EXC — Exelon Corporation

As of 2026-09-13 the tape says the market has soured on the story: EXC is down 13.37% over six months and 6.6% over three, underperforming the XLU utilities ETF, and sits right on top of its 52-week low of $42.58. The trigger was the July 30, 2026 Q2 print — revenue beat, adjusted EPS of $0.43 merely matched, FY guidance of $2.81–$2.91 was reaffirmed, but management cut the data-center/large-load pipeline from 43 GW to 36 GW and imposed stricter Transmission Security Agreements requiring real financial collateral. That removed a chunk of the AI-power optionality that had pushed the stock above $50 in March 2026. Analyst consensus is a cautious Hold (14 Buy / 21 Hold / 2 Sell) with a $48.63 average target, and rising Treasury yields continue to pressure rate-sensitive regulated utilities. The 3.8% dividend yield is doing the heavy lifting on total return, not price.

Revenue is genuinely growing 8–10% YoY on rate-base investment and load: Q2 2026 revenue of $5,967M vs $5,427M in Q2 2025 (+10.0%), and Q1 2026 of $7,242M vs $6,714M (+7.9%), with Q3 2025 also +9.0%. But the earnings translation is where it breaks down. Operating income grew only 5.6% in Q2 2026 ($979M vs $927M) and 4.5% in Q1 2026 ($1,605M vs $1,536M), so operating margin actually compressed — 17.1% to 16.4% in Q2, 22.9% to 22.2% in Q1. Diluted GAAP EPS was flat to the penny: $0.39 vs $0.39 in Q2, $0.90 vs $0.90 in Q1. The reason is visible in the balance sheet: long-term debt climbed from $46.3B (Q3 2025) to $50.3B (Q2 2026) and stockholders' equity from $28.1B to $29.7B, so depreciation and interest on a growing asset base are absorbing the revenue growth before it reaches the per-share line. Free cash flow is structurally negative — three of the last four quarters were negative, TTM roughly -$1.9B (Q3'25 +$163M, Q4'25 -$1,190M, Q1'26 -$634M, Q2'26 -$255M) against $2.2–2.4B of quarterly capex. This is not a one-quarter blip; it is the arithmetic of a wires-and-pipes utility funding a multi-year capital plan with debt and equity issuance. The honest read: top-line growth is real and stable, EPS growth is currently being crowded out, and the trend is decelerating at the earnings line even as revenue accelerates.

Fair, modestly cheap versus its own history, but the discount is earned rather than a mispricing. As of 2026-09-13, EXC trades at 15.75x TTM EPS of $2.716 versus a 5-year average P/E of 17.99 (~12% below its own norm), and 1.756x TTM sales versus a 5-year average P/S of 1.92 (~9% below). On the next-FY consensus EPS of $2.857, the forward P/E is 15.1x. Price/book is 1.49x, EV/EBITDA 10.5x, and the dividend yield is 3.80% ($1.68 annualized on a $43.16 price). The PEG ratio of 3.77 looks alarming but is a poor tool for a rate-base utility whose EPS growth is capped by regulators — it is really just saying growth is ~4–5%, which is true. Against a low-beta (0.389) regulated business with a FERC-authorized transmission ROE near 10.5% and multi-year rate-base growth, ~15x forward earnings is a reasonable-to-slightly-cheap multiple; it is not a bargain because GAAP EPS is flat YoY, FCF is negative, and leverage (debt/equity 1.76x, LT debt/equity 1.69x) means every basis point of long-term rates flows into the funding cost of the capex plan.

This is a clean case of multiple compression, not estimate cuts — and that distinction matters. As of 2026-09-13 the stock sits at $43.16 versus a March 2026 peak above $50, yet full-year guidance was reaffirmed at $2.81–$2.91 on July 30. Revenue is beating consensus (Q2 $5.97B vs $5.66B est.), so the market is not pricing in growth that hasn't shown up; it is pricing OUT the data-center option value after the 43 GW to 36 GW pipeline cut. Translation: the P/E fell from roughly 18x toward 15x while the earnings base held. That cuts both ways. If the collateralized 36 GW queue converts into signed contracts, EPS growth can accelerate above the current 5–7% trend and the multiple re-rates toward $48–$50 without any heroic assumptions. If the pipeline keeps eroding quarter after quarter, then 15x for ~4% GAAP EPS growth with negative free cash flow is not cheap at all — it is roughly fair, and the downside is another leg down toward the $42.58 low. Watch the pipeline conversion rate, not the headline revenue beat.

As of 2026-09-13 the technicals are weak-but-not-washed-out. RSI(14) was 35.97 — approaching oversold but above the 30 threshold that typically marks capitulation. The stock was 2.55% below its 20-day SMA, 5.19% below its 50-day SMA, and 5.95% below its 200-day SMA, meaning it is beneath all three with the moving averages themselves stacked bearishly. Returns were -1.1% over one week, -4.56% over one month, -6.6% over three months, and -13.37% over six months — a persistent, low-volatility grind lower rather than a single violent flush. Price is sitting just $0.58 above the 52-week low of $42.58, which is the line in the sand. Entry-timing implication: there is no reversal signal here. A cautious buyer would want either a genuine oversold flush (RSI below 30) followed by a reclaim of the 20-day SMA, or a decisive close back above the 50-day SMA (roughly the mid-$45s), before treating the downtrend as broken. Buying a stock hugging its 52-week low with flat earnings momentum is catching a falling knife with a 3.8% coupon attached.

