Powell's own filing concedes some rivals have far greater resources and lower costs.
Is it a good business? · 6 · What does its record show?
Sales more than doubled and profit rose
Sales, $B
0.5
0.7
1.0
1.1
1.3
Sep 2022Sep 2023Sep 2024Sep 2025Sep 2026
Sales$471M → $1.26Bsales, year to September 2021 to year to September 2026
Profit0.2% → 21%operating profit per $1
Kept16% → 31%kept per $1 after making it
Capital spending stays near one cent per dollar because Powell leases its factories.
Is it a good business? · 7 · How much room is there to grow?
Demand is large; capacity is the limit
$2.4Bof signed orders waiting
About $1.3Bdeliverable in the next year
22% to 33%data centres' share of commercial power by 2050
Nobody turns the power projections into equipment dollars, so the market cannot be sized from outside.
Is it a good business? · 8 · Who is in charge?
The same two leaders since the lean years
2016Brett Cope took charge
2018Michael Metcalf took over finances
19.7%of each $1 kept as operating profit, fiscal 2025
Under them Powell went from barely breaking even to solid profit.
The story right now · 9 · Why do data centres need Powell?
Data centres need big electrical rooms
+54%data-centre and commercial sales
$934Morders in one quarter, against $312M of sales
$2.7Bof signed work waiting to be built
Data centres may use 22% to 33% of commercial power by 2050, and each needs custom equipment to take power from the grid safely. Powell builds that equipment.
The story right now · 10 · Why can't sales grow faster?
Its own factories set the pace
Record ordersBiggest ever.
Limited factory spaceSkilled workers short.
54% deliverable in a yearThe rest waits.
Sales grow slowerThan orders.
A share of signed orders deliverable within a year back toward the low sixties would prove us too cautious.
The story right now · 11 · What changed lately?
Sales up a third, backlog doubled to $2.7B
+33%sales growth in the September quarter
34.9%kept per $1 after making it, against 31.4% a year earlier
$710Mnew orders, against $271M a year earlier
$2.7Bbacklog, double a year earlier
Strong quarter
The report did not say how much of the backlog can be delivered within a year.
The story right now · 12 · What is everyone arguing about?
A peak margin, or a new normal?
Many investors think
A slowing industrial that ran too far
We think
The margin is at a peak, and we pay for a normal year
Against us: a doubled backlog and a finance chief who says this year's margin holds.
The price and our view · 13 · What is a share worth?
The price sits above even our sunny case
$104what we think a share is worth · the last close is of that
$104our value
Attractive below $83.27
Neutral
Unattractive above $125
$66.20gloomy: growth halves, less kept per $1
$104our estimate: experts' path, then slowing
$132sunny: growth beats the forecast by a quarter
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?
Capacity caps the growth
Capacity limitsBiggest
Orders wait.
Profit at a peakShowing now
Five years ago, near zero.
Rivals add capacityWatch
nVent, Eaton.
Data-centre demand sizeUnknown
Not measured.
Oil and gas is still the biggest slice, and petrochemical sales halved.
The price and our view · 15 · How will we know?
2 February tests the margin
At least 54% of backlog deliverable within a yearGood sign
Kept per $1 stays above 30%Good sign
Orders fall under $400MBad sign
FEB 20272
First-quarter fiscal 2027 results, 2 February. Companies report every three months; each report is a test of the story.
Our view also changes on price alone: if the shares fall to within a fifth of our estimate, we move to Neutral.
The price and our view · 16 · So, what do we think?
A strong builder priced past its best case
$104our value
Attractive below $83.27
Neutral
Unattractive above $125
UnattractiveMedium conviction
Powell builds the electrical rooms between the grid and big users such as data centres, earning double its rivals' returns. Its year ended strongly, but its profit share is at a peak and the price sits above our sunny case.
↑ ↓ to move · C for contents
What it does
More about the company
Who it is
Anything that draws a lot of power, from a data centre to a gas export terminal, needs equipment between the supply and the load to distribute, switch and protect the electricity. Powell, founded in Houston in 1947, designs and builds that equipment, often as a whole prefabricated electrical room.
The alternative is buying switchgear, breakers and controls from several vendors and paying a contractor to assemble them on site, which is slower. In data centres, Powell wins on delivery speed rather than price.
How it makes money
Every job is custom and won by competitive bid, so Powell earns by the project: an order is booked, built over one to three years, and becomes sales as it ships. Big projects rarely repeat with the same customer, and no customer was more than 10% of sales in either of the last two years.
