Forgent Power Solutions, Inc. (NYSE: FPS) — Old Switchgear in a New AI Wrapper, Priced Richer Than Vertiv While Its Sponsor Sells

⚡ Claude’s Take

This block is the author’s own independent opinion and general information only — not investment advice. The analysis that follows takes no position and carries no price target; only this section expresses a view.

Verdict: AVOID at $55 / not-a-short / accumulate only in the low-to-mid $30s. Conviction: medium. A genuinely good demand story wrapped around a two-year-old, sponsor-controlled, leveraged roll-up — and priced richer than Vertiv, the gold-standard data-center pure-play, on every multiple that matters.

Forgent is a real business riding a real boom. It builds the switchgear, transformers, eHouses and power skids that data centers, power producers and utilities cannot build their facilities without; demand is genuinely bottlenecked; backlog is up 157% to ~$2.0 billion; revenue nearly doubled year-on-year; and Adjusted EBITDA margins (~22%) are respectable for the category. None of that is the problem. The problem is the price and the setup. At $55.13 the equity (all 304 million economic units) is worth ~$16.8 billion and the enterprise ~$17.3 billion — about 14x trailing sales and ~68x trailing Adjusted EBITDA. Vertiv, a far larger, cleaner, longer-public franchise, trades at 6.2x sales and 28.7x EBITDA; nVent at 4.6x/21.9x; Powell Industries, the closest listed switchgear pure-play, at 2.9x/14.2x. You are being asked to pay more than double Vertiv’s revenue multiple for a company that has been assembled out of legacy electrical shops since October 2023, has been public for four-and-a-half months, carries ~$490 million of net debt and deeply negative tangible equity, owes its sponsor 85% of future tax savings through a Tax Receivable Agreement, is a “controlled company” run for Neos Partners, and whose sponsor just sold $1.3 billion of stock in a June secondary at $47 — below today’s price. The entire valuation rests on flawlessly filling the ~$5 billion of revenue capacity management says it has built. That can happen; it is also exactly the kind of bottleneck-driven, capacity-shortage tailwind (cf. the turbine and transformer shortages of this cycle) that ends when supply catches up.

Framing: a momentum/theme IPO in distribution. This is not a falling knife and not a deep-value name; it is a freshly-public secular-growth story trading on the AI-power narrative, with the founding owner actively monetizing. The tape is a near-vertical post-IPO advance ($27 → $65 in four months) now rolling over ~16% off the high as the sponsor sells into it. I would not short it — the demand is real, the float is small, short borrow on a controlled name is dangerous, and a clean quarter could squeeze it. But I would not pay 68x EBITDA for a two-year-old roll-up either. Conviction medium. Flips bullish if it sustains 20%+ organic growth toward the $5B capacity at 22%+ Adjusted EBITDA margins and the multiple resets to a Vertiv-like ~25-30x (i.e., much higher earnings into a lower multiple). Flips bearish if backlog conversion slips, a capacity glut compresses pricing, or the Up-C/TRA cash leakage proves as “substantial” as the prospectus warns. Catchy tag: the switchgear is real; the multiple is science fiction, and the sponsor is selling it to you.

📈 Stock Price Action — Five-Year Event Map

Forgent has been public for only ~4.5 months, so there is no five-year history — the “event map” is the post-IPO record from February 5, 2026 to date. The arc is a classic hot-IPO round trip in miniature: priced at $27.00, a quiet first six weeks, then a near-vertical AI-power re-rating to an all-time intraday high of $65.56 on June 4, 2026, and a ~16% pullback to $55.13 (June 26, 2026) as the sponsor’s $47 secondary and a broad tech sell-off cooled the move. The 52-week (life-of-stock) range is ~$25.95–$65.56; the stock sits ~16% below its high and ~104% above its IPO price. Price moves are Fact; attributed drivers are Interpretation.

Cycle narrative: (1–2) The IPO priced at $27 and traded sideways-to-down for six weeks, a normal post-IPO lull as the small public float found hands. (3–4) From April the stock caught the AI-electrification bid that has carried Vertiv, GE Vernova and the transformer names, roughly doubling into a June 4 high of $65.56. (5) Into that strength, Neos brought a ~$2.0 billion secondary at $47.00 — two-thirds of it the sponsor selling — a textbook distribution into momentum. (6) A TD Cowen price-target raise to $73 (June 22) briefly supported the tape. (7) The stock then fell ~16% off its high into June 26 ($55.13, −6.6% that day on heavy volume) as the AI-infrastructure complex sold off and the secondary supply weighed. Each move ties to a dated public event.

1. Executive Summary

Forgent Power Solutions is a U.S. designer and manufacturer of electrical distribution equipment — switchgear, dry-type and liquid-filled transformers, power distribution units (PDUs), electrical houses (“eHouses”), power skids, automatic transfer switches, panelboards, switchboards and tap boxes — sold into three deliberately-chosen end markets: Data Centers (42% of FY25 revenue), the electric Grid (23%), and energy-intensive Industrial (19%), with the balance (“Other,” 16%) in commercial and miscellaneous. The company specializes in engineered-to-order Custom Products (78% of FY25 revenue) and integrated Powertrain Solutions (13%) — i.e., the full electrical “powertrain” that moves power from the utility feed to the racks inside a data center or the lines inside a factory — competing on lead time, customization and breadth rather than price.

The business did not exist in its current form before late 2023. It is a roll-up assembled by private-equity sponsor Neos Partners beginning with an October 31, 2023 acquisition and continuing through a series of bolt-ons (the MGM, PwrQ, States and VanTran transactions). It IPO’d on February 5, 2026 at $27.00 a share through an Up-C structure that keeps Neos in control (a “controlled company”), gives Neos super-economics via a Tax Receivable Agreement that captures 85% of future tax savings, and leaves the public holding Class A stock in a holding company whose only asset is ~80% of the operating LLC.

The operating momentum is real and impressive. Revenue grew 56% from a pro-forma $482.7 million in FY24 to $753.2 million in FY25, and accelerated to $958.4 million in the nine months ended March 31, 2026 (Q3 alone $378.7 million, +103% year-on-year). Backlog reached ~$2.0 billion at March 31, 2026, up 157%. Management says installed capacity can support up to $5 billion of annual revenue with little further capital. Adjusted EBITDA margins run ~22%, comparable to Vertiv and nVent. On a GAAP basis the company is barely profitable (FY25 net loss of $6.2 million) because ~$53 million of annual interest on the roll-up’s debt consumes operating income.

The investment tension is entirely about price and structure, not demand. At $55.13 the enterprise is valued at ~$17.3 billion — ~14x trailing sales and ~68x trailing Adjusted EBITDA — a premium to every comparable, including Vertiv (6.2x/28.7x). The moat is narrow and cyclical (a capacity bottleneck that, like all shortages, attracts supply), the balance sheet carries ~$490 million of net debt and ~-$370 million of tangible common equity, the governance is sponsor-controlled with a TRA cash drain, and the sponsor itself sold $1.3 billion of stock at $47 in June — below the current price. This memo takes no position and sets no target; it lays out what the market is underwriting (essentially a flawless fill of the $5B capacity at sustained 20%+ margins) and what would falsify the bull and bear cases.

