Forgent builds the gear that feeds power to data centres
Switchgear
breakers that cut off a fault
Transformers
lower the grid's voltage
Skids
whole tested systems in one unit
Power arrives at a data centre at high voltage and must be switched, protected and stepped down before servers can use it. Forgent builds the gear that does this.
The company · 2 · What exactly does it sell?
Most of it is built to one project's design
Each $1 of sales, fiscal 2026
70%25%3%2%
Gear built to orderWhole tested systems on skidsStandard gearServicing
A growing share ships as one tested system, so the site plugs in one unit instead of wiring dozens.
The company · 3 · How does it make money?
Paid per order, mostly by data centres
59%of sales from data centres
35 centskept per $1 after making the gear
11%of sales from one customer
Utilities bought 21% and factories 10%, nearly all in North America. Bigger system orders mean fewer, larger buyers.
Is it a good business? · 4 · What gives it an edge?
In a shortage, delivering is the edge
About 44 weeks
the industry's switchgear wait in 2025
1 week
Forgent's fastest shipping time, it says
+161%
its data-centre sales, against 42% building
It makes its own switchgear and transformers, so it ships fast while rivals' gear takes months. The edge lasts only as long as the shortage.
Is it a good business? · 5 · What is the catch?
Big rivals are adding capacity
Vertivmore than doubled its switchgear capacity
Schneider, Eatonnamed in NVIDIA's next rack design
Forgent's own annual report warns prices could fall if capacity outruns demand.
Is it a good business? · 6 · What does its record show?
Growth has paid so far
Sales, $B
0.8
1.4
1.6
Jun 2025Jun 2026last 12 mo
Sales$0.8B → $1.6BJun 2025 to the last 12 months
Profit per $110 → 13 centsoperating profit per $1, fiscal 2026
Gross profit per $135 centskept after making the gear, slipping
Overheads spread over more sales. Building factories cost more cash than the business made last year.
Is it a good business? · 7 · How much room is there to grow?
Data-centre building keeps climbing
$45.2BUS data-centre building, first eight months of 2026
+41.7%on a year earlier
About 40%of data centres built from modules
Nobody publishes how big Forgent's slice is, and the US still imports most of its transformers.
Is it a good business? · 8 · Who is in charge?
A chief from a rival, and a private-equity owner selling
Since 2025Gary Niederpruem, after six years at Vertiv
37%of the vote held by Neos Partners
0 sharesheld directly by any executive or director
The finance chief left on 22 October; Amy Ard, once Proterra's finance chief, took over. Neos has sold in every offering since listing.
The story right now · 9 · Why is power gear so short?
AI data centres need power gear faster than it is made
$45.2Bof US data-centre building in eight months
About 44 weeksto get switchgear in 2025
About 80%of US power transformers supplied by imports, 2025 estimate
Every AI data centre needs a huge power connection, and the gear that brings power in is scarce. A maker that can ship fast wins the orders.
The story right now · 10 · What happens when rivals catch up?
A shortage pays well until capacity arrives
Demand outruns supplyWaits stretch to months.
The fast maker winsSales up 69%.
Rivals add factoriesVertiv more than doubled.
Prices and orders fallIf capacity outruns demand.
Orders already in hand cover this year. The years after depend on orders still coming once the shortage eases.
The story right now · 11 · What changed lately?
Sales rose 69% and the yearly plan went up
$478.6Mquarterly sales, above the forecast
$136Mcash from operations, helped by deposits
$2.45B to $2.65Bthe raised yearly plan
−38.2%the shares in three October weeks
Strong quarter
The shares fell 9.8% on the day of a beat. Orders will now be reported only once a year.
The story right now · 12 · What is everyone arguing about?
Will orders outlast the shortage?
Many investors think
A fast AI power winner, marked down hard
We think
This year is bought; later years need the shortage to last
Against us: it beat forecasts in all three listed quarters, with orders 3.3 times sales.
The price and our view · 13 · What is a share worth?
