PEG — Public Service Enterprise Group Incorporated

The tape is unambiguously bearish: as of 2026-09-13 PEG had fallen 1.78% in a week, 4.59% in a month, 9.17% in three months, and 13.54% in six months, closing at $72.39 against a 52-week low of $72.08. This is a de-rating, not a broken business — Q2 2026 non-GAAP operating EPS rose to $0.86 from $0.77 a year earlier and beat the ~$0.80 consensus, and FY2026 guidance was reaffirmed. The market is instead pricing the structural negatives: rate sensitivity for a capital-hungry utility funding a $22.5–25.5B (2026–2030) capex program, potential equity dilution, and thin free cash flow (FCF was -$408M in Q4 2025). Analyst ratings remain a 16 Buy / 16 Hold / 1 Sell mix with a $83.5 consensus target, so sentiment is split between a constructive sell-side and a market that keeps selling the stock — the sellers are currently winning.

Revenue is lumpy, not trending: quarterly revenue ranged from $2,423M (Q2 2024) to $3,848M (Q1 2026), with Q1 2026 up 19.4% YoY to $3,848M but Q2 2026 down 8.9% YoY to $2,554M — swings driven by weather, power prices, and one-offs rather than a structural decline. The cleaner signal is the GAAP-vs-adjusted split: Q2 2026 GAAP diluted EPS fell to $0.67 from $1.17 a year earlier, yet non-GAAP operating EPS rose to $0.86 from $0.77 (+11.7% YoY), and FY2026 operating guidance of $4.28–$4.40 was reaffirmed against next-FY consensus EPS of $4.368. Operating margin has been two-sided (roughly 27.9% in Q1 2026 versus 18.1% in Q2 2026), and full-year margin reads are distorted by data artifacts (e.g., Q4 2025 gross profit of $4,988M exceeding revenue), so margins are best characterized as stable-to-choppy rather than expanding or collapsing. Bottom line: the latest quarter's reported revenue and GAAP EPS declines look like a blip around a stable, modestly growing adjusted-earnings base — but the underlying growth (~7–8%) is mid-single-digit utility growth, not acceleration.

Cheap-to-fair versus its own history, not versus the market. As of 2026-09-13, PEG traded at 17.9x trailing EPS of $4.03 versus a 5-year average P/E of 21.05x, and roughly 16.6x forward EPS of $4.368 — a genuine discount to its own norm. Price/sales of 2.88x sits below the 5-year average of 3.33x, P/B is 2.09x, and the 3.64% dividend yield is elevated relative to recent history, all consistent with a depressed price. The caveat is real: a trailing PEG ratio of 8.87 and ROIC of just 4.9% show that neither growth nor returns on capital justify a premium multiple, which is precisely why it has de-rated. So the 'cheapness' is partly a legitimate reflection of rate sensitivity and capex-financing risk, not purely a mispricing — cheap versus itself, but not obviously cheap versus a low-growth-utility peer set.

As of 2026-09-13, price is falling faster than estimates: the stock is down 13.54% over six months to $72.39 even though Q2 2026 adjusted EPS beat consensus ($0.86 vs ~$0.80) and guidance was reaffirmed. That gap is the whole thesis question. Either the market is correctly pre-pricing dilution and financing costs tied to the $22.5–25.5B capex program (a structural de-rating), or it is underpricing a reaffirmed ~8% adjusted-EPS growth trajectory at a forward P/E of roughly 16.6x. The trade to watch is estimate revision direction: if next-FY EPS of $4.368 holds, the multiple is the only lever and re-rating to even 19x would imply ~$83; if estimates slip, the low P/E is a mirage and the stock can keep falling. Practical read: this is pricing in more skepticism than the confirmed numbers currently justify, but the burden of proof is on the company to show data-center load and FCF conversion, not on the buyer to hope.

Clean downtrend with RSI approaching — but not at — oversold. As of 2026-09-13, RSI(14) was 34.52, price sat 1.95% below the 20-day SMA, 5.43% below the 50-day, and 9.26% below the 200-day, with 1-week/-1-month/-3-month/-6-month returns of -1.78%/-4.59%/-9.17%/-13.54%. The close of $72.39 is essentially at the 52-week low of $72.08. Implication: there is no trend-reversal signal yet — RSI in the mid-30s leaves room for a further flush below 30 (true oversold) before any durable bounce, so there is no rush to catch the falling knife. A capitulation low near the $72 round number would be a more defensible mean-reversion entry for patient buyers than the current mid-trend level.

