PWR — Quanta Services, Inc.

The news flow is unambiguously bullish: Q2 2026 (reported July 30, 2026) delivered revenue up 41.1% YoY to a record $9.56B, adjusted diluted EPS of $4.24 (+71% YoY) versus ~$3.31 consensus, a record $53.4B total backlog, and a raised FY2026 guide of $39.3–39.7B revenue and $16.45–16.95 adjusted EPS. Piper Sandler upgraded to Strong Buy in September 2026, the analyst consensus sits at Buy (26 buy / 10 hold / 0 sell), and the consensus price target of $789.38 implies ~21% upside. The tension: as of 2026-09-13 the stock was still -8.08% over three months and -4.05% over one month, meaning the market is actively debating how much of the AI/data-center grid supercycle is already capitalized into the price rather than celebrating the beat.

Revenue has compounded steadily from $5.62B in Q3 2023 to $9.56B in Q2 2026, but the more important fact is that YoY growth is accelerating, not just positive: +15.6% in Q3 2025, +19.6% in Q4 2025, +26.3% in Q1 2026, then +41.1% in Q2 2026. That is a genuine inflection, not a base effect. Margins confirm it: Q2 2026 operating income of $694.8M on $9.56B revenue is a 7.27% operating margin versus 5.77% in Q2 2025 — about 150bps of YoY expansion — and operating income grew 78% against revenue growth of 41%, i.e. real operating leverage. The seasonal pattern is intact (Q1 is always the weakest quarter: 4.24% operating margin in Q1 2026, 3.92% in Q1 2025) so the Q1 dip is not deterioration. One soft spot: Q4 2025 operating margin of 6.19% came in slightly below both Q3 2025 (6.89%) and Q4 2024 (6.80%), so the margin story only truly inflected in the most recent quarter. Critical caveat for anyone reading the reported numbers: GAAP diluted EPS of $2.96 in Q2 2026 compares to adjusted EPS of $4.24 — a $1.28 gap driven by intangible amortization from the company's roll-up M&A strategy. The TTM GAAP EPS of $8.906 used to compute the 73.6x trailing P/E materially understates cash earnings power.

Expensive on trailing numbers, defensible on forward numbers, and rich versus its own history. As of 2026-09-13: 73.6x TTM GAAP EPS of $8.906 versus a 5-year average P/E of 45.94x; 2.98x TTM sales versus a 5-year average P/S of 1.63x; 10.14x book; 34.7x trailing EV/EBITDA per the ratio screener. But run it forward on management's own FY2026 adjusted EPS midpoint of $16.70 and the P/E is ~39x; on my FY2027 estimate of ~$19.55 it is ~33x. The trailing PEG of 2.117 looks punitive, but that uses trailing GAAP earnings against trailing growth — on forward adjusted growth (roughly +57% in 2026 per the guidance raise) the forward PEG is well under 1.0. Verdict: this is a premium-priced structural grower, not a value stock. The multiple is justified only if the AI/grid capex cycle keeps compounding, and it offers essentially no cushion if growth normalizes.

The stock is pricing in roughly 1.5–2 years of the current growth rate already. At $650.58 and ~39x FY2026E adjusted EPS, PWR is not waiting for proof — it has pre-paid for mid-teens-plus EPS growth into 2027–2028. That reframes what matters: a beat in Q3 2026 is largely expected and may not move the stock much (note that a blowout Q2 was followed by a -8.08% three-month return). The real swing factor is the forward book. Watch YoY backlog growth (currently a record $53.4B) and 2027 IOU/hyperscaler capex commentary. If backlog growth decelerates below roughly 10% YoY, the multiple compresses faster than EPS grows and the price falls even on good results. If backlog keeps compounding at 15%+, EPS growth alone carries the stock higher. Treat consensus targets near $789 as a growth-delivery scenario, not a base case.

As of 2026-09-13, RSI(14) was 53.37 — squarely neutral, neither overbought nor oversold — with price +2.02% versus the 20-day SMA, +0.61% versus the 50-day SMA, and +9.93% versus the 200-day SMA. That is a mildly constructive, unextended setup: the long-term trend is up, but the stock has done nothing for three months (1-month -4.05%, 3-month -8.08%) and is coiling. The +5.1% session into 2026-09-13 (off a $618.73 prior close) suggests buyers defended the $615–620 zone. Practical read: there is no urgent entry signal here, and no breakdown signal either. The 50-day SMA (~$647 at that date) is the near-term pivot; a decisive close below the 200-day SMA (~$592) would confirm that multiple-compression fear is winning and would be the first real technical warning. A push back above ~$700 with RSI holding under 70 would be the cleaner continuation setup.