2. Business Overview

What it makes. Forgent manufactures the heavy electrical apparatus that sits between the utility connection and the equipment it powers. Its product catalog spans: low-, medium- and paralleling switchgear; transformers (low-voltage, medium-voltage VPI, padmount, PDU and substation, both dry-type and liquid-filled); power distribution units (PDUs); electrical houses (“eHouses”) and UPS eHouses (pre-fabricated, walk-in enclosures containing integrated electrical line-ups); power skids; automatic transfer switches (ATS); generator connection cabinets; panelboards; remote power panels; switchboards; and tap boxes. The company also sells aftermarket and field services — maintenance, testing, repairs, modernization, start-up and commissioning. In FY25 no single product category exceeded 13% of revenue, a deliberately diversified line-up.

How it sells — three product modes. Forgent splits revenue into Standard Products (~5% of FY25), Custom Products (~78%), Powertrain Solutions (~13%) and services (~4%). Standard Products are common designs made in volume for basic applications. Custom Products are “engineered-to-order” for a specific project, involving significant back-and-forth between Forgent’s in-house engineering team and the customer; the company produces more than 1,500 unique designs per year at an average “batch count” of 15 (i.e., on average only 15 units per design). Powertrain Solutions integrate multiple Custom Products — skidded, housed in an eHouse, or designed to operate as a system — into a turnkey electrical powertrain. This engineered-to-order, low-batch, integration-heavy mix is the heart of the business model: it is the part that commands lead-time-driven pricing and that overseas commodity competitors struggle to serve.

Who buys it. Customers include technology/hyperscale companies, power generators (independent power producers), utilities and industrial operators buying directly; OEMs and integrators who embed Forgent’s products into larger systems; EPC contractors building data centers, power plants and transmission/distribution (T&D) infrastructure; and electrical distributors. End-market mix for FY25: Data Center 42%, Grid 23%, Industrial 19%, Other 16% — so 65% from Data Center + Grid (the two structurally fastest, least cyclical markets) and 84% from the three target markets combined. Substantially all FY25 revenue was North American; international is an explicit (but nascent) growth ambition.

Revenue nature — project, not recurring. This is a project-based capital-equipment business, not a subscription or razor/blade model. Revenue is recognized as engineered-to-order projects are built and shipped (a meaningful portion over time, given customization and deferred-revenue/customer-advance dynamics visible on the balance sheet — ~$134 million of short-term deferred revenue at March 31, 2026). Aftermarket services (~4%) are the only genuinely recurring slice. Visibility comes from backlog (~$2.0 billion) rather than contracted recurring revenue; backlog converts as projects are scheduled and delivered, and is exposed to cancellation, push-out and margin re-estimation. The economic driver is therefore the capital-spending cycle of hyperscalers, IPPs and utilities — currently in a historic up-leg, but a cycle nonetheless.

Footprint. Ten manufacturing campuses across five locations — Minnesota (HQ, Dayton), Texas, Maryland, California and Mexico — with ~2,400 full-time employees as of March 31, 2026. Management states it has nearly completed a capacity-expansion program (begun in 2023) that can more than triple FY25 production by end of CY2026 and support up to $5 billion of annual revenue without significant further investment — a critical claim, because the entire bull case is operating leverage on that fixed footprint.

Verdict: A coherent, well-positioned, engineered-to-order electrical-equipment maker aimed squarely at the fastest-growing power-infrastructure end markets, with a sensibly diversified product line and a credible capacity runway. It is a real industrial business — but a project/cyclical one assembled by acquisition two years ago, not a recurring-revenue compounder, and its quality must be judged against the price (–) and the leverage/structure it carries.

3. Industry Dynamics

Structure and demand. Electrical distribution equipment is a mandatory, safety-critical input: every power plant, substation, data center, factory and commercial building needs it, and because failure carries catastrophic consequences (downtime, fire, injury), buyers prioritize reliability and lead time over price. The market has historically grown with non-residential construction and grid investment — low-to-mid single digits in a normal cycle — but Forgent’s three chosen sub-markets are growing far faster. Per the prospectus (citing Omdia, Wood Mackenzie, BCE and Dodge), 2020–2025 CAGRs were ~27% for new data centers, ~15% for power plants, ~11% for T&D, and ~17% for manufacturing plants, versus ~9% for overall non-residential investment. Three secular forces drive this: (i) AI/cloud data-center buildout (the dominant driver of Forgent’s mix); (ii) electricity-demand growth and new generation (IPPs adding gas, solar, storage; reshoring factories); and (iii) grid replacement — NREL/ASCE estimate the majority of in-service grid distribution equipment is >30 years old against a 25–40 year useful life.

The bottleneck — and its capital-cycle hazard. The reason Forgent’s pricing and margins are elevated is that electrical distribution equipment is currently a bottleneck: lead times for transformers and switchgear have stretched dramatically across the industry over 2023–2026, and customers will pay up and book capacity early to secure delivery. This is genuinely favorable today. But through a Marathon “Capital Returns” lens it is precisely the configuration that invites supply and mean-reverts: abnormally high returns and shortage pricing draw capacity from incumbents (Eaton, Schneider, ABB, Siemens, Hitachi Energy, Vertiv, nVent, Powell) and new entrants. Transformer and switchgear capacity is expanding industry-wide. When lead times normalize — historically the back half of every electrical-equipment up-cycle — the shortage premium in price and margin compresses, and a heavily-fixed-cost manufacturer that built capacity for $5 billion of revenue faces operating deleverage if demand merely plateaus. The current shortage is the friend; the resolution of the shortage is the risk.

Competitive intensity. The industry is bifurcated: a handful of large, well-capitalized global players (Eaton, Schneider Electric, ABB, Siemens, Hitachi Energy, plus data-center-focused Vertiv and electrical-protection nVent) with vastly greater R&D, balance-sheet and channel resources; and a long tail of regional and local private fabricators competing project-by-project. Forgent positions itself in the middle: large enough to do the whole powertrain end-to-end and hold capacity, small and nimble enough to customize and turn quickly. It names Vertiv, Hitachi Energy and nVent among competitors and concedes it competes with “a number of smaller private companies in all of the markets that we serve,” and even occasionally with its own OEM/integrator customers. Competition is on performance, reliability, warranty, lead time, customization and price. The big players can — and are — adding data-center-specific capacity; the small players compete on price in commoditized categories.

Regulation. Largely favorable-to-neutral: transformer efficiency standards (DOE) periodically raise the cost of compliant designs (a pass-through risk if Forgent cannot reprice), and grid interconnection delays can slow customer projects (a demand-timing risk). Tariffs cut both ways — they pressure input costs (electrical steel, copper, aluminum) but also encourage the reshoring of manufacturing that drives Industrial demand and disadvantage overseas equipment competitors.

Verdict: A structurally attractive end-market today — mandatory product, accelerating demand, real bottleneck pricing — but a cyclical, capital-cycle-sensitive industry whose current super-normal economics are a function of a temporary capacity shortage that the entire industry is racing to fill. Good place to be in 2026; the durability of the economics past the shortage is the open question, and it is the question the valuation ignores.

