PulteGroup's 10-K (filed 2026-02-04) plus Q4 earnings call reveal a 2-3 year earnings quality deterioration that market hasn't fully priced. Legacy land bought at 2023-2024 peak pricing hits P&L now through 2027, while BTO pivot benefit won't materialize until 2027-2028.

What's in the 10-K

Insurance Reserve Benefit Cliff: $290M Structural Headwind

FY2024: $333.9M reserve reversal (construction defect claims came in lower than actuarial estimates) FY2025: $42.3M reversal

That's a $291.6M YoY cliff (10-K page 26, footnote b). This was a non-recurring SG&A benefit that masked true cost growth. The 10-K notes: "SG&A increased $252.7 million, or 19%, in 2025 compared with 2024. This increase resulted primarily from insurance reserve reversals of $42.3 million in 2025 compared to $333.9 million in 2024."

Normalized SG&A increased ≈$250M YoY despite flat revenue. Management said "managing and balancing our overhead costs consistent with the demand environment" (10-K MD&A), but headcount and technology spending continued through volume decline.

The reserve reversals were actuarial true-ups on IBNR (incurred but not reported) construction defect claims. Lower frequency/severity than modeled = reserve release. But the easy comps are done — you can only reverse reserves once.

Land Charges Accelerating: 2.7× YoY to $126M

FY2024: $52.9M ($34.6M impairments + $18.3M deposit write-offs) FY2025: $125.8M ($77.4M impairments + $48.4M deposit write-offs)

That's $73M incremental pain, 138% increase (10-K MD&A). The impairments are on owned land bought 12-18 months ago at peak pricing. The deposit write-offs are PHM walking from option contracts on land they no longer want at contracted prices.

From 10-K: "We have...taken additional pricing actions in many of our communities, which resulted in $77.4 million of land inventory impairments in 2025. We continue to update the underwriting for our land option contracts prior to buying additional land and have made decisions to walk away from a number of land option agreements, which resulted in write-offs of deposits and pre-acquisition costs totaling $48.4 million in 2025."

The 10-K doesn't disclose unobservable inputs used in fair value measurement for these impairments (no ASC 820 Level 3 disclosure), meaning management deemed them "not significant" under SEC rules. Translation: This is just the beginning. Land bought in 2023-2024 (peak pricing) closes communities in 2025-2027. The 10-K specifically notes 24-36 month absorption except Del Webb (longer).

Geographic Breakdown Shows Texas Concentration Risk (10-K segment data, pages 27-28)

  • Texas revenue: $1.68B (10% of total), down -22% YoY
  • Texas pretax income: $162M (6% of total), down -53% YoY
  • Texas closings: 4,352 units, down -20% YoY
  • Florida also weak: Revenue -9%, pretax income -27%

Texas and Florida combined = $5.9B revenue (35% of total). The Q4 worker noted "regional divergence" but didn't quantify that half the land charge pain likely concentrates in these two geographies.

Credit Facility Expansion (10-K liquidity section)

Effective 2026-02-04, PHM amended Revolving Credit Facility:

  • Upsized from ≈$857M to $1.75B committed capacity
  • Extended maturity to 2031 (from 2026)
  • Expanded accordion feature to $750M (potential $2.5B total)
  • $892.9M available at 12/31/2025

This is smart defensive positioning, not bullish signal. Banks are willing because PHM has fortress balance sheet ($2.0B cash, negative 3% net debt). But timing matters: PHM is expanding liquidity NOW (into softening market) to be opportunistic later (if competitors stumble).

Spec Inventory Strategy (10-K confirms)

10-K MD&A: "Gross margins in 2025 were also unfavorably impacted by our efforts to reduce completed spec inventory to more appropriate levels, which we expect will continue to be an area of focus in 2026. While we have made significant progress in reducing the level of spec inventory during 2025, the level of completed spec inventory remains elevated for the current demand environment."

Confirms spec reduction is happening (spec inventory -18% YoY per Q4 call), but Q4 margins compressed because they're clearing old spec with heavy incentives.

What's NOT in the 10-K (From Q4 Earnings Call)

BTO Margin Premium: "Hundreds of Basis Points"

Q4 call: "BTO carries hundreds of basis points higher gross margins" — but 10-K doesn't quantify this. The 10-K strategy section says: "Maintain an appropriate balance of built-to-order and speculative homes" but provides no breakdown of current mix or margin differential.

Current BTO/Spec Mix: ≈60% Spec / 40% BTO, Targeting Reversal

Q4 call: "Currently ≈60% spec / 40% BTO, targeting reversal to 60%+ BTO / 40% spec" — Not in 10-K. The 10-K only describes BTO conceptually: "Many of our homes are sold on a built-to-order basis where we do not begin construction of the home until we have a signed contract with a customer."

2026 Gross Margin Guidance: 24.5-25%

Q4 call forward guidance (10-Ks don't contain guidance). This implies another 100-130 bps compression from FY2025's 26.3%.

