What changed: orders reached $934 million, book-to-bill was 3.0x and backlog reached $2.4 billion, extending demand visibility into fiscal 2027 and beyond. S1
What did not: Q3 revenue missed the pre-print mean by 1.6%, EPS missed by 4.5%, and management supplied no numeric revenue, EPS, free-cash-flow or backlog-conversion guide. S1S5
Why the stock reset is not enough by itself: the frozen $189.90 after-hours price still equals 27.6x pre-print fiscal 2027 EPS. That multiple needs clean project execution, durable margins and cash conversion—not backlog alone. S5S6J1
Growth-investment scorecard
Fundamentals snapshot
What the company does: Powell engineers and manufactures custom electrical power-distribution, control and monitoring systems for complex, high-availability infrastructure. It reports one operating segment. S2
Market fit: the same medium-voltage and control capabilities serve utilities, LNG, petrochemicals and increasingly behind-the-meter data centers—markets where reliability, engineering depth and schedule coordination matter more than commodity hardware. S1S4
Liquidity: cash and short-term investments were $633.6 million. The last filed quarter showed no revolver borrowings; the current release did not provide a debt detail or cash-flow statement. S1S2
Equity value: approximately $6.92 billion using the frozen after-hours price and 36.43 million shares; approximate enterprise value is $6.29 billion using Q3 cash and no financial debt. Both are analyst calculations pending the 10-Q. S1S5S6J1
Solid operating delivery, but the consensus bar was higher and the cash-quality bridge is missing.
Operating margin is the cleanest disclosed recurring-profitability lens. GAAP net income includes material interest income, while contract-estimate catch-ups require the missing 10-Q.
| Metric | Q3 FY26 | Q3 FY25 | YoY | Q2 FY26 | Sequential | Read | Source |
|---|---|---|---|---|---|---|---|
| Revenue | $311.740M | $286.273M | +8.9% | $296.615M | +5.1% | $5.142M / 1.6% below pre-print mean. | S1S5 |
| Gross profit / margin | $95.299M / 30.57% | $87.899M / 30.70% | -13 bp | $87.936M / 29.65% | +92 bp | Sequential repair; broadly stable year over year. | S1J1 |
| Operating income / margin | $64.076M / 20.55% | $60.124M / 21.00% | -45 bp | $57.581M / 19.41% | +114 bp | Best disclosed operating-quality anchor. | S1J1 |
| Net income | $52.160M | $48.234M | +8.1% | $45.887M | +13.7% | Includes $5.047M net interest income. | S1 |
| Diluted EPS | $1.42 | $1.32 | +7.6% | $1.25 | +13.6% | $0.067 / 4.5% below pre-print mean. | S1S5 |
| Orders / book-to-bill | $934M / 3.0x | $362M / n.a. | +158% | $490M / 1.7x | +91% | Exceptional bookings, but unusually concentrated. | S1 |
| Backlog | $2.4B | $1.4B | +69% reported | $1.8B | +35% reported | Balances are rounded; use company growth percentages. | S1 |
| Q3 CFO / FCF | Not disclosed | — | — | Q2 YTD CFO $94.811M; FCF $90.950M derived | Not comparable | Do not infer CFO from the cash-balance change. | S1S2J1 |
GAAP EPS quality bridge
$5.047M net interest × (1 - 24.54% tax rate) ÷ 36.604M diluted shares.
Analyst quality lens—not an adjusted EPS figure and not a claim that interest is non-recurring.
Q2 filing disclosed a $14.2M favorable first-half operating impact from contract-estimate changes.
Balance-sheet movement
Cash and short-term investments rose $88.672M sequentially to $633.561M, while current liabilities rose $177.370M and other current assets rose $134.113M. Working capital rose to $606.521M, but the current ratio declined to 1.97x from 2.25x. S1J1
Without contract-asset, customer-advance, inventory, receivable and payable detail, none of these movements should be labeled sustainable cash generation.
What is estimate-relevant
Nine-month revenue is $859.539M and diluted EPS is $3.81. Adding the pre-print Q4 mean produces $1.191B revenue and $5.43 EPS—about 0.5% and 1.2% below the respective pre-print full-year means. S1S5J1
The likely direction is a modest FY26 reduction absent call upside; the magnitude cannot be known until estimates refresh.
Margins have held near 20%; bookings have broken away from shipment growth.
Fiscal quarter labels are used throughout. Estimates are pre-print means retrieved August 3; their latest annual revision point was July 23, so they are not a post-earnings consensus.