4. Competitive Position

The claimed moat. Forgent’s differentiation, in its own words, is being “one of only a small number of companies that can engineer and manufacture all of the electrical distribution equipment required for a data center or large manufacturing facility’s powertrain… with some of the highest levels of customization and shortest lead times available.” Three reinforcing claims: (1) breadth — it makes most of what goes into an eHouse or power skid in-house, so it can sell the customer the whole powertrain from one vendor; (2) lead time — its capacity and U.S. footprint let it quote shorter deliveries than capacity-constrained competitors or overseas suppliers; (3) customization/application engineering — late-stage, evolving specs favor a vendor with local engineers close to the customer, which structurally disadvantages overseas commodity manufacturers (no local engineering, can’t hold custom inventory, weeks of ocean freight).

Pressure-testing it (Greenwald lens). In the Greenwald & Kahn taxonomy, durable advantage comes from supply-side cost advantage, demand-side customer captivity, or economies of scale combined with captivity. Forgent’s edge is best described as a modest customer-captivity-plus-lead-time advantage that is currently amplified by a cyclical capacity shortage — not a structural moat:

• Switching costs / captivity: Real but limited. Once a customer specs Forgent into a data-center design and qualifies its equipment, re-specifying mid-project is costly, and the application-engineering relationship has stickiness. But these are project-level, not installed-base, switching costs — each new project is re-competed, and there is no recurring software/consumables lock-in. This is weaker captivity than, say, a control-system or building-management installed base.
• Lead time: A genuine advantage while the industry is capacity-short. It is, however, a function of the shortage, not a permanent structural edge — when peers’ new capacity comes online, the lead-time gap narrows. It is rented from the cycle.
• Scale: Forgent is sub-scale versus Eaton, Schneider, ABB, Siemens and even Vertiv. Its ~$1.2 billion of revenue is a fraction of these players’ electrical franchises; it cannot out-spend them on R&D or out-purchase them on raw materials. Its “scale” claim is relative to small private shops, not the majors.
• Breadth / one-stop powertrain: The most defensible claim — few players self-manufacture the full line-up and integrate it. But the majors can assemble the same scope, and integrators can source components from multiple vendors. It is a real commercial advantage, not an unassailable barrier.

Versus the comp set. Against Vertiv (data-center power/cooling/thermal, ~$10B revenue, deep hyperscaler relationships and a services annuity), Forgent is a narrower, smaller, equipment-only supplier. Against Powell Industries (the closest listed pure-play, ~$1.1B revenue, custom switchgear/eHouses for energy and industrial, net cash), Forgent is similarly-sized and arguably more data-center-levered but far more financially levered. Against nVent (electrical connection/protection/enclosures), it overlaps in enclosures and data-center power. The majors (Eaton, Schneider, ABB, Siemens, Hitachi Energy) are bigger in every dimension. Forgent’s right to win is speed and customization in a shortage, which is precisely the advantage most exposed to mean reversion.

The moat-to-financials test. A moat must show up as a financial outcome that would deteriorate without it. Forgent’s ~22% Adjusted EBITDA margin and shortage pricing are the candidate. But the margin is in line with — not above — Vertiv (~21.5%), nVent (~21%) and Powell (~20%), so there is no evidence of super-normal moat economics; Forgent earns category-average margins. And on full invested capital (including the ~$817 million of goodwill and intangibles the roll-up paid), returns are only ~WACC. The high incremental returns are on tangible operating assets — typical of an engineered-to-order manufacturer — but the price paid to assemble the company captured most of that value for the sellers. By the playbook’s own standard, this is a narrow, cycle-amplified competitive position, not a durable moat.

Verdict: A real but narrow competitive position — breadth + lead time + application engineering — that is currently flattered by a capacity shortage and earns category-average (not super-normal) margins. Durable enough to be a good operator; not durable enough to justify a premium to the best-positioned, larger, cleaner names in its own peer group. Crowded market, genuine-but-rentable differentiation.

5. Growth History and Forward Opportunities

History — explosive, but acquisition-built and short. Forgent’s reported history is only ~2.5 years and is fractured by Predecessor/Successor acquisition accounting. On a comparable pro-forma basis, revenue grew 56% from $482.7 million (pro-forma FY24) to $753.2 million (FY25), then accelerated to $958.4 million in 9M FY26 (vs. ~$515.6 million in 9M FY25, +86%), with Q3 FY26 revenue of $378.7 million up ~103% year-on-year. This is a blend of (i) organic volume from the data-center/grid boom, (ii) price/lead-time premium, and (iii) the annualization and integration of the MGM/PwrQ/States/VanTran acquisitions. The ~$2.0 billion backlog (+157%) points to continued near-term growth; quarterly revenue has stepped up sequentially every quarter through FY26.

The forward opportunity — fill the $5B capacity. Management’s stated runway: capacity additions (largely complete) can support >3x FY25 production by end CY2026 and up to $5 billion of annual revenue with little incremental capital. The growth levers: (1) data-center capacity additions as hyperscalers and neoclouds keep building (the secular driver); (2) share gain from capacity-constrained competitors — explicitly a thesis (“take share from competitors that are capacity constrained”); (3) grid replacement super-cycle; (4) reshoring-driven Industrial demand; (5) further M&A (the roll-up is explicitly ongoing — “acquire companies that increase our scale”); and (6) international expansion (nascent, from a ~0% base). Pricing/lead-time tailwinds add to volume growth while the shortage persists.

Quality of the growth. Mixed. The demand is high-quality (secular, mandatory product, fast-growing markets), and incremental margins on the fixed footprint should be attractive if volumes fill capacity. But the growth is (a) cyclical and bottleneck-dependent — a portion of both volume and price is shortage-driven; (b) partly acquired, with the attendant integration and purchase-accounting noise; and © lumpy/project-based, dependent on large data-center awards (one >10% customer in Q3 FY26) and exposed to backlog cancellation/push-out. It is not the steady, recurring, high-visibility growth that would warrant a software-like multiple. The bull case requires the secular demand to outrun the industry’s capacity response for several more years — plausible, but not assured, and fully priced.

Verdict: High-rate growth of mixed quality. The trajectory is genuinely impressive and the near-term backlog supports it, but it is cyclical, partly acquired, project-lumpy and shortage-amplified — high-octane, not high-durability. The valuation treats it as the latter.

6. Financial Quality

Income statement. FY25: revenue $753.2 million, gross profit $278.1 million (36.9% gross margin), GAAP operating income $36.6 million (4.9% margin, depressed by a Q4 FY25 charge), and a GAAP net loss of $6.2 million — the loss driven entirely by $52.8 million of interest expense on the roll-up’s debt. 9M FY26 shows the operating inflection: revenue $958.4 million, gross margin ~34.6% (a modest compression from FY25 as mix/scale shifts), operating income building to $39.3 million in Q3 alone, and — critically — interest expense falling to ~$10.8 million in Q3 (from ~$21.0 million in Q2) as IPO proceeds and the December 2025 refinancing reduced the cost of debt, turning the company GAAP-profitable (Q3 net income $18.3 million, aided by a ~$12.4 million favorable tax/discrete item). On an adjusted basis: Adjusted EBITDA $169.2 million in FY25 (22.5% margin) and $210.2 million in 9M FY26 (21.9%); TTM (Apr-2025–Mar-2026) Adjusted EBITDA ~$253 million and TTM Adjusted Net Income ~$150 million. Adjusted EBITDA margins are healthy and category-competitive; the gap between GAAP (~breakeven) and adjusted (~$150M net) is mostly interest, depreciation/amortization of acquired intangibles, and one-time/SBC add-backs — normal for a recently-levered roll-up, but a reminder that GAAP earnings power is currently thin and leverage-sensitive.