Del Webb Revenue Contribution: 22-24% of Closings

Q4 call said Del Webb is "22-24% of closings, targeting 25%" and "routinely delivers highest gross margins." But 10-K only discloses customer mix: "During 2025, 38%, 40%, and 22% of our home closings were to first-time, move-up, and active adult customers, respectively."

The 22% active adult ≈ Del Webb, but 10-K doesn't break out Del Webb revenue separately or quantify margin premium. At 22% of 29,572 closings = 6,500 units. At $573K average ASP (company average), that's ≈$3.7B revenue (22% of total). But Del Webb ASP is likely higher (active adult buyers, more amenitized, longer absorption). Without quantification, can't assess if Del Webb buffer offsets entry-level compression.

Peer Comparison (10-Ks)

Lennar (LEN): $23.1M deposit write-offs in FY2025 vs $5.1M in FY2024 (4.5× increase). No material land impairments disclosed. LEN is more disciplined on land option contracts.

D.R. Horton (DHI): $6.7M impairments in Q3 2025 only. Full-year likely <$30M. DHI operates at massive scale with better land option leverage.

M/I Homes (MHO): $51M in Q4 2025 charges ($40M inventory impairments + $11M warranty). $30M of impairments on ≈1,000 lots in entry-level (ASP <$375K), concentrated in Austin/San Antonio Texas. MHO's $40M in one quarter = 32% of PHM's full-year $126M, but MHO is 1/10th PHM's size.

PHM's $126M land charges are 2.6× sector average relative to size. The Texas concentration (Austin/San Antonio weakness echoed in both PHM and MHO) suggests geographic, not company-specific problem.

What This Means

Thesis Timeframe, Not Thesis Validity

The BTO pivot thesis is correct: BTO generates "hundreds of bps" higher margins than spec (Q4 call). The balance sheet fortress is real: $2.0B cash, $892.9M available on $1.75B credit facility, negative 3% net debt.

But there's a 2-3 year valley between here and payoff:

  • 2025-2026: Legacy land (bought 2023-2024 peak) flowing through P&L at compressed margins. Land charges will remain elevated ($100M+ range) as communities underwritten at higher prices/lower incentives are reunderwritten in current reality.

  • 2027-2028: BTO pivot matures. New communities on cheaper land (acquired 2025-2026 distress pricing) with structural BTO premium pricing start closing. Margin inflection.

The $400M Earnings Quality Deterioration:

  • Insurance reserve benefit cliff: $290M
  • Land charge acceleration: $73M
  • Combined: $363M pretax headwind

FY2025 pretax income: $2.75B Normalized for these items: $2.39B (13% lower than reported)

And this gets WORSE in 2026 before it gets better. Q4 call guided to gross margin 24.5-25% (vs 26.3% actual in FY2025). That's another 100-130 bps compression on $16-17B revenue base = $170-220M incremental gross profit headwind.

Stock Pricing:

  • Current: $134.57
  • YTD: +21.8%
  • Consensus target: $140.92 (only +4.7% upside)
  • Range: $115-159

Street is NEUTRAL with minimal upside despite PHM being "number one TSR among homebuilders for both 1 year and 10 years" (per Q4 call). Why?

Analysts see the 2-3 year margin trough. Recent actions:

  • RBC Capital (2026-01-30): Sector Perform, $115 target (15% downside)
  • Citigroup (2026-02-03): Neutral, $136 target (minimal upside)

Only bulls are Oppenheimer ($149), Wells Fargo ($150), UBS ($159). But consensus is 8 Hold, 8 Buy/Strong Buy, 0 Sell. That's "wait and see," not "accumulate."

Investment Implication

Position-dependent decision:

If original thesis modeled 2-3 year valley + fortress balance sheet survival = HOLD through trough, conviction ≥75%.

If original thesis assumed 2026 margin recovery = Thesis timeframe extended, reassess size. Near-term pain deeper than expected. The $400M earnings quality deterioration + 2026 guide to further margin compression (per Q4 call) + elevated land charges = reduce position or wait for Q1 2026 margins to confirm BTO progress vs legacy land compression is dominating.

LR 0.75 rationale:

This isn't actionable short (LR 0.6 or lower). It's a timing/holding period adjustment. The 75% reflects:

  • 25% bull case: Market already prices valley, BTO pivot accelerates faster than expected, land charges peak in 2025
  • 75% bear case: 2-3 year trough is deeper/longer than market expects, land charges persist through 2026-2027, BTO margin benefit materializes but doesn't fully offset legacy land pain until 2028

The thesis is intact (BTO pivot + fortress balance sheet). But the path is rockier and longer than Q4 call implied. Size for surviving 2-3 years of margin compression, not for immediate recovery.

Sources: PHM 10-K (2026-02-04), PHM Q4 2025 earnings call, LEN 10-K (2026-01-28), DHI 10-K (2025-11-19), M/I Homes Q4 2025 earnings