Eight-quarter operating trajectory
Revenue, gross profit and operating income in USD millions; operating margin on the right axis. S1S5
Accessible data table
| Quarter | Revenue | Gross profit | Operating income | Operating margin |
|---|---|---|---|---|
| Q4 FY24 | $275.063M | $80.434M | $56.116M | 20.40% |
| Q1 FY25 | $241.431M | $59.524M | $35.572M | 14.73% |
| Q2 FY25 | $278.631M | $83.432M | $58.919M | 21.15% |
| Q3 FY25 | $286.273M | $87.899M | $60.124M | 21.00% |
| Q4 FY25 | $297.983M | $93.526M | $63.245M | 21.22% |
| Q1 FY26 | $251.184M | $71.418M | $42.771M | 17.03% |
| Q2 FY26 | $296.615M | $87.936M | $57.581M | 19.41% |
| Q3 FY26 | $311.740M | $95.299M | $64.076M | 20.55% |
Orders, backlog and book-to-bill
Orders and backlog in USD millions; backlog is period-end and company-rounded. S1S3S7
Accessible data table
| Quarter | Orders | Backlog | Book-to-bill |
|---|---|---|---|
| Q1 FY26 | $439M | $1.6B | 1.7x |
| Q2 FY26 | $490M | $1.8B | 1.7x |
| Q3 FY26 | $934M | $2.4B | 3.0x |
Ten-quarter revenue and earnings view
Six reported quarters plus four future pre-print estimate quarters. Calendar-period provider data were relabeled to Powell's September fiscal year-end. S1S5
Accessible data table
| Quarter | Status | Revenue | Diluted EPS |
|---|---|---|---|
| Q2 FY25 | Reported | $278.631M | $1.19 |
| Q3 FY25 | Reported | $286.273M | $1.32 |
| Q4 FY25 | Reported | $297.983M | $1.407 |
| Q1 FY26 | Reported | $251.184M | $1.133 |
| Q2 FY26 | Reported | $296.615M | $1.25 |
| Q3 FY26 | Reported | $311.740M | $1.42 |
| Q4 FY26 | Pre-print mean | $331.704M | $1.619 |
| Q1 FY27 | Pre-print mean | $276.072M | $1.255 |
| Q2 FY27 | Pre-print mean | $327.351M | $1.45 |
| Q3 FY27 | Pre-print mean | $370.230M | $1.835 |
EPS surprise: the cadence changed
Three beats followed by two misses. Estimates are historical pre-print means; EPS is 3-for-1 split-adjusted. S1S5
Accessible data table
| Quarter | Estimate | Actual | Surprise |
|---|---|---|---|
| Q3 FY25 | $1.257 | $1.32 | +5.0% |
| Q4 FY25 | $1.26 | $1.407 | +11.6% |
| Q1 FY26 | $1.013 | $1.133 | +11.8% |
| Q2 FY26 | $1.363 | $1.25 | -8.3% |
| Q3 FY26 | $1.487 | $1.42 | -4.5% |
Three-year growth estimates consensus
Fiscal 2026-2028 are pre-print means—not guidance and not yet refreshed after Q3. S5
Accessible data table
| Fiscal year | Basis | Revenue | Revenue growth | EPS | EPS growth |
|---|---|---|---|---|---|
| 2025A | Reported | $1.104B | 9.1% | $4.953 | 20.9% |
| 2026E | Pre-print mean | $1.197B | 8.3% | $5.493 | 10.9% |
| 2027E | Pre-print mean | $1.426B | 19.2% | $6.875 | 25.2% |
| 2028E | Pre-print mean | $1.651B | 15.8% | $8.132 | 18.3% |
The language improved; the guide did not become numeric.
A semantic upgrade is evidence of confidence, not a forecast range. Margin language can be quantified only as an analyst-derived reference.
| Topic | Prior Q2 posture | Current Q3 posture | Delta | Decision use | Source |
|---|---|---|---|---|---|
| Fiscal 2026 | Expected “solid” performance. | Expected “another very strong year.” | Confidence improved | Qualitative; no revenue or EPS range. | S1S3 |
| Fiscal 2027 | Demand visibility discussed through FY28 for the large data-center award. | Confidence explicitly extends into FY27. | Duration strengthened | Conversion timing remains undisclosed. | S1S4 |
| Gross margin | Expected consistent with FY25 performance; FY25 was 29.37% derived. | Expected consistent with trailing-twelve-month levels; about 30.08% derived. | Implied reference +71 bp | Inference only; not a formal guide lift. | S1S3J1 |
| Capacity | Houston/Ohio leases and greenfield alternatives discussed. | Jacintoport completion expected by fiscal year-end; production ramp thereafter; greenfield evaluation continues. | Execution moved forward | Timing, capex and returns remain unquantified. | S1S4 |
| Cash / capex | Q2 YTD CFO and capex filed. | No Q3 CFO, FCF or full-year capex range. | Disclosure gap | Working-capital underwriting remains open. | S1S2 |
POWL is becoming an AI-power story—but it remains a custom-project manufacturer.