• AI and hyperscale data-center power demand is the dominant driver — every incremental gigawatt of compute requires transmission upgrades, substations, and interconnection work, and hyperscaler capex announcements effectively set the ceiling on Quanta's opportunity set.
• Utility capital expenditure cycles: investor-owned utility rate cases and multi-year capex plans are the baseline for the backlog; federal transmission permitting reform (or its absence) determines how quickly that capex converts into awarded contracts.
• Interest rates: higher rates raise utility cost of capital and can defer discretionary projects. So far the mission-critical nature of grid hardening and data-center interconnection has overridden that headwind, but a sustained move higher in long yields is a real brake.
• Craft labor availability and wage inflation — skilled lineworkers and electrical trades are the industry's binding constraint; Quanta's scale, training pipeline, and contractual escalators let it pass through wage inflation, but labor scarcity caps how fast revenue can actually be executed.
• Sector rotation and multiple risk: as a long-duration, high-multiple industrial, PWR is exposed to any rotation out of growth/momentum names or a broad de-rating of AI-adjacent infrastructure stocks.

• Q3 2026 earnings (expected late October or early November 2026): the key reads are backlog conversion, adjusted EPS progression against the $16.45–16.95 FY2026 range, and whether management raises guidance for a second time.
• 2027 capital expenditure budgets and investor days from major investor-owned utilities and hyperscalers — the single best forward visibility signal for multi-year backlog.
• Further sell-side revisions following the July 2026 guidance raise, including the Piper Sandler Strong Buy upgrade in September 2026 and any movement of the consensus target (currently $789.38) toward the $940 high end.
• Capital structure and M&A: the August 2026 senior notes offering (long-term debt was $5.42B at 2026-06-30, up from $4.10B at 2024-12-31) funds continued roll-up activity, which is the company's other growth engine alongside organic demand.
• Any acceleration or pause in announced data-center and power-generation projects, which would directly move the backlog growth rate WAll Street is now underwriting.

• Valuation and multiple risk is the biggest one: 73.6x trailing GAAP P/E, 34.7x trailing EV/EBITDA, 10.1x book, and a market cap of $97.8B. There is no valuation cushion — a growth miss gets punished disproportionately.
• Growth normalization: 41.1% revenue growth in Q2 2026 is not a sustainable run rate. If YoY revenue growth decelerates into the single digits, the ~39x forward multiple will not hold.
• Execution and margin risk on fixed-price EPC contracts — cost overruns, weather, and labor inflation could compress the thin 6.2% TTM operating margin (14.4% gross margin) quickly, and the 200bp-separated adjusted figures show how much of reported earnings depends on add-backs.
• Balance-sheet and liquidity: the current ratio of 1.098 is thin, cash of $506M sits against $5.42B of long-term debt, and the acquisition roll-up has steadily increased leverage — funding more deals with debt in a higher-rate environment shrinks the margin for error.
• Customer concentration and cyclicality of the AI capex cycle itself: hyperscaler spending is a consensus trade today, but it is still capital-markets-funded and could pause abruptly, and the TTM-adjusted figures assume sustained demand.
• Earnings-quality/watch-item: the Q2 2026 GAAP-to-adjusted EPS gap ($2.96 vs $4.24) means the headline 73.6x trailing P/E and the 34.7x EV/EBITDA both overstate how expensive the business truly is — but it also means the bull case depends on investors continuing to accept the adjusted framing.

This is a genuinely excellent business at a genuinely full price as of 2026-09-13. The operating story is as good as it gets — accelerating revenue (+41.1% YoY in Q2 2026), expanding margins, a record $53.4B backlog, and a raised guidance range — and its position at the center of the AI/data-center and grid-modernization buildout gives it multi-year visibility most industrials can only dream about. But my own blended fair value lands around $672 versus a $650.58 spot price, roughly 3% upside, which means the market has already paid for the growth. Own it as a core structural-infrastructure position for the multi-year theme, not as a value entry, and size it with the understanding that a deceleration in backlog growth would compress the multiple before it ever shows up in EPS.

Blend of two forward-looking approaches. (1) 60% weight: 34x applied to my FY2027E adjusted EPS of ~$19.55 (FY2026 guidance midpoint of $16.70 grown ~17%, reflecting moderating but still strong backlog-driven growth) = ~$665 — a premium multiple versus the market justified by the structural grid/AI capex cycle and $53.4B backlog, but below the 45.94x 5-year average trailing P/E. (2) 40% weight: 19.5x applied to FY2027E adjusted EBITDA of ~$5.6B = ~$109.2B enterprise value, less net debt of ~$4.92B ($5.42B long-term debt less $506M cash as of 2026-06-30), divided by ~153M diluted shares = ~$682. Weighted result: ~$672, roughly 3% above the 2026-09-13 price of $650.58 and well below the $789.38 sell-side consensus.