Canadian Solar is far more expensive than its market cap suggests: $7.8 billion of debt and $1.4 billion of minority interest bring its enterprise value to $8.7 billion. x.com 

Canadian Solar has no thesis without FEOC and 45X compliance, because those rules gate every leg of its business, battery storage included. Storage faces the strictest FEOC thresholds, Canadian Solar makes batteries in China and its anode material is under a US AD/CVD case, so its Kentucky battery plant is a compliance effort, not a hedge. US non-Chinese wafer supply is tiny and committed, with Corning's output going to T1 Energy, Suniva and Heliene; its own would take years and billions. x.com 

Canadian Solar's US energy-storage business is rewriting its valuation, and the shares look cheap against the sum of its parts. e-STORAGE delivered a record 2.7GWh in Q3 2025 with $3.1B of backlog, lifting gross margin to 17.2%, and US cell, module and battery plants under CS PowerTech target FEOC tax-credit rules. Modules, storage and project development are worth about $2.5-4B versus a $1.2-1.3B market cap. The discount reflects about $6.65B of debt and FEOC uncertainty. x.com  x.com 

Canadian Solar's US business will keep carrying a conglomerate discount until it is fully ring-fenced or spun off. Forming CS PowerTech to retake control of US operations is a legal restructuring, not a clean capital separation, so while the US arm sits under the parent, top-level capital allocation risk drags down the whole valuation. A 90% holdco discount would be an extreme edge case. x.com  x.com 

Eos Energy, Fluence and FuelCell Energy are best avoided because capital-intensive energy technology requires super-high capex. These small companies have no way to compete with industrial giants that have far greater borrowing power. Bloom Energy is the exception, with lower capitalization and a unique product delivered within 60-90 days of contract signing. x.com 

Canadian Solar's pullback is an opportunity: management is aggressively targeting the US market, where it brings utility-scale solar experience, and its Australian solar and storage business is strongly government-supported. It belongs in the same tier as T1 Energy, Tigo Energy, Nextracker and Array Technologies. x.com 

Fluence Energy's $5.6B backlog exceeds its roughly $3.4B market cap by about 65%, so contracted future revenue is worth more than the business today, assuming execution holds. FY26 revenue guidance is $3.2-3.6B, and the backlog excludes a 12GW hyperscaler pipeline; announcing two hyperscaler deals in Q2 or Q3 could drive a bigger re-rating. Its software business, growing ARR 22% on top of the installed base, deserves at least 3x sales versus 0.5x for the whole company. x.com  x.com 

Canadian Solar is probably as hurt as anyone by China's polysilicon policy, which at the very least piles more uncertainty onto a story already bogged down with it. The policy also dings JinkoSolar by hitting its workarounds for entering the US market. Chinese solar has still risen on the broader anti-involution story, a bid that reads as a sentiment shift and growing confidence in anti-involution. x.com 

Canadian Solar may be one of the market's most overlooked AI infrastructure plays, because it is building the energy and storage backbone AI data centers badly need while investors chase chipmakers. x.com 

A polysilicon price floor plus 15% Section 232 tariff, effective December 4, 2026, favors US producers and hurts those reliant on cheap Chinese poly. Corning's Hemlock is the US champion the policy was written for, and First Solar uses no polysilicon, so rivals' costs rise while its stay flat. Daqo's $5/kg Chinese poly is the target, Sunrun faces higher panel costs with subsidies gone, and Canadian Solar's US plants depend on imported inputs. Installers' Q4 prints may be inflated by stockpiles. … Show more