Summary
- GDS reported RMB837.6 million of Q2 net income after recording RMB959.9 million of income from equity-method investees. The company says that investee line arose mainly from a dilution gain on its DayOne holding, but it does not disclose the exact DayOne amount.
- The gain was not the same as cash received. GDS's core operating test was less exuberant: revenue rose 6.5%, gross profit fell 3.6%, adjusted EBITDA rose 2.5%, and adjusted EBITDA margin narrowed from 47.3% to 45.5% as utilities consumed more revenue.
- With RMB46.1 billion of short- and long-term debt, capex guidance raised to RMB10 billion and net committed area running well ahead of net utilized additions, the useful receipt is how quickly funded construction becomes occupied, cash-generating capacity.
A line larger than the bottom line
The cleanest way to read GDS's second quarter is to resist the first number in the highlights. Net income of RMB837.624 million was a sharp reversal from the prior-year loss. Yet the income statement also recorded RMB959.916 million as GDS's share of results of equity-method investees. That single line exceeded the whole company's net income by RMB122.292 million.
Scale is the point; subtraction is not. The investee figure cannot simply be removed to manufacture an “underlying loss.” GDS reported RMB439.171 million of income from operations, RMB366.816 million of net interest expense, and RMB91.243 million of income before tax and the share of equity-method investees. It then recorded RMB213.535 million of tax expense and the RMB959.916 million investee contribution before arriving at net income. Tax, interest and other items have their own timing and causes.
The company's wording also sets an attribution limit. GDS says the investee income arose mainly from a dilution gain on its DayOne investment after DayOne completed its Series C convertible preferred-share issue. “Mainly” does not mean the full RMB959.916 million is an exact disclosed DayOne gain. The release does not provide that component.
This distinction matters because the net-income headline combines two economic surfaces. One is GDS's capital-intensive China data-centre operator. The other is a minority interest in an international platform whose latest financing changed GDS's ownership percentage and the accounting value of what it retained.
Dilution can create income without a sale
DayOne announced in June that its final Series C closing had taken total gross proceeds to US$4.5 billion. Coatue and Hillhouse led the round and became its two largest shareholders. New capital enlarged DayOne's equity base. An existing investor such as GDS can therefore own a smaller percentage after the issue even if it does not sell shares in that step.
The accounting effect is easy to confuse with cash. When a new investor buys shares at a price above the carrying value implicit in the existing stake, the issue can raise the measured value of the old investor's remaining interest while diluting its percentage. GDS recognised income associated mainly with that change. DayOne received the Series C proceeds; GDS did not receive US$4.5 billion.
There was a separate cash event, and its separation makes the Q2 boundary clearer. In the first quarter, DayOne completed a US$385 million repurchase of GDS ordinary shares at the Series C issue price. GDS said the sale recycled about 95% of the principal it had invested at nearly 6.5 times money. Q1 equity-method-investee income of RMB2.136 billion arose mainly from both DayOne dilution and that partial sale.
After an April Series C upsizing, GDS said its remaining DayOne interest was worth more than US$2.2 billion at the issue price and represented about 19.9% as of 29 April. That is meaningful financial flexibility. It is still a transaction mark, not a guaranteed exit price, and it exposes GDS to DayOne's execution, financing and future ownership changes.
The first governance question is thus whether future reporting keeps three things separate: cash proceeds from actual share sales or distributions; accounting gains or losses on the retained interest; and the operating contribution, if any, that flows through equity-method results. A combined investee line cannot answer all three.
The operating receipt had thinner margins
GDS's China business was growing, but costs grew faster. Revenue increased 6.5% to RMB3.088 billion as data centres continued to ramp. Cost of revenue rose 9.6% to RMB2.424 billion. Gross profit consequently fell 3.6% to RMB664.2 million, and gross margin declined from 23.8% to 21.5%. GDS attributed the pressure mainly to utility costs taking a larger share of revenue.