Margins and operating leverage. Gross margin ~35–37% and Adjusted EBITDA margin ~22% sit in line with Vertiv/nVent/Powell. The bull case rests on operating leverage: with capacity for ~$5 billion of revenue largely built, incremental volume should drop through at high contribution margins, lifting Adjusted EBITDA margin toward the mid-20s and beyond. The evidence is suggestive but not yet conclusive — 9M FY26 Adjusted EBITDA margin (21.9%) is actually below 9M FY25 (24.5%), as mix shift, the cost of scaling, and slight gross-margin compression offset volume leverage. The thesis that margins expand with scale is unproven in the public record so far.

Cash flow. Here the quality concerns sharpen. FY25 operating cash flow was only $45.0 million against $101.4 million of reported EBITDA — working capital consumed $34 million (receivables +$78.5M, inventory +$34.5M) as the business scaled, and capex of $84.1 million (capacity build) produced negative free cash flow of ~-$39 million. FY24 OCF was negative. So Forgent has been a cash consumer, funding growth capex and working capital with debt and IPO proceeds — typical of a fast-scaling capital-equipment roll-up, but a long way from the “cash machine” implied by the multiple. FCF should inflect positive as the capacity build completes (management says major capex is largely behind it) and if working capital normalizes — but that is a forward assumption, not a demonstrated result. Net income is also diverging favorably from cash in FY25 (EBITDA >> OCF), the opposite of the usual red flag but a sign of heavy working-capital and capex intensity during the build.

Balance sheet — the structural weak point. At March 31, 2026: cash $93.8 million; total debt ~$584 million (plus ~$123 million of capital/finance leases); net debt ~$490 million (~$613 million including leases). Net debt/Adjusted EBITDA is ~1.9x ex-leases — not extreme, and improved by the IPO — but the equity is thin: total equity $582.6 million includes $136 million of minority interest, and against $516.6 million of goodwill + $300.3 million of intangibles ($817 million), tangible common equity is roughly -$370 million (TCE ratio -22.6%). The June secondary’s primary proceeds (~$628 million) were used to redeem Opco units from the sponsor’s affiliates, not to pay down debt — so leverage is largely unchanged post-secondary, and the June 23 credit-agreement amendment re-set the facility terms. Liquidity is adequate (current ratio ~1.6x, revolver availability), but this is a goodwill-heavy, negative-tangible-equity, leveraged balance sheet — the financial profile of a young LBO, not a fortress.

ROIC/ROE. Aggregator ROE/ROIC figures are distorted (thin/negative tangible equity, minority interest, discrete tax items). Reconstructed: on full invested capital (~$1.0–1.1 billion of net debt + equity, before adding back the goodwill the roll-up paid), adjusted NOPAT of ~$130–160 million implies ROIC in the low-double-digits — roughly at or modestly above WACC. On tangible operating capital the returns are high (asset-efficient engineered-to-order manufacturing), but the ~$817 million of goodwill/intangibles paid to assemble the company means the owner’s return on the price paid is ~WACC — the Greenwald signature of a roll-up where the sellers captured the value. There is no super-normal return on capital here.

Verdict: Economics are decent but do not yet visibly improve with scale, and the cash and balance-sheet quality is weak. Healthy category-level adjusted margins; thin GAAP earnings; negative-to-just-turning free cash flow; ~$490M net debt and ~-$370M tangible equity; ROIC ~WACC on the capital deployed. This is a financially-fine, leverage-sensitive young roll-up — not a high-return compounder, and certainly not one whose demonstrated economics justify a 68x EBITDA multiple.

7. Capital Allocation

The track record is short and sponsor-driven. Forgent has existed for ~2.5 years and has been public for ~4.5 months, so the capital-allocation record is essentially the Neos roll-up program plus an IPO and a secondary. The judgments available:

• Acquisitions (the core activity). Neos assembled Forgent by buying electrical-equipment manufacturers (MGM, PwrQ, States, VanTran and the founding business) from late 2023, funded with ~$517 million of debt and ~$436 million of equity in FY24, deploying ~$742 million on acquisitions that year. The multiples paid are not cleanly disclosed but the result — $817 million of goodwill and intangibles — confirms meaningful prices. Integration risk is live and ongoing; the prospectus flags acquisitions prominently as a continuing strategy (“acquire companies that increase our scale”). The roll-up “worked” in the sense that it created a $1.2B-revenue platform riding a boom; whether it created value per share depends entirely on the exit multiple, which is the public’s risk now.
• Capacity capex. ~$84 million in FY25 to build toward $5 billion of revenue capacity. If demand fills it, this is excellent capital allocation (operating leverage on a fixed base); if demand plateaus, it is stranded fixed cost. The bet is reasonable given backlog, but it is a bet.
• IPO (Feb 2026) and secondary (June 2026). The IPO raised $1.51 billion gross; the structure (Up-C + TRA) was engineered to maximize sponsor economics. The June secondary was ~$2.0 billion, two-thirds of which (~$1.30 billion) was the sponsor selling, with the ~$628 million primary used to buy out more sponsor-affiliated Opco units rather than de-lever or fund growth. In substance, the recent “capital allocation” has been returning capital to Neos, not reinvesting in the business or strengthening the balance sheet.

Incentives and structure — the red flags. This is a “controlled company” run for Neos: dual-class (Class B carries votes but no economics), Neos retains super-rights (board, special-meeting, written-consent and consent rights over >$100M actions so long as it holds 35%), a classified board, and DGCL-203 protections that do not apply to Neos. Most consequentially, the Tax Receivable Agreement obligates Forgent to pay the TRA participants 85% of the cash tax savings realized from the Up-C step-ups and future unit exchanges — a payment stream the company itself calls “substantial” and expects to run for more than fifteen years after the sponsor fully exchanges. That is a structural, long-dated diversion of cash from public shareholders to the sponsor that does not appear in EV or in EBITDA, and it materially reduces the cash that public holders ultimately receive per dollar of tax benefit.

Dividends/buybacks. A nominal dividend has been paid (largely a pre-IPO/structuring artifact; FY25 dividends ~$13 million); there is no established public-company return program, and given the leverage, negative tangible equity and TRA obligations, capital return to common holders should be a low priority for years.

Insider behavior. The dominant signal is the sponsor selling: $1.3 billion in the June secondary at $47, into strength, four months after the IPO. There is no evidence of open-market insider buying. For a freshly-public, sponsor-controlled name, aggressive early distribution by the controlling owner is a meaningful negative tell on how insiders view the risk/reward at these levels.

