PHM — PulteGroup, Inc.

The tape is the sentiment here: as of 2026-09-13 PHM has lost ground over the 1-week (-4.97%), 1-month (-8.44%), 3-month (-3.99%) and 6-month (-0.78%) windows, trades below all three major moving averages, and RSI(14) of 36.83 signals persistent distribution rather than capitulation. That negativity sits awkwardly against the news flow - the Q2 2026 report on 2026-07-22 beat on EPS (~$2.43-2.48 vs. $2.36 consensus), net new orders rose 6% YoY to 7,536 homes worth $4.1B, home-sale gross margin recovered 60bp sequentially to 25.0%, and Wolfe upgraded to Outperform with a $129 target in late August. The disconnect looks rate-driven rather than company-specific: mortgage-rate volatility is crushing affordability, and investors are extrapolating the 11% revenue and 45% operating-income declines from Q2. Analyst consensus (20 Buys vs. 3 Sells, $145.86 average target) is not bearish at all - the bearishness is entirely in the price action.

The trend is decelerating, and it is not a blip. Revenue peaked at $4.92B in Q4 2024 and has declined YoY every quarter since Q3 2025: -1.6% (Q3'25), -6.3% (Q4'25), -12.4% (Q1'26) and -9.6% (Q2'26, $3.98B vs. $4.40B). Operating margin has compressed relentlessly: 20.6% (Q3'23) to a 23.5% peak in Q4'24, then 17.3%, 18.3%, 17.3%, 16.3%, 12.9% and 11.1% in Q2 2026 - roughly 1,240bp of peak-to-trough erosion in six quarters, with the worst two readings being the last two. Gross margin on FMP's definition fell from 30.3% (Q2'24) to 27.2% (Q2'25) to 20.7% (Q2'26); the company's own home-sale gross margin metric shows the same direction, 27.0% to 25.0%. EPS has followed: Q4 2024 $4.43, Q4 2025 $2.56, Q1 2026 $1.79, Q2 2026 $2.43, down ~20% YoY. Free cash flow was positive in every quarter from 2023 through Q1 2026 (Q4 2025 alone threw off $739.8M) but flipped negative at -$12.9M in Q2 2026 on a $17.1M operating cash flow against $29.9M of capex. The genuinely important counter-signal: net new orders grew 6% YoY, the Q2 EPS beat consensus, gross margin rose 60bp sequentially, and net income ($472.0M) actually exceeded operating income ($442.5M) thanks to non-operating income. So this is a decelerating business showing the first tentative signs of stabilization - not yet an inflection, and Q3 (2026-10-22) is the test.

Fair in absolute terms, rich versus its own history. As of 2026-09-13 the stock trades at 12.03x TTM EPS of $9.944 versus a 5-year average P/E of 7.67x - a ~57% premium - and at 1.39x TTM sales versus a 5-year average of 1.15x, which is why the deterministic valuation overlay tags it 'Rich vs. its own history.' But that historical average is a distorted anchor: PHM earned $14.71 in 2024 and $11.11 in 2025, so the low average P/E was a denominator artifact of peak-cycle earnings, not evidence of chronic cheapness. On forward numbers the picture is more balanced: consensus next-FY EPS of $10.151 puts the stock at 11.65x, with 1.74x book against a ~15% ROE, ~14.3% ROIC and only 0.177x debt-to-equity. The PEG of -0.443 is meaningless with negative trailing growth. These are not distressed multiples for a 14% ROIC builder with a net-debt position of roughly $0.5B, but they are not cheap either - and unlike 2022-2023, there is no deep-value cushion if the 2027 earnings stabilization that consensus assumes fails to materialize.

The key trap here is that headline EPS is flattering a deteriorating operating business. Q2 2026 operating income fell 45% YoY ($805.5M to $442.5M) while EPS fell only ~20% ($3.03 to $2.43), because aggressive buybacks ($373M repurchased in Q2 alone, roughly 6% of market cap annualized) and non-operating income are cushioning the per-share line - net income of $472.0M actually exceeded operating income of $442.5M. Watch operating margin and gross margin, not EPS. On price, the market is paying 11.7x forward earnings for EPS of $10.151 that assumes flat-to-up earnings against a trailing $9.944 that is still falling; that is a bet on stabilization, not a discount for deterioration. If Q3/Q4 2026 confirms margin stabilization and continued order growth, a 13x multiple on $10.15 supports ~$132 and the current price looks like a reasonable entry. If gross margin slips below the 25% area and orders roll over, the multiple compresses toward the 9-10x that builders historically get in a confirmed downcycle, which is $91-102 - so the downside from a missed estimate is roughly twice the near-term upside. The balance sheet and buyback put a floor under the story; the earnings trajectory does not yet put a ceiling on it.