Growth and profit
What growth costs
For years, growing made Powell more profitable: sales doubled while office, selling and research costs grew far less. That effect has stalled; since 2024 those costs have grown a little faster than sales, so recent margin gains came from the factory floor. It spends little on buildings, leasing a plant rather than building one.
Its record
The numbers
Is it growing?Steady
Sales grew 14% in the year to Sep 2026, about the same pace as the 9.1% in the year before.
Is it profitable?Strong
Operating margin is 21%, up from 8.9% in the year to Sep 2023.
Does growth pay off?Pays off
Running costs take 10¢ of each sales dollar, down from 15¢ in the year to Sep 2022.
Does it earn good returns?Strong
Each dollar put into the business earns 28% a year after tax, up from 2.4% in the year to Sep 2022.
Is it making cash?Strong
It turned 33% of sales into free cash, $415M over the year.
How steady is it?Uneven
Yearly sales growth has ranged from +9.1% to +45% over the last 5 years.
Are shares being diluted?Stable
The share count barely moved in 2.0 years, and pay in shares costs 0.5% of sales.
Why it wins
The evidence
The scarce thing in this industry is not design but factory capacity and the skilled workers to run it. So buyers commit early to secure a delivery slot: last quarter Powell booked orders worth 3.0 times its sales, and its backlog takes one to three years to deliver.
The same squeeze shows at rivals. nVent says its own large orders can take more than a year to build.
Who could beat it
The rivals in full
Its rivals are much larger, and Powell's own filing concedes some have far greater resources and lower costs. nVent, Eaton and Hubbell are the listed ones, and nVent's $1.0B purchase of a switchgear maker targets Powell's two growth markets. If rivals add capacity faster than demand grows, the delivery-slot advantage goes.
Each rival
nVent Electric, which in 2025 bought the enclosures, switchgear and bus systems businesses of Avail Infrastructure Solutions for approximately $1 billion in cashNVT
The delivery slot itself - a competitor buying rated switchgear capacity directly attacks the scarcity that lets Powell win on speed, and Avail's business is described in nVent's own filing as serving power utilities and data centres, Powell's own two growth markets.
Eaton and Hubbell, the two other US-listed multinationals selling power distribution equipment into the grid, data centres and industrialsETNHUBB
The order itself, on competitively bid work - Powell's own filing concedes that some competitors 'are significantly larger and have substantially greater global resources' and may have lower cost structures.
Siemens Energy, Schneider Electric and ABB, the European multinationals in the same class Powell's Item 1 namesENR.DESU.PAABBN.SW
The same bid, on international and large utility work, from balance sheets far larger than Powell's; none files with the SEC, so none appears in the cohort and the comparison set understates the true competitive field.
How big it could get
The market and who has signed
The market
The demand is large and hard to size. Government projections have data-centre servers rising from about 7% of the electricity American commercial buildings use in 2025 to 22% to 33% by 2050, but nobody converts that into equipment dollars.
The limit is Powell's own capacity. It made $1.10B of sales in fiscal 2025, and only about $1.3B of its $2.4B backlog can be delivered in the next year.
Who has signed
Three large orders last quarter, none with a named customer: a data-centre power system worth more than $400M, a $75M fertiliser plant and a $60M gas export terminal. The data-centre job is only the first phase of a larger programme, and the later phases have not been awarded.
Who runs it
Leaders, pay and the cash
Who runs it
Brett Cope has run Powell since 2016 and chairs its board; Michael Metcalf has run its finances since 2018. Under them it went from barely breaking even to keeping about a fifth of each sales dollar as operating profit. Insiders own only about 2%; the largest holder, Thomas Powell, is not on the board. Bonuses and share awards ride on a profit measure and on safety, with goals set below results already reached, and nothing pays on the share price.
What they did with the money
It has spent lightly on capacity: $12.4M to expand a Houston yard by 62%, and a leased plant due in 2027 rather than a new build it put at $70M to $100M. It has no borrowings and $634M of cash.
The last report
The latest report
Powell's sales rose 33% and its backlog doubled to $2.7 billion
What it did to the thesis Strong quarter
It answered our three questions. Margins rose and orders stayed far above last year, in the case's favour. Delivery slowed as a share of the backlog, though not in dollars, and part of the margin was one-off. The price stays dearer than our best case.
The report
In the fourth quarter of fiscal 2026, to September 30, sales rose 33% to $397M and Powell kept 34.9 cents of each sales dollar after making it, against 31.4 cents a year earlier. Orders of $710M took the backlog to $2.7B, double a year earlier.
