 7.3% in 2025 after falling 66.7% and 50.7% in the two prior years [5]. Daqo booked a positive gross margin in Q4 2025 for the first time in the trough.

The rebound did not hold. By the end of Q1 2026 spot had fallen back to roughly RMB35–37/kg — below production cost again — and Daqo took a $98.4 million inventory impairment, driving gross margin to negative 521% [6]. The late-2025 recovery, in other words, was the direct product of a policy push, and it reversed when the follow-through did not arrive on schedule.

Spot is the midpoint of reported RMB market ranges converted at ~RMB7.0/USD; "Daqo cost" is FY2025 full production cost, "price floor" is the mooted RMB53–54/kg price-law minimum. Sources: Q2 2025 presentation [7]; FY2025 Annual Report [8]; Q4 2025 call [9]; Q1 2026 call [10].

What enforcement would take

Management has been explicit about the mechanism it is waiting on. Under the amended Pricing Law, sales "should not be below the industry level cost," and Daqo's read is that the floor would sit at "at least RMB53–54 per kilogram" — roughly $7.6/kg, and about 15% above its own $6.61/kg full cost [11]. If that floor were set and enforced, Daqo would return to reported profit and, as one of the lowest-cost producers, expect to regain share. That is the bull mechanism in a sentence.

The obstacle is the distance between rhetoric and enforcement. The policy scaffolding is real and building: Beijing has formally designated anti-involution a national priority in the 15th Five-Year Plan, revised the Anti-Unfair Competition Law and drafted a Pricing Law amendment, and proposed a mandatory national standard imposing strict per-unit energy-consumption limits on polysilicon production — an instrument that would fall hardest on high-cost, high-energy producers rather than on Daqo [12]. But as of the Q1 2026 call the price had not yet been set: a fresh government "round of price determination" was expected around mid-2026, with the enforcement method still under discussion and Daqo, in its own words, "in observation mode" [13]. The tell is in the futures curve: at the time of the Q4 2025 call the polysilicon future traded at RMB46.3/kg — near Daqo's full cost, and well below the RMB53–54 the company hoped policy would deliver [14]. The market, in short, was not pricing enforcement.

The demand ceiling

Even a policed floor would only stop prices falling; it does not create the demand that clears a 600,000-tonne inventory mountain. By the end of Q1 2026, industry polysilicon inventory had climbed to nearly 600,000 metric tons — with Tier-1 producers alone holding at least three months of stock — even as industry monthly supply fell to about 93,000 tonnes at just 39% utilization [15][16]. Supply has been cut hard; the overhang persists because demand is not growing fast enough to absorb it.

The demand picture is one of decelerating growth, not decline. Global solar installations reached roughly 580 GW in 2025 — a record — but growth slowed to about 9% after rising in the mid-30s-percent range in the prior two years [17]. And the demand that exists is increasingly policy-timed rather than organic. China added a record 93 GW in May 2025 alone — then just 14 GW in June — as developers rushed to beat a May 31 cutoff for market-based pricing, pulling demand forward and leaving an air-pocket behind it [18].

Annual growth decelerated from roughly +36% (2024) to +9% (2025). Source: FY2025 Annual Report, Item 3 Risk Factors [19].

Two structural headwinds sit on top of the deceleration. First, China is removing the export subsidy that props up its module makers: the value-added-tax rebate on exported solar products falls from 9% to 6% on April 1, 2026 and is abolished entirely from January 1, 2027 [20]. Polysilicon itself is sold under a 13% VAT with no refund, so the cut does not hit Daqo directly — it bites indirectly, by raising the delivered cost of Chinese modules abroad and dampening the export demand that pulls polysilicon through the chain [21]. Management flagged exactly this dynamic on the Q1 2026 call: a January–February export rush to beat the rebate deadline, then downstream buyers holding off on polysilicon purchases amid uncertainty about demand after April 1 [22]. Second, the entity-list and UFLPA barriers established earlier (The Polysilicon Cycle) keep Daqo's Xinjiang-made polysilicon out of the largest premium-priced end market, so the US installation base does little for its order book.

How Daqo is playing it, and what to watch

Daqo has turned its low cost into a waiting strategy rather than a share-grab. Through the trough it has repeatedly declined to sell below cost: in Q1 2026 it sold only 4,482 tonnes against 43,402 produced, recognizing a $5.96/kg ASP on that thin volume while holding the rest as inventory in anticipation of a policy-led recovery [23]. It has tied its production plan directly to enforcement: guidance is to hold utilization at 50–55%, but management has said plainly that if the price law is not enforced and rivals keep selling below cost, "then we will lower our utilization" and sell closer to market [24]. The company's own volumes have become a live signal of whether it believes the recovery is real.

That framing yields a short list of checkable markers, each with a filing line and a threshold:

Source: synthesized from Q1 2026 and Q4 2025 earnings calls and the FY2025 Annual Report, as cited above [25].

The measured read: the recovery the whole case waits on is real but conditional, and the condition is policy rather than the cost curve doing its work unaided. The strongest fact for the bulls is that the mechanism has already fired once — prices jumped more than 50% in H2 2025 when Beijing pushed [26], and Daqo briefly turned gross-margin positive [27] — so this is a demonstrated lever, not a hope. The strongest fact against them is that it reversed within two quarters, that a 600,000-tonne inventory overhang and decelerating demand give the free market no help, and that enforcement kept slipping into "observation mode." What would settle it is the mid-2026 price determination: a credible above-cost floor with teeth turns Daqo's balance sheet from a defensive buffer into a return; another lapse leaves the company idling plants and drawing down cash while it waits for the next push.