The price sits about a quarter below our value
$27.66what we think a share is worth · the last close is of that
$27.66our value
Attractive below $22.13
Unattractive above $33.19
$7.30gloomy: orders stop when the shortage ends
$27.66our estimate: this year, then slowing
$55.40sunny: direct AI orders keep growth near 30%
Value per share in three futures. None is a prediction of the share price.
We estimate what a share is worth from the cash the business should earn over many years. Our rule: buy only when the price is at least a fifth below our estimate, as a cushion for being wrong.
The price and our view · 14 · How could we be wrong?
The shortage ends, and one kind of customer
The shortage endsBiggest
Prices and orders could fall together.
An owner who keeps sellingShowing now
Neos sold in every offering.
Copper in fixed pricesWatch
Many orders cannot pass on metal costs.
Cancellation termsUnknown
Not public for $3.0B of orders.
Data centres are 59% of sales, and orders will now be reported only once a year.
The price and our view · 15 · How will we know?
February shows whether profit per sale climbs
Sales on pace for the raised plan, 22 cents or more keptGood sign
A cut to the yearly plan below $2.4BWe turn negative
Restated results or weak accounting controlsWe turn negative
FEB 202715
Second-quarter fiscal 2027 results. Companies report every three months; each report is a test of the story.
Our view also changes on price alone: we turn Neutral if the shares rise back to a fifth below our estimate.
The price and our view · 16 · So, what do we think?
Power gear in a shortage, a year already ordered
$27.66our value
Attractive below $22.13
Unattractive above $33.19
AttractiveMedium conviction
Forgent builds the switchgear and transformers that bring grid power into data centres, and sales grew 69%. The price sits about a quarter below our value, enough cushion for a shortage that will ease: Attractive.
↑ ↓ to move · C for contents
What it does
More about the company
Who it is
Power arrives at a data center at high voltage and must be switched, protected and stepped down before servers can use it. Forgent builds the gear that does this: switchgear, which holds the breakers that cut off a fault; transformers, which lower the voltage; and the panels that carry power to the racks.
Most of it is built to one project's design. A growing share ships as a whole tested system on a skid, so the site plugs in one unit instead of wiring dozens.
How it makes money
Forgent is paid per order, mostly for custom gear: about 70% of sales last year, with whole skids another 25%. Data centers bought about 59%, utilities 21% and factories 10%, nearly all in North America.
Profit after making the gear was 35 cents on each dollar of sales. One customer alone was 11% of sales, and bigger system orders mean fewer, larger buyers.
Growth and profit
What growth costs
Growth has paid so far: as sales grew 89%, operating profit rose from about 10 cents to 13 cents per dollar of sales, because overheads spread over more sales. Building factories cost more cash than the business made last year; the company expects capital spending of about 3% of sales this year.
Its record
The numbers
Is it growing?Fast
Sales grew 83% over the twelve months to Sep 2026, slower than the 89% in the year to Jun 2026.
Is it profitable?Healthy
Operating margin is 14%.
Does growth pay off?Pays off
Running costs take 20¢ of each sales dollar, down from 27¢ in the year to Jun 2025.
Does it earn good returns?Decent
Each dollar put into the business earns 14% a year after tax, up from 6.3% in the year to Jun 2025.
Is it making cash?Solid
It turned 7.5% of sales into free cash, $122M over the last twelve months.
Is the balance sheet safe?Light debt
It owes $381M more than the cash it holds, about 1.6 years of the cash its operations bring in.
Is profit turning into cash?Real cash
Free cash flow was 1.0 times net income.
Why it wins
The evidence
In a shortage, being able to deliver is the advantage. Industry switchgear took about 44 weeks to arrive in 2025, and Forgent says it can design a custom unit in hours and ship it in as little as a week, because it makes its own switchgear and transformers.
That shows up as share: its data center sales grew 161% while US data center building grew about 42%. But it lasts only as long as the shortage, and nobody outside the company measures its delivery times.
Who could beat it
The rivals in full
Vertiv, Schneider Electric, Eaton and Hubbell sell the same gear to the same buyers, and Vertiv has more than doubled its switchgear capacity since 2021. Forgent's own annual report warns prices could fall if the industry adds capacity faster than demand grows. NVIDIA's next rack power design names Eaton, Schneider and Vertiv as partners, not Forgent.