Verdict: Too short to fully judge, but what is visible tilts negative for public shareholders. The roll-up was competently assembled and the capacity bet is rational, but the recent capital flows have run to the sponsor (secondary + Opco redemptions), the governance is sponsor-controlled, and the TRA is a long-dated cash drain. Management’s operating pedigree (ex-Vertiv/Caterpillar/Schneider) is strong; the capital-allocation alignment with minority holders is weak. Net: a competent operator inside a structure built to favor Neos.

8. Changes and Headwinds — Last Two Years

Because Forgent is ~2.5 years old, “the last two years” is essentially its entire life:

• Formation and roll-up (Oct 2023 – 2025): Neos acquired the founding business (Oct 31, 2023, “Inception” Sept 8, 2023) and bolted on MGM, PwrQ, States and VanTran, creating the platform. Predecessor/Successor accounting splits the early periods.
• Capacity build (2023–CY2026): A multi-site expansion (ten campuses, five locations) to support up to $5 billion of revenue, now nearly complete — the key enabler of the forward thesis.
• Demand inflection (FY25–FY26): Revenue +56% (FY25) then +86% (9M FY26); backlog +157% to ~$2.0 billion — the AI/data-center electrification wave hitting the order book.
• December 2025 refinancing: A new/amended credit agreement (Dec 19, 2025) that lowered interest cost (visible in the Q3 FY26 interest drop), followed by Amendment No. 1 on June 23, 2026 (8-K) re-setting facility terms.
• IPO (Feb 5, 2026) at $27.00 — Up-C structure, TRA, controlled-company governance established.
• Secondary (June 1, 2026) at $47.00 — ~$2.0 billion, two-thirds sponsor selling; the defining recent event and the source of current technical overhang.
• Sell-side initiation/PT raise: TD Cowen Buy, PT raised to $73 (June 22, 2026).

Headwinds/risks emerging: (i) the secondary overhang and likely further sponsor selling as lock-ups roll off; (ii) input-cost inflation (electrical steel, copper, aluminum) and tariffs pressuring margins if not repriced; (iii) AI-capex sentiment — the stock now trades as an AI-infrastructure proxy and fell ~16% off its high into late June on a tech sell-off; (iv) backlog conversion risk; (v) transformer efficiency-standard compliance costs; (vi) utility interconnection delays slowing customer projects. None has yet impaired the fundamentals, but the stock’s beta to AI-infrastructure sentiment is now high.

Verdict: The two-year record is one of rapid, sponsor-engineered build-out into a favorable demand wave — fundamentally strengthening the operating story while structurally (leverage, TRA, sponsor distribution) and technically (overhang, AI-sentiment beta) adding risk. On balance the changes have made FPS a bigger, faster-growing, but also richer and more sentiment-driven security.

9. Risk Analysis (Risk Matrix)

Net: The dominant risk is valuation — a premium-to-Vertiv multiple on a younger, levered, sponsor-controlled name leaves little margin for error — compounded by the cyclical/shortage nature of the margins and the sponsor-overhang/structure overlay. Catastrophic loss is unlikely; a sharp de-rating is not.

10. Valuation Discussion (Embedded Expectations)

Where it trades. At $55.13 (June 26, 2026) on 304.4 million total economic units (244.1M Class A + 60.3M Class B/Opco), equity value is ~$16.78 billion; adding ~$490 million net debt gives EV ~$17.27 billion (~$17.4 billion including finance leases, and before the off-balance-sheet TRA claim). Against trailing fundamentals:

The table is the thesis. Forgent is priced at more than double Vertiv’s revenue multiple and ~2.4x Vertiv’s EBITDA multiple, and ~5x Powell’s EBITDA multiple, despite earning the same ~22% margins, being smaller, younger, more levered, sponsor-controlled, and burdened by a TRA. Vertiv is the most generously-valued large, established, pure-play data-center power franchise in the public market — and FPS trades at a large premium to it.

Embedded expectations — what the $55 price requires. Reverse-engineering the multiple: to justify ~$17.3 billion of EV at a “normal” premium-industrial multiple, Forgent must deliver the $5 billion revenue capacity at sustained mid-20s% Adjusted EBITDA margins. At $5 billion × 24% = ~$1.2 billion of Adjusted EBITDA, the current EV would be ~14x that future number — reasonable only if (a) the $5B is filled (roughly 4x current revenue), (b) margins expand and hold despite the shortage resolving, © it happens within a few years, and (d) the market still pays a premium multiple at that scale. In other words, today’s price already capitalizes a near-flawless, multi-year execution of the entire capacity plan. On trailing numbers (~68x EBITDA) there is essentially no current-cash-flow support; the valuation is a pure forward/option-on-execution.

Scenario analysis (illustrative; not a target). Holding ~304M units and ~$490M net debt:

• Bear (~$25–32): Shortage resolves, growth decelerates to ~10–15%, margins compress to high-teens, and the multiple re-rates toward Powell/nVent (3–5x sales, 14–20x EBITDA). FY27 revenue ~$1.4B × ~20% = ~$280M EBITDA at ~18x → EV ~$5.0B → equity ~$4.5B → ~$15/unit; a softer de-rate to ~6x sales/~22x on ~$1.5B revenue → ~$28–32/unit. This range simply marks FPS to a Vertiv-to-nVent multiple on near-term numbers.
• Base (~$38–50): Growth stays strong (revenue to ~$1.6–1.8B in FY27) at ~22–23% margins (~$370–410M Adjusted EBITDA), and the multiple compresses to a still-premium ~25–32x EBITDA / ~6–8x sales → EV ~$10–13B → equity ~$9.5–12.5B → ~$31–41/unit, with the high end (~$45–50) if it sustains a Vertiv-like ~30x on rising estimates.
• Bull (~$70–95+): The $5B capacity fills faster than the industry adds supply; margins push to mid-20s (~$1.1–1.2B Adjusted EBITDA at scale); the market keeps paying ~18–20x that forward number → EV ~$20–24B → equity ~$19.5–23.5B → ~$64–77/unit (TD Cowen’s $73 sits here), with blue-sky above $90 if it becomes “the next Vertiv.”

The asymmetry is unfavorable at $55: the base case clusters below the current price, the bull case roughly matches the sell-side target, and the bear case is a long way down. The market is underwriting the bull path as the base case.

What the market is pricing correctly vs. incorrectly. Correctly: the demand is real, the backlog is real, the capacity is real, and the margins are real. Incorrectly (in my read): it is extrapolating shortage-era growth and pricing as durable, ignoring the cyclicality of the bottleneck, under-weighting the TRA/Up-C cash leakage and sponsor distribution, and paying a premium to better-positioned, cleaner, larger peers for a four-month-public, sponsor-controlled roll-up. No price target; no recommendation — but the embedded expectations leave little room for anything short of flawless execution.

11. Variant Perception

Consensus belief. Forgent is “the next Vertiv” — a pure-play picks-and-shovels winner of the AI-power buildout, with a long runway to fill $5B of capacity, category-leading growth, ex-Vertiv management, and a backlog that de-risks the next several years. Sell-side is positive (TD Cowen Buy, $73). The stock’s ~104% post-IPO advance reflects this consensus.