The market opportunity is real. The underwriting hinge is whether engineering, labor, suppliers and project controls can scale without losing margin or tying up disproportionate cash.
Behind-the-meter data centers
The more-than-$400M award validates Powell's position outside the data-center building, connecting on-site generation to campus power. It was announced after Q2 and booked in Q3; it is not a second award. S1S3S4
Hinge: milestone execution and follow-on phases, neither of which is guaranteed.
Utility and grid investment
Electric Utility revenue grew 18% year over year as grid hardening, generation additions and load growth support medium-voltage investment. S1
Hinge: utilities broaden the demand base but still carry permitting, schedule and equipment constraints.
LNG and process industries
Q3 included an approximately $60M LNG export award and an approximately $75M petrochemical/fertilizer order, even as Petrochemical revenue fell 49% year over year. S1
Hinge: bookings can be strong while revenue remains lumpy.
Concentration math
The three identified mega-orders total more than $535M, or more than 57.3% of Q3 bookings; the data-center order alone was more than 42.8%. These are lower bounds because the award was described as above $400M. S1J1
High concentration is not automatically bad: a large project can improve factory loading and engineering reuse. But it increases customer, schedule, cancellation, change-order and working-capital sensitivity.
Capacity flywheel—or bottleneck
Management previously said people and supply chain would be the main constraints on another mega-award, and acknowledged it cannot meet every requested schedule. Leased capacity can bridge demand faster than a greenfield plant; training and engineering productivity decide whether capacity becomes profitable throughput. S4
The next useful disclosures are schedule adherence, utilization, rejected/deferred bids, engineering cycle time and capex returns.
Competitive comparison: where Powell is advantaged and exposed
| Dimension | POWL positioning | Versus diversified electrical peers | Underwriting implication | Evidence |
|---|---|---|---|---|
| Engineering / customization | High-touch, project-specific systems and integration. | Potentially stronger fit for complex power architectures; less standardized volume. | Supports differentiation but raises execution and labor risk. | S2S4J1 |
| Installed capacity | Multi-plant North American footprint expanding through leases and Jacintoport. | Smaller scale than global diversified peers; more flexibility than one-site specialists. | Upside is capacity-constrained; fixed-cost ramp matters. | S1S4 |
| End-market breadth | Utilities, oil and gas, LNG, petrochemical and data centers. | Narrower product portfolio but cross-market use of the same engineering core. | Breadth reduces pure-play risk; projects remain cyclical and lumpy. | S1S2 |
| Pricing / supply chain | Copper hedging and selective price increases; steel, aluminum and components actively managed. | Less broad procurement scale than major peers. | Contract terms and pass-through protections are critical. | S2S4 |
| Disclosure | One segment; limited customer, project-margin and backlog-timing detail. | Less granular than many diversified industrial disclosures. | Requires a wider margin of safety around forward earnings. | S2J1 |
The Q3 call is missing; the prior-call promises define the test.
Every item below is explicitly prior-call evidence from May 5, not fresh Q3 commentary. The right-hand column converts it into a falsifiable next check.
Questions management must answer on August 4
- What portion of the $2.4B backlog converts in FY27 versus later?
- How much of the $400M+ data-center PO remains, and what milestones or cancellation protections apply?
- How many distinct data-center customers and funded campuses support backlog?
- How much cash will contract assets, inventory, labor and supplier prepayments consume?
- Can gross margin hold near TTM levels without favorable project closeouts?
- How many opportunities are delayed because Powell cannot meet requested schedules?
- Will management give numeric FY27 revenue, margin, capex or FCF parameters?
Answer-quality rubric
High quality: quantified conversion years, customer/project bounds, cash bridge, margin-normalization bridge and capacity milestones.
Medium quality: directional cadence plus at least one measurable range.
Low quality: repeating strong demand, confidence or “visibility” without timing, economics or cash requirements.
This rubric prevents a strong narrative from being mistaken for a de-risked earnings path. J1
The stock had already capitalized much of the AI-power option.
Daily closes are split-adjusted. The chart ends at the August 3 regular close; the $189.90 after-hours observation is a separate timestamped snapshot.
