The Equipment Layer Reprices: Vertiv Down 33% on a "Timing" Debate While Eaton Raises Guidance
Update (Sep 28, 2026): the customer-concentration question is now answered from the filings — and the answer cuts against the demand-stall reading. Vertiv's FY2025 Form 10-K (filed Feb 13, 2026) and Q2 2026 Form 10-Q (filed Jul 29, 2026) contain no customer-percentage disclosure at all: no "one customer accounted for X% of net sales" language appears in the revenue note (Note 3) or the segment note (Note 13). Under ASC 280-10-50-42, an issuer must disclose any single customer that is 10% or more of consolidated revenue — the absence of that disclosure implies no single Vertiv customer reached 10% of FY2025 net sales of $10,229.9M (i.e., none hit ~$1.02B). Concentration is addressed only qualitatively, in a risk factor:
"Large customers, such as larger communication network, cloud/hyperscale, neocloud, and colocation data center providers, comprise a material portion of our customer base and generally have greater purchasing power than smaller customers." — Vertiv FY2025 Form 10-K
So the TIKR-style thesis-breaker ("hyperscaler capital spending, made sharper by Vertiv's customer concentration") is real but unquantifiable at the named-customer level from public filings: the exposure is a broad hyperscaler/neocloud mix, not a single-name dependency above 10%.
What the filings DO show is a demand picture that contradicts the derating. Three numbers from the Q2 2026 10-Q:
- Current deferred revenue doubled in six months: $1,814.7M (Dec 31, 2025) → $3,633.7M (Jun 30, 2026) — customers are putting cash UP FRONT against future delivery. The cash-flow statement shows deferred revenue contributed +$1,822.7M to operating working capital in H1 2026; operating cash flow was $1,866.6M, up 198% YoY.
- Backlog roughly doubled in 2025: $15.0B at Dec 31, 2025 vs $7.2B at Dec 31, 2024, with "the majority… considered firm and expected to be shipped within the next 12 to 18 months" — but the 10-K warns: "Orders may be subject to cancellation or rescheduling by the customer," and customers "have the right in some circumstances, usually with penalties or other termination consequences, to reduce or defer firm orders in backlog."
- Q2 2026 revenue $3,274.3M (+24.1% YoY), net income $497.8M (+53.5%), and FY2026 capex guidance was raised to $550–570M (from $425–525M in the 10-K) to build more capacity.
The balance sheet is the opposite of the leveraged players in this map. In March 2026 Vertiv issued $2.1B of senior UNSECURED notes — 4.850% due 2036, 5.650% due 2046, 5.800% due 2056, and 5.950% due 2066 — repaid its entire $2,076.1M term loan, and replaced its secured ABL with a $2.5B unsecured revolver (4.00x max net leverage covenant). The 10-Q states: "We do not have any guarantees or other off-balance sheet financing arrangements, including variable interest entities." Per the /markets/VRT/2026/09/28 view: total debt $3.26B vs cash $2.81B, debt/EBITDA 1.2x, Q2 EPS $1.52 beat vs $1.43 est. A bond market that sells a data-center equipment maker 40-year unsecured paper at 5.95% is not pricing credit stress in the equipment layer.
Where this leaves the timing-vs-demand debate: the stock at $253.28 is -32.6% from its 52-week high (3-month -18.6%) on a 57.2 P/E — the derating is a valuation/timing event, not (yet) a filing-level demand event. The bear case survives in exactly two places: (1) the qualitative purchasing-power/penalty-laden-contract risk with hyperscalers, and (2) the backlog's cancellation rights. The Oct 28 Q3 print — revenue vs the $3.75B midpoint and the deferred-revenue trajectory — remains the settler; a second miss would convert "timing" into pattern, but as of the June 30 filings the cash was flowing in the opposite direction of a stall.