CSIQ · Canadian Solar Inc.

Current Thesis

The solar-policy-relief rally that carried the whole complex off its April lows is rolling over, and CSIQ is the weakest hand in it. The name ran to a 52-week high of $34.59, broke above its upper Bollinger Band on 2026-05-08 (RSI ~77), then hit the wall: the Q1 print on 2026-05-14 carried a net loss, the Q2 gross-margin guide collapsed to 13–15% from 25.1%, and a ~$200M convertible notes offering flagged a capital need. On 2026-06-05 the stock fell 11.86% to $17.16, slicing through the 50-day MA (~$25) and pinning itself to the 200-day (~$17.74). This is a rolled-over structure with a cheap-looking multiple the value-trap pattern the playbook exists to avoid, not an accelerating narrative to ride. The one live bull leg is e-STORAGE ($3.5B contracted backlog / 34 GWh as of 2026-05-08). That story is real but it does not own the tape right now; margin compression, dilution, and a sector-wide de-rate do. Watch, don't chase.

Bull Case

• e-STORAGE backlog is the franchise: $3.5B contracted / 34 GWh under long-term service agreements as of 2026-05-08; 20+ GWh cumulative shipped as of 2026-03-31. Storage is the high-multiple growth segment, decoupled from module commodity pricing.
• Q1 (2026-05-14) hit high end of guide: revenue $1.1B with 2.5 GW modules + 2.1 GWh storage recognized, both above company guidance. Demand is there at the top line.
• Q2 shipment ramp guided up: 3.1–3.3 GW modules and 2.8–3.2 GWh storage for Q2, a sequential volume step-up; FY US guide 6.5–7.0 GW modules + 4.5–5.5 GWh storage.
• Tariff refund tailwind is cash, not just accrual: IEEPA refunds began arriving in cash with more expected into Q2 a real balance-sheet event, not paper.
• Cheap on price-to-sales: ~$1.16B market cap against $5.48B ttm revenue. If margins stabilize and the policy overhang clears, the re-rate room is large (high analyst PT $37).

Bear Case

• Margin collapse is the headline: Q2 gross-margin guide of 13–15% vs Q1's 25.1% and Q1 itself was juiced by an 860bps tariff-refund accrual. Underlying module economics are deteriorating fast.
• Unprofitable and bleeding: Q1 net loss of $32M (-$0.71/diluted share); ttm EPS -$2.52, ttm net income -$102M. Plus a $29M FX loss in the quarter and negative cash flow.
• Dilution overhang: a ~$200M convertible notes offering signals the balance sheet needs feeding a tell that contradicts the "cheap" thesis.
• Wrong side of US policy: China-exposed supply chain into US tariffs and the active TOPCon Section 337 case (337-TA-1494, procedural updates late May 2026). FSLR gets paid via Section 45X domestic-manufacturing credits; CSIQ does not have that shield.
• Leadership transition mid-storm: founder Shawn Qu stepped to Executive Chairman/CTO; Colin Parkin named CEO on 2026-05-14 a handoff into a margin air-pocket.
• Analyst tape is cold: consensus Hold; Freedom Broker initiated/cut to a $16 PT on 2026-05-19; median PT clustered ~$17–21, i.e. roughly zero upside to spot.

Setup & Price Structure

Spot $17.16 (close 2026-06-05), down -11.86% on the day. 52-week range $9.41–$34.59 so the stock is now ~50% off its May high and sitting roughly 32% off the lows. The move above the 50-day on 2026-05-01 marked the start of the policy rally; the upper-Bollinger break on 2026-05-08 marked the blow-off; the post-earnings fade plus the 2026-06-05 gap-down completed the round trip. Price has lost the 50-day (~$25) decisively and is now resting on the 200-day (~$17.74) the last shelf before air down to the $9–10 base. MACD histogram flipped negative on 2026-06-01, confirming momentum had already rolled before the June 5 break. The whole complex moved together that day (FSLR -8%) on higher Treasury yields and ITC phase-down fear, so this is a sector de-rate, not an idiosyncratic blow-up which makes a knife-catch here a bet on the macro/policy tape, not on CSIQ's own setup. No clean higher-low has printed; there is no momentum entry until the structure repairs.

Catalyst Calendar (next 30 days)

• ~2026-07-04 (est., HARD policy date): commercial-solar projects must begin construction by July 4, 2026 to qualify for the full 30% ITC. A pull-forward of bookings into the deadline, then a demand cliff thesis behind it binary for the whole sector's H2 narrative.
• Ongoing TOPCon Section 337 case (337-TA-1494): ITC procedural updates were posted late May 2026; any ruling or schedule headline moves the China-exposed names (CSIQ included) on the day.
• Treasury-yield / risk-off tape (daily): rate-sensitive renewables trade inversely to the 10Y; another yield spike pressures the group regardless of company news.
• No company-specific binary in-window: Q1 already reported 2026-05-14; next earnings (Q2) is ~mid-August 2026, outside the 30-day window so no earnings-blackout blocker right now.

What Would Change Our Mind

Bullish flip (turns this into an actionable momentum re-entry): a daily close back above the rolled-over 50-day MA (~$25) accompanied by a Q2 gross-margin guide that stabilizes above 15% and a sector that is leading, not lagging, the tape. Short of margin repair, even a bounce is a dead-cat to fade. Bearish confirmation: a daily close below the 200-day (~$17.74) / the $17 shelf opens the path back to the $9–10 base and confirms the value trap. The e-STORAGE backlog growing past $3.5B with rising attach margins would be the fundamental tell that the franchise is real but the tape, not the backlog, gates entry here.

Correlation Notes

CSIQ is a high-beta expression of the clean-energy / solar-policy narrative and trades as a pack with FSLR, ENPH, SEDG, RUN and the TAN ETF confirmed by the 2026-06-05 broad selloff. Within that pack it is the laggard: US-manufacturing names (FSLR, paid via Section 45X) are structurally insulated, while China-supply-chain names (CSIQ) carry the tariff/337 overhang and weaker margins, so CSIQ underperforms on down days and is a lower-quality vehicle on up days. Inverse correlation to the 10Y Treasury yield is the dominant macro driver. The e-STORAGE segment gives a secondary read-through to the grid/data-center power-demand theme, but that linkage has not yet been the price driver. For pairs purposes, CSIQ is the high-beta/low-quality leg against FSLR as the quality leg of the same trade.

[recent archive last 6 entries]

(unchanged — see prior file)