Builders buy modular power enclosures and utilities buy distribution transformers from Eaton rather than Forgent.
Hubbell (PCX and Systems Control)HUBB
Data centers buy modular power from PCX and utilities buy substation control buildings from Systems Control.
Hitachi Energy, Prolec GE and WEG6501.TGEVWEGE3.SA
Utilities and plants buy transformers from these makers, Prolec now wholly GE Vernova's, as their US plants expand.
IEM and Mission Critical Groupprivate
Private US makers of custom switchgear and power skids, the closest in shape to Forgent, win the same data center projects.
How big it could get
The market and who has signed
The market
US builders spent $45.2B on data centers in the first eight months of 2026, up 41.7%. Forgent's gear is a slice of that, but nobody publishes how big a slice, so the market cannot be sized precisely.
Two trends help: about 40% of data centers are now built from factory-made modules, which management says outside research expects to reach 60% by 2030, and the US still imports most of its transformers.
Who has signed
Customers have placed $3.0B of orders, which management says are all firm purchase orders, though cancellation terms are not public. A frontier AI lab placed its first direct order and a hyperscaler signed a framework agreement; neither is named, sized or dated.
Who runs it
Leaders, pay and the cash
Who runs it
Gary Niederpruem became chief executive in May 2025 after six years at Vertiv, a rival. Ryan Fiedler, finance chief since July 2025, stepped down on 21 October 2026, citing family time; Amy Ard, head of finance since January and once Proterra's finance chief, took over.
No executive owns shares directly; their stake is in the private-equity owner's partnerships, mostly unvested. Last year's bonus paid on sales and profit. The owner, Neos Partners, holds 37% of the vote.
What they did with the money
Every dollar the company raised in its share sales went to buy out part of its private-equity owner, so the company kept none. It built 1.8 million square feet of factories, plans $35M more in Tijuana, refinanced its loan more cheaply, and names acquisitions as its first use of cash.
The last report
The latest report
Forgent's September-quarter sales rose 69% and it raised its yearly plan
What it did to the thesis Strong quarter
It showed the ordered year turning into sales on time and lifted it a little. Management's order figures point to next year too, but are checked only yearly.
The report
Sales in the first quarter of fiscal 2027 rose 69% to $478.6M, above the $445M to $465M the company had forecast. Profit before interest, tax and depreciation, as the company counts it, was $104.9M, 21.9% of sales.
Customers paid $159M more in deposits, so cash from operations reached $136M. The company raised its plan for the year to $2.45B to $2.65B of sales, which it says its orders fully cover.
On the call management said orders were about double a year earlier and about 40% of the order book falls after this year, figures the filings no longer give.
What's happening now
The background
The next test is the December quarter, reported in February 2027: whether the raised plan for the year holds and the profit margin climbs toward the 24% the plan implies. Its private-equity owner has sold shares in each follow-on offering since listing and expects to sell more.
Where we disagree
Both sides in full
The market
A fast AI power winner growing far faster than its market, which has now been marked down 62% from its high.
We believe
The year ahead is bought and the next is nearly free at this price; the years after depend on orders outlasting a shortage everyone is adding capacity to end.
The argument
The orders are real, but they are this year's sales, not growth still to come, and the price already counts them. What is not counted is the next year, which depends on orders arriving after rivals add capacity, and those orders will be reported only once a year. At the lower price that risk is mostly paid for.
Against us: the company beat forecasts in each of its three quarters as a listed company, and orders ran at 3.3 times sales in the latest one.
What it's worth
How we value it
A business with a year already in firm orders is valued on the guide for that year and slower growth after it, slowing toward 4% by year ten, each year closing 30% of the gap, with the margin set below the shortage's peak.
What the price assumes
Sales growth in the first year, then slowing as ours does, at the same profit per sale.