Trending down, with no oversold washout yet. As of 2026-09-13, PHM at $118.26 sits about 5.9% below a ~$125.7 20-day SMA, 7.1% below a ~$127.4 50-day SMA and 5.6% below a ~$125.3 200-day SMA (levels back-solved from the reported percentage gaps), which means overhead resistance is a tight cluster at roughly $125-127 sitting just above the stock. RSI(14) of 36.83 is weak but above the 30 oversold line, so there is still room for another leg lower before a momentum washout - this is not a 'buy the RSI' setup. The 52-week low at $108.49 is roughly 8% below spot and is the level to watch for a retest; a daily close back above $125-127 would be the first credible technical evidence that a bottom is forming. Until then, there is no technical reason to anticipate a turn.

• Mortgage-rate volatility and affordability: the single biggest swing factor for PHM's volumes and pricing, and the proximate cause of the recent de-rating; any sustained move lower in the 30-year rate would be the most powerful positive catalyst available.
• Structural US housing shortage and favorable demographics: years of underbuilding plus household formation continue to favor large national builders with strong balance sheets, letting PHM take share from fragmented regional players even in a soft market.
• Input costs, tariffs and labor: lumber and imported-material costs plus skilled-labor scarcity keep build costs sticky, which matters more when ASPs are already falling 3% YoY and buyers need rate buydowns to close.
• Competitive intensity and incentive spending: builders are competing on mortgage rate buydowns and price concessions, which is the mechanism compressing gross margins industry-wide from 27% to 25% at PHM; resale inventory normalizing adds a second source of competition.
• Credit and mortgage-financing conditions: PHM's supplementary mortgage/title operations and its new $625M Truist mortgage repurchase facility (August 2026) tie part of the P&L to warehouse funding and GSE liquidity conditions.

• Q3 2026 earnings on 2026-10-22 before the open: the critical read on autumn selling trends, cancellation rates, gross-margin guidance, and whether the +6% Q2 order growth was a one-quarter event or the start of a recovery.
• Mortgage-rate direction: a decisive drop in the 30-year rate would restore affordability and drive both order growth and gross-margin relief, the cleanest path to the ~$10.15 consensus 2027 EPS.
• Capital return execution: the $0.26 quarterly dividend (record 2026-09-15, payable 2026-10-02) plus continued repurchases - $373M in Q2 2026, roughly 6% of market cap annualized - mechanically lift per-share earnings even if net income is flat.
• Estimate revisions after Q3: with only $10.151 consensus next-FY EPS against $9.944 trailing, there is little slack in the model; a guidance cut or raise would likely move the stock more than the reported quarter itself.

• Continued gross-margin compression: home-sale gross margin already fell 200bp YoY to 25.0% and FMP-derived operating margin roughly halved YoY to 11.1% in Q2 2026 - if incentives and price concessions persist, the flat-EPS consensus for the next fiscal year is simply wrong.
• Higher-for-longer rates: sustained mortgage-rate pressure would keep volumes and ASPs under pressure (closings already -8% YoY, ASPs -3% YoY) and could force spec-inventory writedowns.
• Earnings quality and cash conversion: TTM free cash flow turned negative in Q2 2026 (-$12.9M), the accruals ratio of +2.49% as of 2026-06-30 shows net income running slightly ahead of operating cash flow, and net income exceeded operating income - the profit line is being helped by items below the operating line.
• Valuation without a cushion: at 12.03x TTM versus a 7.67x 5-year average, the stock offers no margin of safety if EPS undershoots; a re-rating to the 9-10x builders historically receive in a confirmed downcycle implies $90-100.
• Land and competitive risk: paying up for land into a slowing market, plus aggressive incentive spending by peers, could extend the margin reset well beyond the current two quarters.

PulteGroup is a genuinely well-run builder - ~14.3% ROIC, ~15-17% ROE, 0.177x debt-to-equity, a 5.27 current ratio, and roughly $0.5B of net debt - but it is not a buy right now, and it is not a screaming short either. Revenue is down 11% YoY, operating income is down 45%, gross margin has compressed 200bp+, free cash flow just went negative, and the stock trades at a 57% premium to its own 5-year average P/E with a broken chart (below all three moving averages, RSI 36.8, all momentum windows negative). The bull case is real but still unproven: orders +6% YoY, a 60bp sequential margin recovery, an EPS beat, and a consensus that already models flat-to-up 2027 earnings. That is a 'wait for the Q3 2026 print on October 22 to confirm the turn' situation, not a buy-the-dip one - the risk/reward is roughly balanced with the downside from a guidance miss ($90-100) larger than the near-term upside from stabilization (~$130).

Method: justified forward P/E, cross-checked against book value and EV/EBITDA. I apply 12.0x to a FY2027 EPS estimate of $10.15 (essentially consensus, after haircutting for continued gross-margin pressure but crediting ~4-6% annual per-share accretion from buybacks) = $121.80, rounded to $122. I use 12x rather than the 7.67x five-year average because that average is distorted by peak-cycle 2024 EPS of $14.71 and is not a valid anchor for a company now earning $9.944 TTM. Cross-checks bracket the figure: ~1.8x projected 2027 book value per share (~$68) equals ~$122, and ~9.5x a normalized ~$2.3B EBITDA less ~$0.5B net debt gives ~$113, so the three approaches cluster in the $113-122 band rather than resting on one multiple.