LGI Homes is a $1.1B Sun Belt entry-level builder. The Q1 2026 10-Q filed April 28 disclosed two non-obvious mechanics that sharpen the cohort-level Japanese acquisition pattern in motion since February. Trading at 0.541x book.

What the filing says

Home sales revenue $319.7M (-9.0% YoY). 881 closings (-11.5%), ASP $362,924 (+2.9%). GAAP gross margin 18.7% vs 21.0%, but $4.7M of the decline is inventory impairment in Florida and Central — ex-impairment 20.2%, adjusted 23.4% (nearly flat). Adjusted EBITDA margin improved +230 bps to 7.6% on cost discipline (selling expenses -22.9%, G&A -10.7%). Book value $90.49/share. Wholesale revenue -45.4% on institutional demand withdrawal. LGI Living SFR arm operational with 35 Q1 closings against a $50M Evergreen Residential facility (expandable to $125M).

Two findings the headline misses

Cancellation rate 45.6% vs 16.3% prior year. Peer Q1 2026 band: DHI 16%, LEN 13%, KBH 12% ("lowest in four years," per CEO Q1 call), MTH 11%, PHM 12%, TMHC 10%, MHO 8.4%. LGIH is a 3-5x outlier — not a sector demand collapse, an idiosyncratic contract-to-close failure.

Credit covenant restricts share repurchases AND dividends through December 31, 2026. Q1 buybacks: zero. $157.3M authorization stranded. Across the below-book cohort (BZH, CCS, LGIH, MTH, KBH, MHO), 5 of 5 peers actively repurchased Q1 2026 (BZH $15.1M, CCS ≈$36.5M, MTH $131M acceleration, KBH $50M, MHO ≈$50M). LGIH alone is mechanically locked.

What the market thinks

Reverse-engineering the current price against probability-weighted operating outcomes (E[NoAcq] ≈ $45-51, E[Acq] ≈ $110) implies P(Japanese acquisition by Dec 31, 2026) of 5-8%. Our prior is 18% on the LGIH-specific subset of cohort prediction (basket prediction at 55% across 5 candidates). EV over 8 months: 18% × +148% (1.34x book takeout per Sumitomo/TPH precedent = $121.26) + 50% × -15% (cancellation drag) + 22% × -3% (sideways) + 10% × +25% (sector rally) ≈ +21% expected. Market-implied EV ≈ +7%. Probability gap ≈10-13 percentage points; return gap ≈14 percentage points.

Why the gap exists

Cross-ticker work was unsynthesized at filing date: peer cancellation rates required 7+ separate Q1 filings, and the buyback covenant differential required reading 5 credit agreements. The covenant expiry December 31, 2026 is a deadline mechanic — once it lifts, LGIH can defend the discount via buyback. Three Japanese acquirers with unmet multi-year deployment targets (Sumitomo 23K/yr by 2030, Daiwa 10K+ by 2026, Sekisui expanding) face an 8-month window before that mechanical edge evaporates. Four Japanese→US homebuilder deals announced in 5 weeks Feb-Mar 2026 (Sumitomo/TPH $3.97B, Daiwa/UHG, Iida/Wright, Daiwa/JK Monarch) establish the pace.

Risks, ranked

1. Operational bear deepens. 1,699-unit backlog at material conversion risk if cancellation rate persists. Q2-Q3 closings could materially miss.
2. Cohort thesis cools. Pace was 4 deals in 5 weeks Feb-Mar; April produced no Japanese deal. May-June pace will signal whether the wave continues or pauses.
3. Sector exposure dilutes the idio thesis. Regression: 70% XHB-driven variance (β=1.94), only 30% idio. Naked long is effectively a leveraged builder bet, not the structural setup.
4. Squeeze risk both directions. 20.2% short interest, 5.0 days to cover.

Catalysts

• 2026-06-30 — Net debt/cap retention metric date (CEO bonus gates on covenant compliance)
• ≈2026-08-15 — Q2 2026 10-Q (resolves cancellation trajectory and Q2 revenue YoY)
• 2026-12-31 — Cohort acquisition window deadline + covenant expiry
• Anytime — Japanese acquirer 8-K or 13D

What would change our mind

• Q2 cancellation rate <25% — idio bear reverses, M&A urgency reduces, reframes as single-driver upside on operating recovery
• Second below-book cohort builder acquired (BZH, CCS, MTH, KBH, MHO) — confirms the cohort pattern is closing; LGIH within-cohort uplift becomes the operative variable
• Strategic review announced or activist 13D — internal trigger eliminates the "no behavioral signal" bear case
• Q2 cancellation rate >50% — confirms franchise damage; takeout multiple discounts vs the TPH precedent
• Japanese acquirer Form 4 / 13D in LGIH — re-rates the binary

Evidence