USD daily close with zoom/pan and event markers. Price data are market observations, not evidence of intrinsic value. S6
Print-day sequence
Prior close: $208.68. August 3 regular close: $219.72, up 5.29%. Frozen after-hours: $189.90 at 18:45:45 ET, down 13.57% from the regular close and 9.00% from the prior close. S6
The regular session preceded the 16:28 ET filing. After-hours trading is thinner and volatile; the move is a sentiment signal, not a stable valuation anchor.
Event interpretation
The >$400M data-center award was first disclosed on May 4 and entered Q3 bookings. The incremental Q3 information is the full $934M order total, the other two mega-orders and the resulting $2.4B backlog—not a second data-center win. S1S3
The three-for-one split changed per-share presentation, not enterprise economics; all comparative EPS figures here are split-adjusted. S1S6
| Date | Event | What changed | What did not | Source |
|---|---|---|---|---|
| May 4, 2026 | Q2 results and $400M+ data-center award | Validated behind-the-meter data-center exposure and multi-year duration. | The award had not yet been booked at Q2-end. | S3 |
| July 6, 2026 | CEO retention RSU award | 36,000 restricted stock units awarded, vesting through 2030 subject to service. | Not an open-market purchase and not an earnings signal. | S8 |
| August 3, 2026 | Q3 results | Record orders/backlog; data-center award booked; two other mega-orders disclosed. | No numeric guide, current cash flow, filing detail or call Q&A. | S1 |
| August 4, 2026, 11:00 ET | Scheduled earnings call | Potentially resolves conversion, concentration and cash questions. | Call outcome unknown at this report's cut-off. | S9 |
Sector and portfolio read-throughs
Electrical equipment
Analyst inference: the order validates continued demand for complex medium-voltage distribution around power generation and AI campuses. It is directionally supportive for electrical peers, but it does not prove their share, price or margins. S1J1
The reset removed excess, not execution risk.
All multiples use the frozen $189.90 after-hours observation and pre-print denominators. They are provisional because post-print estimates have not refreshed.
Current valuation bridge
| Metric | Value | Formula / caveat | Source |
|---|---|---|---|
| Equity value | $6.92B | $189.90 × 36.431M shares | S5S6J1 |
| Approx. enterprise value | $6.29B | Equity value - $633.561M cash; no financial debt assumed from last filing | S1S2J1 |
| TTM P/E | 36.4x | $189.90 ÷ approx. $5.21 TTM EPS | S1S5J1 |
| NTM P/E | 30.8x | $189.90 ÷ $6.159 pre-print rolling EPS | S5S6J1 |
| NTM EV / EBITDA | 22.2x | $6.29B ÷ $283.533M pre-print NTM EBITDA | S5J1 |
| NTM P / FCF | 49.6x | $6.92B ÷ $139.491M pre-print NTM FCF | S5J1 |
| FY26E P/E | 34.6x | $189.90 ÷ $5.493 pre-print mean | S5S6J1 |
| FY27E P/E | 27.6x | $189.90 ÷ $6.875 pre-print mean | S5S6J1 |
| FY28E P/E | 23.4x | $189.90 ÷ $8.132 pre-print mean | S5S6J1 |
| EV / TTM sales | 5.4x | $6.29B ÷ $1.158B TTM revenue | S1S5J1 |
What is priced in
At 27.6x pre-print FY27 EPS, the market still assumes that backlog converts with limited margin leakage and that FY27 earnings grow roughly 25%. S5J1
Potentially mispriced upside: repeatable data-center designs, follow-on campuses and faster capacity productivity could push earnings toward the high end or beyond.
Potentially mispriced downside: customer concentration, fixed-price true-ups, labor/supply bottlenecks or working-capital absorption could make the backlog less valuable than its headline size.
The after-hours decline improves the entry math, but the bear/base sensitivity still says the expected value is highly dependent on delivery.
Forward-multiple peer context
| Company | NTM P/E | NTM EV / EBITDA | Comparability note | Source |
|---|---|---|---|---|
| POWL at frozen after-hours price | 30.8x | 22.2x | Custom-project exposure; current Q3 filing and post-print estimates pending. | S5S6J1 |
| Eaton | 29.3x | 23.0x | Diversified electrical and aerospace; much larger scale. | S5 |
| Vertiv | 33.7x | 24.9x | Higher-purity data-center power/cooling, different service and product mix. | S5 |
| nVent | 27.6x | 19.8x | Electrical connection/protection and cooling exposure; acquisition mix differs. | S5 |
| Hubbell | 22.0x | 16.7x | Utility/electrical benchmark with lower project concentration. | S5 |
| Four-peer median | 28.4x | 21.4x | Context, not a mechanically correct valuation anchor. | S5J1 |
Provider snapshots retrieved August 3; timestamp and underlying estimate-vendor identity were not exposed. POWL is around a 9% NTM P/E premium and a 4% NTM EV/EBITDA premium to the median at $189.90. Peer business mixes are materially different. S5J1
Illustrative FY2027 EPS / multiple sensitivity — not price targets
$6.10 low-end pre-print EPS × 22x = $134.20, about 29% below the frozen after-hours price. Assumes conversion or multiple disappointment. S5J1
Formula: FY2027 EPS assumption multiplied by the stated P/E. Scenario multiples are analyst assumptions; the exercise ignores future cash, debt and share-count changes and is not a formal target-price model. J1
The next evidence must turn visibility into conversion.