Strongest bull case. The electrification of compute is a multi-year, possibly decade-long, super-cycle; electrical distribution equipment is a hard bottleneck; Forgent has already built the capacity (so it can capture share without dilutive capex) and has the engineering and breadth to win whole-powertrain awards. If it fills $5B at mid-20s margins, today’s price is ~14x a future EBITDA number — and a true “next Vertiv” could command more. Management has done exactly this once before (building Vertiv out of Emerson Network Power). Backlog +157% and Q3 revenue +103% say the demand is here now.

Strongest bear case. You are paying a premium to Vertiv for a two-year-old, leveraged, sponsor-controlled roll-up with category-average margins, ~WACC returns on invested capital, negative tangible equity, negative-to-just-turning free cash flow, a TRA siphoning 85% of tax savings to the sponsor for 15+ years, and a controlling owner who just sold $1.3 billion of stock at $47 — below the current price. The margins and growth are shortage-amplified and mean-reverting; the industry is racing to add the very capacity that would erode Forgent’s lead-time edge; and the valuation requires flawless multi-year execution that no public track record yet supports. A multiple even halfway to Vertiv’s halves the stock.

The 3–5 assumptions that matter most:

1. Does the AI/data-center capex super-cycle persist long enough to fill $5B of capacity? (Demand durability.)
2. Do margins hold/expand as the industry’s capacity shortage resolves? (Pricing durability — the crux.)
3. Does the multiple stay near or above Vertiv’s, or re-rate toward Powell/nVent? (Valuation — the dominant driver of returns from here.)
4. How much cash does the Up-C/TRA structure divert from common holders, and how aggressively does Neos keep selling? (Structure/overhang.)
5. Can a four-month-public roll-up execute the capacity fill without integration, working-capital or quality stumbles? (Execution.)

Falsifying evidence. Bull falsified by: decelerating order intake/backlog, Adjusted EBITDA margins compressing below ~20%, lengthening peer lead times (shortage resolving), or further large sponsor sales. Bear falsified by: sustained 20%+ organic growth with margins expanding toward mid-20s, FCF turning solidly positive, Neos’s stake stabilizing, and the company demonstrably winning whole-powertrain hyperscaler awards at premium prices into 2027.

The factor/positioning read (FactorsToday returns no data — the stock lacks the ≥252-day history its model requires — so this is built from the AZI price tape only): FPS trades as a high-beta, momentum/AI-infrastructure proxy with a near-vertical post-IPO advance now rolling over ~16% off its high, a tiny public float relative to total units, and heavy single-day volume on down days (14.5M shares June 26). That is the profile of a crowded momentum/theme trade in early distribution, not an abandoned value name — consistent with consensus being offsides on price/durability, not on demand. It supports the “AVOID-here, don’t short” framing in Claude’s Take.

12. Fact vs. Interpretation

13. Open Questions

1. Present value of the TRA liability and its annual cash cost to common holders — the prospectus calls it “substantial” but the magnitude/timing depend on exchanges and tax law.
2. Backlog quality — cancellation/termination terms, margin embedded, and the conversion schedule of the ~$2.0B (how much converts in FY27 vs. beyond).
3. Customer concentration by name — who is the >10% Q3 FY26 customer, and how concentrated is the data-center book among a few hyperscalers/neoclouds?
4. Acquisition multiples paid for MGM/PwrQ/States/VanTran and the organic-vs-acquired split of the 56%/86% growth.
5. Margin bridge — why did 9M FY26 Adjusted EBITDA margin (21.9%) fall below 9M FY25 (24.5%), and is mid-20s margin at scale realistic as the shortage resolves?
6. Capex/FCF inflection — is the capacity build truly complete, and when does free cash flow turn durably positive?
7. Lock-up schedule and Neos’s intended pace of further selling.
8. Pricing power post-shortage — what share of current price/margin is shortage premium vs. structural?

14. What Must Be True

Bull case — what must be true: (1) AI/data-center/grid capex stays strong enough to fill Forgent’s ~$5B capacity over the next several years (revenue roughly 3–4x from here); (2) Adjusted EBITDA margins hold or expand to the mid-20s even as the industry adds capacity and the shortage eases; (3) backlog converts at expected margins with limited cancellation; (4) the company executes the capacity fill cleanly as a newly-public operator; and (5) the market continues to award a premium (~Vertiv-like or better) multiple as it scales.

• Falsification test: Two consecutive quarters of decelerating order intake/flat-to-down backlog, or Adjusted EBITDA margin printing below ~20%, or visible lead-time normalization across peers (the shortage resolving) — any one breaks the durable-super-cycle premise.

Bear case — what must be true: (1) the capacity bottleneck resolves over the next 12–24 months as the whole industry adds switchgear/transformer capacity, compressing shortage pricing and margins; (2) growth decelerates from shortage-amplified rates toward mid-teens; (3) the multiple re-rates toward Powell/nVent (cutting the stock materially); and/or (4) sponsor overhang, TRA leakage and leverage weigh on per-share value.

• Falsification test: Sustained 20%+ organic growth into FY27 with Adjusted EBITDA margins expanding above ~23% and free cash flow turning solidly positive, while peers’ lead times stay extended — that would show the economics are durable, not shortage-rented, and break the mean-reversion thesis.

Synthesis: The bull and bear cases agree the demand is real and the business is competent; they disagree on durability and price. The crux is whether ~22% margins and 80%+ growth are structural or shortage-rented, and whether a four-month-public, sponsor-controlled roll-up should trade at a premium to Vertiv. At $55, the price has already decided that question in the bull’s favor — which is the asymmetry this memo flags.

15. Source Appendix

See the separate Source Appendix (Appendix B in the combined report) for the full citation list. Primary sources: Forgent Power Solutions 424B4 prospectuses (IPO, filed 2026-02-06; secondary, filed 2026-06-01), Form 10-Q for the quarter ended March 31, 2026 (filed 2026-05-14), Form 8-K filed 2026-06-26, and Forms 3/4 — all SEC EDGAR, CIK 0002080126. Quantitative data: ROIC.ai (FPS, VRT, NVT, POWL statements, ratios, enterprise value), AZI price CSV and news feed. Industry data points (Omdia, Wood Mackenzie, BCE, Dodge, NREL, ASCE) as cited within the prospectus. FactorsToday and ROIC transcript coverage were unavailable for FPS given its short public history; Google Drive context was not retrievable.

APPENDIX A — Standard Diligence Questionnaire

Forgent Power Solutions, Inc. (NYSE: FPS) — prepared 2026-06-27. Answers are grounded in primary sources; Fact / Interpretation / Assumption labels are applied where it matters.

General

What thoughtful questions have other investors asked about this company? The core debates: (1) Is the ~22% Adjusted EBITDA margin structural or a temporary shortage premium that mean-reverts as the industry adds transformer/switchgear capacity? (2) Can Forgent fill its stated ~$5B revenue capacity, and over what timeframe? (3) Is a premium-to-Vertiv multiple defensible for a two-year-old, leveraged, sponsor-controlled roll-up? (4) How much cash does the Up-C/TRA structure divert to Neos, and how fast will the sponsor keep selling? (5) How real and convertible is the ~$2.0B backlog, and how concentrated among a few hyperscalers? (Interpretation, from the disclosure profile and peer-set framing.)

