KC · Kingsoft Cloud Holdings Limited

Current Thesis

Kingsoft Cloud’s artificial-intelligence infrastructure story now depends on sustaining operating profit and repairing a rebound already invalidated below $11.00. The 2026-09-11 adjusted close of $9.96 extends the deterioration recorded at $10.64 on 2026-08-28. The operating evidence remains stronger than the price structure: results released on 2026-08-19 showed second-quarter revenue of RMB3,072.0 million, up 30.8% year over year, and operating profit under generally accepted accounting principles (GAAP) of RMB23.0 million.

As a price narrative, the narrative is dead — the 2026-09-11 close remains below the published $11.00 invalidation threshold despite the 2026-08-19 profit inflection. This is an interpretation of the broken rebound, not a conclusion that demand for computing capacity has disappeared. A close above the pre-results reference of $12.33, recorded on 2026-08-07, together with another profitable operating quarter would contradict that interpretation.

The funding evidence has advanced since the previous report. Kingsoft Cloud’s monthly return, submitted on 2026-09-04, records no change in issued shares during August. That resolves the anticipated August issuance check; it does not establish how subsequent infrastructure spending will be financed. August monthly return

Bullish and bearish views on Kingsoft Cloud Holdings Limited

The model's bull view on Kingsoft Cloud Holdings Limited (KC), in brief: AI demand reached reported scale. The 2026-08-19 results reported second-quarter AI cloud gross billings of RMB1,327 million, up 82% year over year and equivalent to 56% of public-cloud revenue. Gross billings are the disclosed demand measure; they are not interchangeable with… The bear view: The published rebound condition failed. Adjusted daily bars show $10.64 on 2026-08-28 and $9.96 on 2026-09-11, both below the previously published $11.00 thesis-break level. The later observation extends an existing failure; it does not establish a new support shelf. Capital… Both cases follow in full.

Bull Case

• AI demand reached reported scale. The 2026-08-19 results reported second-quarter AI cloud gross billings of RMB1,327 million, up 82% year over year and equivalent to 56% of public-cloud revenue. Gross billings are the disclosed demand measure; they are not interchangeable with recognized revenue.
• Operating profitability became measurable. The 2026-08-19 release reported second-quarter gross margin of 15.2%, versus 12.8% in the first quarter, and GAAP operating profit of RMB23.0 million. A subsequent operating loss would contradict the interpretation that this marks durable operating improvement.
• August brought no issued-share increase. The return filed on 2026-09-04 reports 4,531,784,801 issued ordinary shares at both July and August month-end. It provides no evidence of an August equity expansion. August monthly return

Bear Case

• The published rebound condition failed. Adjusted daily bars show $10.64 on 2026-08-28 and $9.96 on 2026-09-11, both below the previously published $11.00 thesis-break level. The later observation extends an existing failure; it does not establish a new support shelf.
• Capital spending challenges the plan. On the 2026-08-19 call, management reported second-quarter capital expenditure including capitalized leasing arrangements of RMB3.3 billion after RMB3.0 billion in the first quarter, while reiterating a full-year RMB6.2 billion base case. The inference is financing pressure; a new offering or a disclosed increase in financing requirements would confirm it.
• Profitability remains incomplete and uneven. The 2026-08-19 release reported a second-quarter net loss of RMB93.0 million and enterprise-cloud revenue of RMB714.3 million, down 1.3% year over year. The positive operating quarter therefore does not establish net profitability or growth across both cloud segments.

Setup & Price Structure

The 2026-09-11 adjusted close was $9.96, with the shares 45.3% below the reported 52-week high of $18.21 and down 7.2% over three months. The 14-day relative strength index (RSI) was 34.1. These observations describe weak price momentum; no moving-average level or trading-volume series is available to establish a base or expanding participation.

Analyst attention continued after the earnings release: StreetInsider reported on 2026-09-01 that Goldman Sachs analyst Timothy Zhao reiterated an $18.00 price target. The subsequent $9.96 reference close does not demonstrate convergence toward that forecast. An analyst target is evidence of published expectations, not evidence of investor inflows. Goldman Sachs target reiteration

Current short interest, options positioning and fund-flow measurements are missing. The available coverage observations are too few to support a claim of retail crowding or institutional accumulation.

Catalyst Calendar (next 30 days)

• ~2026-09-30, unconfirmed publication window: The interim report for the six months ended 2026-06-30 remains the disclosure anticipated in the 2026-08-30 research note. A company-confirmed publication date has not been established. Its relevance is the financing and lease detail behind the capital spending reported on 2026-08-19; the calendar assumption fails if publication does not occur in that window.

No company-confirmed upcoming earnings date was identified as of 2026-09-13; the investor-relations calendar still lists the completed 2026-08-19 second-quarter call. The August securities return was filed on 2026-09-04 and is no longer an upcoming catalyst. Company investor relations HKEX filing record

What Would Change Our Mind

The July–August rebound has already lost its published price condition: a daily close below $11.00 invalidates that leg, and the 2026-09-11 close of $9.96 satisfies it. Lowering the threshold would change the thesis rather than repair the evidence.

Reconsideration requires observable repair: a daily close above the 2026-08-07 reference close of $12.33 and a subsequent results release retaining positive GAAP operating profit. A gross margin below the second quarter’s 15.2% or a return to operating losses would instead weaken the operating-improvement case established on 2026-08-19. One profitable quarter is too small a sample to establish durability.

Correlation Notes

This remains a single-company setup; no measured peer-price or correlation series accompanies the 2026-09-11 observations. A broader China technology or AI infrastructure advance cannot therefore be counted as confirmation.

The economic connection to Xiaomi and Kingsoft is observable in management’s 2026-08-19 disclosure: ecosystem revenue grew 28% year over year, while revenue from top non-ecosystem customers grew 51%. Those figures describe customer exposure. They do not establish how KC’s shares correlate with either company or with a technology index.