• Interest rates and Treasury yields: EXC is a capital-intensive, rate-sensitive regulated utility. Rising long-term yields (driven by revived inflation concerns) have compressed sector multiples and directly raise the cost of the ~86% of 2026 debt financings management has already executed against a $50.3B long-term debt stack.
• Grid strain and load growth: Severe summer 2026 heat pushed PJM Interconnection near capacity limits, and management publicly described the system as operating close to the edge. That is a structural tailwind for transmission, high-voltage buildout, and battery storage — the exact capex EXC earns a regulated return on.
• Data-center demand realignment: The 43 GW to 36 GW pipeline cut, combined with stricter Transmission Security Agreements requiring posted collateral (roughly 4 kW backed by $1B of collateral in one cited arrangement), is filtering speculative projects out of the queue. Near-term narrative negative, longer-term quality positive.
• Regulatory outcomes at the state level: FERC-authorized transmission ROE (~10.5%) is attractive, but actual margin expansion depends on state commissions — the pending BGE electric distribution rate case in Maryland and proposals before the New Jersey BPU will dictate allowed revenue requirements.
• Sector rotation away from regulated utilities: Capital chasing the AI-power theme has favored merchant generators and IPPs with unregulated upside; regulated wires names like EXC have lagged XLU peers as investors pay for torque rather than stability.

• Q3 2026 earnings, expected late October/early November 2026 — the key read is whether FY2026 guidance of $2.81–$2.91 stays intact and whether the 36 GW data-center queue converts at a healthy rate.
• Maryland Public Service Commission ruling on BGE's July 2026 electric distribution rate case, plus New Jersey BPU action on Atlantic City Electric's proposed transmission-connected battery storage project.
• PJM and MISO competitive transmission project awards, where Exelon is bidding alongside partners such as Invenergy — a win adds visible rate-base growth at the FERC ROE.
• Announcement of additional collateralized large-load/data-center contracts; each signed deal re-establishes the growth narrative that was damaged in July 2026.
• A sustained decline in long-term Treasury yields, which would mechanically lift the entire regulated utility sector's multiple and close part of EXC's discount to its own 5-year average P/E of 17.99.

• Further data-center pipeline erosion: the queue already fell from 43 GW to 36 GW in one quarter. Another cut would confirm the July 2026 announcement was a trend rather than a one-time filter, and would likely take out the $42.58 52-week low.
• Funding cost and leverage: $50.3B of long-term debt against $29.7B of equity (debt/equity 1.76x), roughly $1.9B of negative TTM free cash flow, and a current ratio of 0.999. Refinancing a capex-heavy plan into a higher-rate environment pressures earned ROE and per-share earnings regardless of how well the underlying assets perform.
• Adverse rate-case outcomes in Maryland, Illinois, or New Jersey — disallowances or allowed ROE cuts would directly impair the rate-base growth model that justifies the current multiple.
• Earnings quality gap: 8–10% revenue growth is translating into exactly zero GAAP EPS growth ($0.39 flat in Q2 2026, $0.90 flat in Q1 2026) as depreciation and interest consume the upside. If that persists, the 15x P/E is not cheap.
• Weather, storm restoration costs, and commodity/hedging volatility, which are recurring earnings-swing factors for a six-utility footprint serving ~11 million customers.
• Execution risk on a $9B+ annual capex program: overspending or schedule slippage delays the rate-base growth that the entire investment case depends on.

Exelon is a decent regulated utility but not a compelling buy as of 2026-09-13. At $43.16 you are paying 15.1x next-year EPS for a 3.80% dividend yield and a business that grew revenue 10% YoY in Q2 2026 yet delivered literally flat GAAP EPS of $0.39 — because $50.3B of debt, $1.9B of negative trailing free cash flow, and rising depreciation are eating the rate-base growth before it reaches shareholders. The stock has de-rated from above $50 in March 2026 to within $0.58 of its 52-week low while guidance held, which tells you the market repriced the data-center option value after the pipeline was cut from 43 GW to 36 GW, not that fundamentals collapsed. My calculated fair value of roughly $47 implies about 9% price upside plus the dividend — real, but not enough to compensate for a downtrend that is still intact (RSI 35.97, price below the 20-, 50- and 200-day SMAs) and for the very real possibility that the pipeline shrinks again. This is a hold-with-a-cautious-tilt, best bought either closer to the $42.58 floor or after a rate-case win or a fresh data-center contract proves the growth story.

Method: justified forward P/E applied to next-fiscal-year consensus EPS. I used next-FY estimated EPS of $2.857 (consistent with management's reaffirmed FY2026 guidance midpoint of $2.86) against a 16.5x multiple — a deliberate discount to EXC's own 5-year average trailing P/E of 17.99, because GAAP EPS has been flat YoY for two consecutive quarters and long-term rates are pressuring the funding cost of the capex plan; 16.5x is also consistent with the stock's current 15.75x TTM. 16.5 × $2.857 = $47.14. Cross-check: applying the 5-year average P/S of 1.92 to next-FY consensus revenue of $25.43B implies a ~$48.8B market cap, or about $47.4 on ~1.03B shares — within a dollar of the P/E-derived figure, which gives me reasonable confidence in the ~$47 level.