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? Closer to a cyclical high relative to the company’s own short life and to the industry’s shortage-driven pricing — margins and growth are amplified by an electrical-equipment bottleneck (Interpretation). Absolute earnings are still scaling, but the quality of the current margin/price is peak-of-cycle-flavored.

Driven by the external environment or internal actions? Both: external (AI/data-center capex super-cycle, grid replacement, reshoring, equipment shortage) and internal (capacity build, roll-up integration, lead-time positioning). The external driver dominates the demand; internal execution determines share capture (Interpretation).

How stable are revenues? Project-based and lumpy, not recurring — only ~4% is aftermarket services. Visibility comes from ~$2.0B backlog rather than contracts; revenue is exposed to project push-out/cancellation and to the capital-spending cycle of hyperscalers, IPPs and utilities (Fact/Interpretation).

Outlook for products/services? Strong near-term: demand for switchgear, transformers, PDUs, eHouses and power skids is bottlenecked; backlog +157%. Medium-term depends on whether the capacity shortage persists (Interpretation).

How big will this market be — growing, shrinking, domestic or international? Forgent’s target markets grew at 11–27% CAGRs (2020–2025) vs ~9% for non-residential investment; the electrical-equipment opportunity is large and growing. Revenue is ~100% North American today, with international expansion an explicit but nascent ambition (Fact).

Business Quality & Competitive Moat

Is the industry getting more or less competitive? More: the current shortage and super-normal economics are drawing capacity from the majors (Eaton, Schneider, ABB, Siemens, Hitachi Energy, Vertiv, nVent) and new entrants — the capital cycle at work (Interpretation, Marathon lens).

How profitable is the business (ROIC, ROE)? Adjusted EBITDA margin ~22% (category-average). On full invested capital including the ~$817M of goodwill/intangibles paid in the roll-up, ROIC is ~low-double-digits — roughly at WACC. On tangible operating assets, returns are high (asset-efficient ETO manufacturing), but the price paid captured most of that value for sellers (Interpretation). Reported ROE is distorted by thin/negative tangible equity and minority interest.

How profitable is the industry — how many competitors, barriers to entry? A bifurcated industry: a few large global players plus a long tail of regional/local private fabricators. Barriers are moderate — capital, engineering talent, capacity and lead-time relationships — but not insurmountable; the current shortage is a temporary entry deterrent, not a permanent moat (Interpretation).

Can the business be easily understood? Yes — it makes and sells heavy electrical equipment for data centers, the grid and factories. The structure (Up-C, TRA, Predecessor/Successor accounting, minority interest) is complex, but the underlying business is straightforward (Fact).

Can it be undermined by foreign low-cost labor? Partly insulated: custom, late-spec, lead-time-critical products are hard for overseas commodity competitors to serve (no local engineering, can’t inventory custom, weeks of freight). Standard/commoditized categories (~5% of revenue) are more exposed (Fact/Interpretation).

Do brands matter? Reputation and qualification matter (safety-critical product, high consequence of failure), but this is a spec/relationship/lead-time business, not a consumer brand. The “Forgent” corporate brand is two years old (Interpretation).

Nature of competition? Performance, reliability, warranty, lead time, customization and price. Forgent competes primarily on lead time and breadth in a shortage (Fact).

Customers’ switching costs? Project-level, not installed-base: re-specifying mid-project is costly and the application-engineering relationship is sticky, but each new project is re-competed and there is no recurring software/consumables lock-in (Interpretation).

Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? The capacity footprint (built for ~$5B revenue) and the engineering/customer relationships are partly captured in intangibles/goodwill; the operating leverage of the under-utilized capacity is an off-balance-sheet “asset” if demand fills it (Interpretation).

Off-balance-sheet liabilities? Yes, materially: the Tax Receivable Agreement obligates Forgent to pay 85% of realized tax savings to Neos’s affiliates for 15+ years — a “substantial,” long-dated cash claim not in net debt or EBITDA. Also ~$123M of finance/capital leases and normal purchase/operating commitments (Fact).

How conservative is the accounting? Heavy reliance on non-GAAP (Adjusted EBITDA/Net Income with multiple add-backs), Predecessor/Successor splits, purchase-accounting intangible amortization, and EGC reduced-disclosure status. Revenue includes over-time recognition with deferred revenue/customer advances. Not aggressive per se, but the GAAP-to-adjusted gap is large and warrants scrutiny (Interpretation).

How CapEx-hungry is the business? Was capex-heavy during the 2023–CY2026 capacity build (~$84M FY25 = ~11% of revenue, driving negative FCF). Management says major capacity capex is largely complete, so the model should become more capital-light from here — a key, unproven forward assumption (Fact/Assumption).

Capital Allocation & Management

How much FCF does the business generate, and how is it used? Negative-to-just-turning: FY25 free cash flow ~-$39M (OCF $45M less $84M capex); FY24 OCF negative. Cash has funded growth capex and working capital. No established public-company capital-return philosophy; near-term priorities should be de-levering and the TRA (Fact/Interpretation).

Significant acquisitions recently? The company is a series of acquisitions (MGM, PwrQ, States, VanTran + founding business, 2023–2025), with ~$742M deployed in FY24 and ~$817M of resulting goodwill/intangibles; further M&A is an explicit ongoing strategy (Fact).

Buying back shares? No — the opposite: the sponsor is selling (June secondary, ~$1.3B at $47). The ~$628M primary proceeds were used to redeem sponsor-affiliated Opco units, not to repurchase public stock or de-lever (Fact).

Issuing large amounts of new shares to insiders? The Up-C/Class B/Opco-unit structure and the TRA are designed to deliver outsized economics to Neos; equity-based compensation exists but is modest in cash terms so far. The defining issuance event was the IPO/secondary (Fact/Interpretation).

Compensation policy of directors/management? Reduced-disclosure EGC; controlled-company governance. Management’s operating pedigree is strong (ex-Vertiv/Caterpillar/Schneider). Detailed incentive-metric alignment with minority holders is not yet established in a public proxy; the controlling structure favors Neos (Interpretation/Open Question).

Motivations of management? Operating team is incentivized to scale and execute the Vertiv-style playbook; the controlling sponsor (Neos) is incentivized to maximize and monetize value — visible in the rapid IPO-then-secondary distribution (Interpretation).

Valuation & Market Data

Is the stock an ADR, MLP, or K-1 issuer? No ADR, no MLP. It is a C-corp in an Up-C structure; public holders own Class A common stock and receive a 1099, not a K-1 (the K-1/partnership economics sit with the Continuing Equity Owners at the Opco level) (Fact).

Dividend policy? No meaningful common dividend policy established; a nominal pre-IPO/structuring dividend (~$13M FY25) was paid. Given leverage, negative tangible equity and TRA obligations, a public dividend is not a near-term priority (Fact/Interpretation).

How profitable is the business? ~22% Adjusted EBITDA margin; GAAP near-breakeven (FY25 net loss $6.2M) due to interest; turning GAAP-profitable in Q3 FY26 as interest fell post-IPO (Fact).