Adjusted gross profit told the same directional story. It fell 0.7% to RMB1.499 billion, while adjusted gross margin declined from 52.0% to 48.5%. Adjusted EBITDA increased 2.5% to RMB1.406 billion, but its margin narrowed from 47.3% to 45.5%. Corporate cost savings offset part of the utility pressure, not all of it.
Adjusted EBITDA is useful here because GDS's definition explicitly excludes the share of results of equity-method investees. It offers a view of the operating engine without the DayOne-related line. It is not cash flow: it also excludes interest, tax, depreciation, leases and other charges that matter to a company building physical infrastructure.
The bridge exposes the tension. Operations produced RMB439.2 million of income, but net interest expense was RMB366.8 million. The quarter's tax charge of RMB213.5 million rose mainly because of an intra-group transfer undertaken in preparation for a possible second asset injection into the C-REIT. These are not reasons to dismiss reported profit. They show why an investee valuation gain and operating conversion must be read as different ledgers.
Funding must meet the delivery queue
At 30 June, GDS had RMB14.927 billion of cash and cash equivalents. Short-term debt was RMB9.210 billion and long-term debt RMB36.922 billion, or RMB46.131 billion in total by simple addition. Subtracting cash gives RMB31.204 billion, but that arithmetic should not be confused with a company-defined net-debt measure. During Q2, GDS obtained RMB4.907 billion of new debt and refinancing facilities.
The reason for the funding appetite is visible on the construction side. GDS raised 2026 capital-expenditure guidance from about RMB9 billion to about RMB10 billion. Area under construction jumped 43.9% from the first quarter to 170,355 square metres, and 89.2% of it was pre-committed. This is a substantial customer-backed development programme, not speculative floor space in the ordinary sense.
It still has a conversion clock. GDS added 59,317 square metres of net committed and pre-committed area in Q2, but only 21,307 square metres of net utilized area. Those measures do not share a same-quarter maturation window, so the 38,010-square-metre arithmetic gap is not a delivery miss. It does describe the expanding distance between commercial promise and currently occupied capacity that lenders and shareholders must fund through construction and customer ramp-up.
By quarter-end, GDS had 784,802 square metres committed or pre-committed, 684,977 square metres in service and 542,236 square metres utilized. The utilization rate was 79.2%. A high commitment rate can provide visibility, but the income and cash economics arrive only as capacity enters service, customers move in and power consumption converts into revenue at an adequate margin.
That final condition is important. More utilized area will not automatically repair profitability if utility costs keep rising faster than revenue or if financing consumes most operating income. The physical and financial queues have to clear together.
The next receipt needs two columns
GDS has created real option value from DayOne. It separated an international platform, attracted outside capital, sold part of its position at a high multiple and retained a stake marked at more than US$2.2 billion. None of that is fictitious simply because the Q2 dilution gain was non-cash.
The problem comes only when the valuation surface substitutes for the operating one. A reader looking at RMB837.6 million of net income could miss that the equity-method-investee line was larger, adjusted EBITDA margin fell, gross profit contracted, and the core build programme still depended on heavy capital and refinancing.
The next useful disclosure would therefore have two columns. The first would reconcile DayOne ownership, cash realised, carrying value, transaction marks and any recurring share of operating results. The second would reconcile GDS China commitments into construction, service, utilization, revenue, gross profit, interest and cash generation.
If both columns improve, GDS will have two engines: a valuable minority asset and a data-centre platform converting AI-led demand into returns. If the DayOne column keeps supporting the bottom line while utility, financing and ramp-up constrain the China column, headline profit will remain a poor proxy for operating progress.
Sources
Member Briefing
Deeper Profile Context
Sign in with the right membership level to unlock the full briefing and source notes.
Only for Strategic Circle
Strategic Circle
Open to all readers. Unlock profile briefings after joining and signing in.
Join Strategic CircleOnly for Leadership Alliance
Leadership Alliance
For qualified IP-asset owners and management; sign in to unlock alliance briefings.
Join Leadership Alliance