On the call the finance chief said about $1.3B of the backlog, roughly half, ships within a year, against 60% a year earlier, because the new orders are big and slow. About 2.3 points of the quarter's margin came from one-off project closeouts; the base margin is near 30%, which he expects to hold next year.
What's happening now
The background
The call on 17 November put about half the backlog inside the next year and part of the quarter's margin down to one-off closeouts. The next check is the first fiscal 2027 quarter on 2 February.
Where we disagree
Both sides in full
The market
A slowing industrial that ran too far: sales grew 9% while orders tripled, and only 3 of 10 analysts say buy.
We believe
The fall removed the excess without creating a bargain. The growth ceiling management describes is real, and set by its own capacity.
The argument
The market is reading Powell against its own history, when it barely made money. Against live rivals it is mid-pack on price and earns double their returns, which is why we are not bearish.
We would be too cautious if the deliverable share of backlog recovers toward the low sixties, growth outpaces rivals again, and the later data-centre phases come to Powell.
What it's worth
How we value it
Valued on the analysts' revenue path for three years, each later year closing 30% of the gap to 4%, on official operating profit with stock pay a cost and tax paid in full; the thesis's own evidence is the check on that path.
What the price assumes
Sales growth in the first year, then slowing as ours does, at the same profit per sale.
The price assumed
42.6%
Our estimate
18%
Measured at the close of
16 Nov 2026
The three cases
Bear case$66.20
Growth runs at half the analysts' path, 9%, 8% and 7%, the margin settles a fifth lower at 7.7%, and new capital earns only the 8% rate: the downturn test.
Our value$104
Revenue grows 18%, 16% and 14% over three years on fiscal 2026's $1.26B, the analysts' path (C37), then fades toward 4%, closing 30% of the gap each year; the margin settles at its five-year average of 9.6%, not today's cycle peak, though the $2.7B backlog (C39) and the CFO's statement (C40) hold it high in the early years.
Bull case$132
Growth runs a quarter above the analysts' path, 22%, 20% and 18%, the margin reaches 10.5%, and new capital earns 15%.
What could go wrong
Every risk
The risks
Not enough skilled workers
The chief executive expects craft labour to be a challenge in 2027 and 2028, which would delay deliveries.
Margins have stopped rising
Gross margin was lower than a year earlier in each of the last two quarters, as one-off gains from closing projects fade.
Orders can be cancelled
Customers may walk away from the backlog, usually paying costs and a margin rather than the full order.
Insiders selling
Insiders made 79 open-market sales and no purchases in the past year.
What we still do not know
Whether the later phases of the data-centre programme come to Powell. None has been awarded.
Whether the leased plant, due in 2027, lifts how much of the backlog converts each year.
What changes our view
Every route to a new view
Every route to a new view
→ Neutral
The share price falls to about $125, back within 20% of what we think the business is worth.
→ Attractive
The share price falls to about $83, 80% of what we think the business is worth.
→ Neutral
Gross margin holds at fiscal 2026's 31.3% or better through the next two quarters, with at least $1.3B of the backlog deliverable within a year. Quarterly orders stay above $500M, so the order book is a level, not two big orders.
The next test
It supports the case if
First-quarter gross margin at or above 30%, backlog convertible within twelve months at or above $1.3 billion, and orders above $500 million.
It pressures the case if
Gross margin back below the 28.4% of fiscal Q1 2026, backlog convertible within twelve months below $1.3 billion, or orders below $400 million with no mega order.
We would also be wrong if
Gross margin falls against the year before for three quarters in a row.
The verdict
Scores and how we weighed it
The scores
Growth
3 of 4
Business quality
3 of 4
Financial position
4 of 4
Execution record
3 of 4
Per share capture
4 of 4
Potential
3 of 4
Risk profile
3 of 4
Valuation
1 of 4
Evidence quality
2 of 4
How we weighed it
ValueNegative
The price sits well above our value and above the bull case, even after a strong report.
BusinessPositive
Returns on capital are more than twice its rivals'.
GrowthPositive
Sales grew in each quarter faster than the last, and the backlog doubled in a year.
RisksNegative
The margin is at a peak, skilled workers are short and orders can be cancelled.
RivalsPositive
Mid-pack among listed rivals despite the best returns of the group.
The groupMixed
Power names move together on AI power news.
Latest newsPositive
The last report beat what the thesis needed on margin and orders; on the call part of the margin proved one-off, and only about half the backlog ships within a year.
In plain words
Backlog
Work customers have signed for that has not been built and billed yet. It turns into sales over the coming quarters.
In plain words
Margin
The share of each dollar of sales a company keeps as profit after paying what the product cost to make.
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.