Is net income diverging from cash from operations? Yes — in FY25, EBITDA ($101M) far exceeded OCF ($45M) due to heavy working-capital build and capex; adjusted net income (~$150M TTM) also diverges from GAAP (near-breakeven). The divergences reflect scaling intensity and the GAAP/adjusted gap, not (so far) earnings-quality manipulation, but both warrant monitoring (Interpretation).

Risks & Downside

What factors would cause the stock to decline? Multiple de-rating toward peers (the dominant risk at ~68x EBITDA); shortage resolution compressing margins; AI-capex digestion; further sponsor selling/overhang; backlog conversion disappointment; input-cost/tariff margin pressure; integration stumbles (see matrix).

Risk of a catastrophic loss? Low in the near term — real revenue, real backlog, moderate leverage (~1.9x), adequate liquidity. The realistic downside is a sharp valuation de-rating, not insolvency (Interpretation).

Chance of a total loss? Low. It would require a severe, sustained collapse in data-center/grid demand combined with a refinancing failure — not the base case given current demand and leverage (Interpretation).

Recent News & Events

Has the business environment changed recently? The operating environment remains strongly favorable (AI/data-center power demand, backlog +157%). The market environment shifted in late June 2026 — a tech/AI-infrastructure sell-off pulled FPS ~16% off its June 4 high; the June secondary added supply/overhang (Fact).

Significant acquisitions? None newly announced post-IPO; the roll-up history (MGM/PwrQ/States/VanTran) predates the IPO. Further M&A remains a stated strategy (Fact).

Change in accounting policies? None disclosed beyond the ongoing Predecessor/Successor and Up-C/TRA accounting established at IPO (Fact).

Recent changes — new markets, facilities, management? Capacity-expansion program (ten campuses/five locations) nearly complete; December 2025 debt refinancing and June 23, 2026 credit-agreement Amendment No. 1; IPO (Feb 5, 2026, $27) and secondary (June 1, 2026, $47); TD Cowen Buy/PT $73 (June 22, 2026). International expansion is an early-stage ambition (Fact).

APPENDIX B — Source Appendix

Forgent Power Solutions, Inc. (NYSE: FPS) — sources accessed 2026-06-27. Primary sources first; all SEC filings under CIK 0002080126.

Primary — SEC filings (EDGAR)

1. Form 424B4 (secondary offering prospectus) — filed 2026-06-01, accession 0001193125-26-249223. Cover/Up-C structure, TRA, Continuing Equity Owners (145,218,850 Class A + 60,310,039 Class B), 80.19% Opco ownership, “controlled company”; offering of 42,280,000 shares at $47.00 ($1,987,160,000; $1,304,602,662 to selling stockholders, $627,910,438 to the Company to redeem Opco units); business overview, end-market mix (DC 42% / Grid 23% / Industrial 19% / Other 16%), product mix (Standard 5% / Custom 78% / Powertrain 13% / services 4%), backlog ~$2.0B (+157%), capacity to ~$5B, 10 campuses / 5 locations, ~2,400 employees; competition (Vertiv, Hitachi Energy, nVent, smaller privates); summary financials and non-GAAP reconciliations (Adjusted EBITDA: PF2024 $99.2M/20.6%, FY25 $169.2M/22.5%, 9M FY26 $210.2M/21.9%); customer concentration; risk factors. URL: https://www.sec.gov/Archives/edgar/data/2080126/000119312526249223/olympus_iii_424b4.htm
2. Form 424B4 (IPO prospectus) — filed 2026-02-06, accession 0001193125-26-040029. IPO of 56,000,000 Class A shares at $27.00 ($1,512,000,000 gross); use of proceeds; Up-C/TRA formation. URL: https://www.sec.gov/Archives/edgar/data/2080126/000119312526040029/d890989d424b4.htm
3. Form 10-Q (Q3 FY26, quarter ended March 31, 2026) — filed 2026-05-14, accession 0002080126-26-000018. Quarterly statements; customer concentration (one >10% in Q3 FY26, none for 9M FY26); interest-expense trajectory; balance sheet (goodwill $516.6M, intangibles $300.3M, net debt ~$490M, negative tangible common equity). URL: https://www.sec.gov/Archives/edgar/data/2080126/000208012626000018/fps-20260331.htm
4. Form 8-K — filed 2026-06-26 (event June 23, 2026), accession 0001193125-26-283773. Item 1.01 — Amendment No. 1 to the Credit Agreement dated December 19, 2025 (Forgent Power LLC). URL: https://www.sec.gov/Archives/edgar/data/2080126/000119312526283773/na-20260623.htm
5. Forms 3 / 4 (Feb–June 2026) — insider/sponsor ownership and the June secondary transactions. EDGAR CIK 0002080126.
6. SEC EDGAR submissions metadata — company name, NYSE listing, SIC (Electrical Industrial Apparatus), filing history. https://data.sec.gov/submissions/CIK0002080126.json

Quantitative data sources

7. ROIC.ai — FPS company profile, income statement (annual + quarterly), balance sheet, cash flow, profitability ratios, per-share data, enterprise value; comparative enterprise-value/multiples for Vertiv (VRT), nVent (NVT) and Powell Industries (POWL). Third-party aggregated data; reconciled to filings.
8. AZI price history CSV (azitrading.com) — FPS daily OHLCV since 2026-02-05; used for the price-action event map. Valuation-percentile (valuation_index) unavailable (insufficient public history).
9. AZI news feed — FPS headlines (May–June 2026), including “Forgent Rides AI Infrastructure Boom” (2026-05-18) and “TD Cowen Maintains Buy on Forgent Power Solutions, Raises Price Target to $73” (2026-06-22).

Industry data (as cited within the prospectus)

10. End-market growth CAGRs 2020–2025 — Omdia (data centers ~27%), Wood Mackenzie (power plants ~15%), BCE (T&D ~11%), Dodge Construction Network (manufacturing ~17%); overall non-residential ~9%.
11. Grid-equipment age/useful life — NREL and ASCE (majority of in-service distribution equipment >30 years old vs 25–40 year life).

Peer comparables (ROIC, latest FY/TTM)

12. Vertiv Holdings (VRT) — EV ~$63.2B; EV/sales 6.2x; EV/EBITDA 28.7x; EBITDA margin ~21.5%.
13. nVent Electric (NVT) — EV ~$18.0B; EV/sales 4.6x; EV/EBITDA 21.9x; EBITDA margin ~21%.
14. Powell Industries (POWL) — EV ~$3.2B; EV/sales 2.9x; EV/EBITDA 14.2x; EBITDA margin ~20%.

Coverage gaps / unavailable sources

15. FactorsToday — no factor-loading/leaderboard data returned (model universe requires ≥252 trading days; FPS has been public ~4.5 months). Price-action read built from the AZI CSV only.
16. ROIC earnings-call transcripts — none listed (company too newly public). No transcript sweep possible; management framing taken from the prospectus/MD&A.

Analytical frameworks applied

18. Greenwald & Kahn, Competition Demystified — moat-type taxonomy (the lead-time/customization edge classified as narrow, cycle-amplified customer captivity, not a durable barrier; ROIC-on-invested-capital ≈ WACC test).
19. Marathon / Chancellor, Capital Returns — supply-side capital-cycle analysis (shortage pricing attracts capacity and mean-reverts).