• Interest rates: as a capital-intensive utility funding a $22.5–25.5B (2026–2030) capex program, PEG's valuation and cost of capital are highly rate-sensitive; stable-to-lower rates would relieve both the financing burden and the multiple, higher-for-longer rates would extend the de-rating.
• AI/data-center power demand: roughly 11,800 MW of large-load inquiries sit in PEG's PJM footprint, and carbon-free nuclear generation positions it as an indirect AI-infrastructure play — a genuine structural tailwind, though mostly prospective rather than in current earnings.
• New Jersey regulation (NJBPU): timely recovery on grid modernization, energy efficiency, and gas system modernization (GSMP) filings is essential to protecting ROE; adverse rate-case outcomes would cap earnings growth.
• Sector rotation: defensives and rate-sensitive utilities have been out of favor versus growth/AI names, and PEG's 0.52 beta and 3.64% yield make it a candidate for further rotation pressure if risk appetite stays strong.
• PJM power and capacity prices: wholesale energy and capacity market outcomes directly affect PSEG Power's nuclear margins, adding a commodity-driven swing factor on top of the regulated utility base.

• Q3 2026 earnings on ~November 3, 2026 — the key check on FY2026 guidance adherence, data-center load-growth tracking, and whether the GAAP/non-GAAP gap narrows.
• Announcement of long-term power purchase agreements (PPAs) between the nuclear fleet and hyperscalers — a positive re-rating catalyst if signed at attractive terms.
• NJBPU decisions on grid modernization and GSMP filings, which determine the pace of rate-base recovery and ROE support.
• Federal Reserve rate path — easing would relieve the cost-of-capital pressure driving much of the current de-rating.
• Progress/updates on the GridSmart Battery Program and other grid-modernization capex, which support the multi-year rate-base growth story.

• Rate sensitivity and financing risk: the large capex program likely requires debt and equity issuance, and higher-for-longer rates both raise the cost of capital and pressure the multiple — a scenario the 6-month 13.54% decline may already partly price.
• Equity dilution from funding the $22.5–25.5B capex plan, which would dilute the per-share earnings growth that underpins the forward P/E.
• Regulatory risk: unfavorable NJBPU rate-case or cost-recovery outcomes could squeeze allowed returns and earnings growth.
• Thin and volatile free cash flow (e.g., -$408M in Q4 2025 and only +$81M in Q1 2026), which means the dividend and capex are increasingly reliant on external financing.
• GAAP-vs-adjusted earnings divergence (Q2 2026 GAAP EPS of $0.67 versus adjusted $0.86) and a trailing PEG ratio of 8.87/ROIC of 4.9% — evidence the reported quality of earnings and returns on capital are below what a premium multiple would require.

PEG is a low-beta (0.52), 3.64%-yielding regulated utility with real AI/nuclear optionality (11,800 MW of PJM large-load inquiries) that has been de-rated to a discount versus its own history and is sitting at a 52-week low. The honest problem: the deterministic score flags weak trailing fundamentals (ROIC 4.9%, weak revenue growth, lumpy/negative FCF) and the tape is unambiguously bearish, while the $83.5 bullish consensus and reaffirmed guidance have so far failed to stop the selling. That is the textbook setup for either a value opportunity or a value trap, and right now the evidence — negative FCF in several quarters, a capital-hungry capex program, rate sensitivity, and no technical reversal — tilts toward the trap until the cash flow or the trend turns. Not a buy into weakness, and not a clean short given the dividend and defensive profile; a cautious, wait-for-confirmation stance.

I applied a forward P/E of 18.5x to next-FY consensus EPS of $4.368 (18.5 x $4.368 = $80.85). The multiple sits between PEG's depressed current forward P/E of ~16.6x and its 5-year average trailing P/E of 21.05x — a modest premium to the broad utility peer group justified by the nuclear/AI-data-center optionality and reaffirmed ~8% adjusted-EPS growth, but below the historical average to reflect the rate-sensitivity, equity-dilution, and FCF risks that drove the de-rating. The result implies ~12% upside to the 2026-09-13 price of $72.39 and lands just below the $83.5 analyst consensus.