The price assumed
34.0%
Our estimate
80%
Measured at the close of
16 Nov 2026
The three cases
Bear case$7.30
Orders stop once the shortage ends (C12): FY27 lands a little under the raised guide at 72%, then no growth, the margin falls back to FY26's 12.8%, and new capital earns only 8%.
Our value$27.66
FY27 grows the raised guide's 80% midpoint, which management says backlog fully covers (C23); FY28 25% and FY29 15%, below the analysts' 34%, as Vertiv and others add capacity (C12); the operating margin settles at 18%, under the shortage's guided EBITDA margin (C20), once the shortage pricing fades.
Bull case$55.40
Direct orders from an AI lab and a hyperscaler (C15) keep growth near the analysts' 34% and 30%, and the margin holds at 21% as modular builds rise toward 60% of data centres (C16).
What could go wrong
Every risk
The risks
The shortage ends
Prices rose because supply fell short; as Vertiv, Eaton and Forgent itself add factories, prices and orders could fall together.
One kind of customer
Data centers are 59% of sales and one customer 11%; a pause in AI building hits most of the company at once.
An owner who keeps selling
Neos holds 37% of the vote, can block major decisions, and has sold in every offering since February.
Copper in fixed prices
Many orders carry no clause passing on metal costs, and copper rose last year and again this quarter.
Less to see
Orders will be reported once a year, so a slowdown could show first in a missed quarterly forecast.
What we still do not know
How Forgent's delivery times compare with rivals' product by product, measured by a buyer rather than by the company.
What a customer can do to cancel or delay an order in the $3.0B backlog, and at what cost.
How much of last year's margin came from shortage pricing that would reverse when delivery times return to normal.
Why the finance chief left seven months after listing, without severance, weeks before the first quarter's report; the company says family time and no disagreement.
What changes our view
Every route to a new view
Every route to a new view
→ Neutral
The share price rises back to about $22.13, 80% of what we think the business is worth.
→ Unattractive
The company cuts its fiscal 2027 sales forecast below $2.4B. Orders taken in the year to June 2027 come in below that year's sales.
→ Unattractive
A filing restates past results, reports a weakness in its accounting controls, or the auditor leaves.
→ Unattractive
The share price rises to about $33.19, 20% above what we think the business is worth.
The next test
It supports the case if
Revenue on a pace for the $2.45B to $2.65B guide, an adjusted EBITDA margin of 22% or more, and the FY27 guide held or raised.
It pressures the case if
An adjusted EBITDA margin below 22%, or the FY27 guide cut, or cash from operations negative once customer deposits stop growing.
We would also be wrong if
The company cuts its fiscal 2027 sales forecast below $2.4B.
The profit margin stays below 21% after the first quarter of fiscal 2027.
Industry switchgear delivery times fall below six months while Forgent's prices fall.
The verdict
Scores and how we weighed it
The scores
Growth
3 of 4
Business quality
3 of 4
Financial position
3 of 4
Execution record
3 of 4
Per share capture
3 of 4
Potential
2 of 4
Risk profile
3 of 4
Valuation
4 of 4
Evidence quality
3 of 4
How we weighed it
ValuePositive
The price has fallen past the margin of safety under our value with no new fact against the business; the bear case sits far below and the bull far above.
BusinessPositive
It builds data-centre power gear fast and is winning share in a shortage.
GrowthMixed
Firm orders cover this year and, management says, much of next, with orders about doubling; growth slows as rivals add capacity.
RisksNegative
The shortage will end, one kind of customer dominates, the controlling owner keeps cashing out and copper sits in fixed prices.
RivalsMixed
Its filed rivals have not been re-measured since the fall, so the comparison is stale.
The groupMixed
Contractors and power gear for data centres were marked down together.
Latest newsPositive
The first quarter beat the forecast and raised the year; since then the shares have fallen with no company news.
In plain words
Our estimate of value
We forecast how much cash the business can earn over the next ten years and beyond, then ask what that stream is worth today to an investor who wants an 8% yearly return. We do it three times: a gloomy future, an expected one and a sunny one. We call a share Attractive only when its price is at least 20% below the expected estimate, and Unattractive when it is 20% above.