Dated events are separated from inferred monitoring windows. The underwriting can strengthen only through observable delivery, not stronger adjectives.
Dated / bounded catalysts
- August 4, 2026, 11:00 ET: Q3 call; conversion, margin, cash and capacity questions. S9
- Near-term SEC filing window: Q3 10-Q; contract-estimate changes, liquidated damages, working capital, debt and backlog detail.
- Fiscal year-end: expected completion of Jacintoport fabrication expansion, followed by production ramp. S1
- FY2027 reporting: first measurable conversion of the expanded backlog and capacity base.
Observable falsifiers
- FY27 backlog conversion is materially back-end loaded without customer protections.
- Gross margin falls below the roughly 30% TTM reference because of unfavorable project true-ups or mix.
- Working capital absorbs cash faster than operating income grows.
- Data-center follow-on phases do not materialize, or concentration remains unquantified.
- Jacintoport or leased-capacity ramps slip, or utilization costs dilute margin.
- Two consecutive EPS misses extend into another quarter without an offsetting guide reset.
KPI watch list
- Backlog conversion by fiscal year and end market.
- Order-size distribution and customer concentration.
- Gross and operating margin excluding favorable contract catch-ups.
- Contract assets, customer advances, inventory and operating cash flow.
- Engineering headcount/productivity, supplier lead times and on-time delivery.
- Capacity capex, commissioning dates and incremental revenue/margin.
- Post-print FY2026-FY2028 consensus revisions.
Open evidence and research status
Evidence confidence: medium-high for the print; medium for the forward view. SEC-filed results and historical statements were checked. The current call, transcript, presentation and 10-Q were unavailable.
Research posture: wait for proof / do not chase. Reopen the report after the call and filing; then refresh the cash-quality bridge, Q&A debate map, conversion schedule and valuation denominators.
Central thesis: Powell has scarce engineering exposure to several high-growth power markets. The stock becomes attractive only when backlog duration is paired with credible recurring margin and cash conversion at a price that compensates for project concentration.
Source register, assumptions and limitations
Primary filings control reported financials. The prior transcript supports call-only narrative. Fiscal.ai and Yahoo observations are timestamped and labeled. Analyst calculations are separated as J1.
Filed August 3, 2026; SEC acceptance 16:28:02 ET. Current results, balance sheet, orders, backlog, end-market growth, capacity update and qualitative outlook.
Filed May 7, 2026. Historical statements, cash flow, working capital, contract-estimate changes, liquidated damages, revolver status, commodity/tariff risks and segment disclosure.
Filed May 4, 2026. Prior outlook and first disclosure of the more-than-$400M data-center award.
Event 551945, May 5, 2026, 107 segments. Prepared remarks and full Q&A reviewed. Used only as prior-call evidence; no company-authored transcript was found and speech-to-text artifacts are possible.
POWL identifier NasdaqGS-POWL; retrieved August 3, 2026. Historical statements, share count, pre-print revenue/EPS/EBITDA/FCF means, and ETN/VRT/NVT/HUBB peer snapshots. Latest POWL annual revision point was July 23; estimates were not post-print. Peer snapshot timestamps and underlying estimate-vendor identity were not exposed.
Daily split-adjusted history through the August 3 regular close and one-minute extended-hours snapshot through 18:45:45 ET. After-hours quotes can be thin and volatile.
Orders, backlog, book-to-bill, cash and no-debt statement used for fiscal 2026 demand trajectory.
CEO retention award details; distinguished from open-market insider buying.
Q3 webcast timing and current earnings-material inventory checked August 3. The page listed the Q3 release and webcast, not a Q3 presentation or transcript.
Comparative market tables and historical statement cross-check.
Formulas are shown near relevant sections: growth and margins from exact reported values; interest-income EPS lens; TTM values; equity and enterprise value; P/E and EV/sales; order concentration; residual fiscal-year math; and EPS/multiple sensitivities. These are not company guidance or consensus.