HHH carries 37,000 acres in West Phoenix (Teravalis) on the balance sheet at roughly $16K/acre cost basis. In 2026, Tract — a private developer — closed on 2,069 acres four miles away at ≈$66K/acre, targeting a $20B / 1.8GW data-center campus; Arizona Land Consulting paid $135K/acre on an adjacent 160-acre parcel; Microsoft Goodyear DC is already operational with a $90M expansion announced. HHH trades at $64.28, a 38% blended discount to the $104 intrinsic value management published for the first time on May 8, 2026. The gap on Teravalis alone — $50K/acre delta × 37,000 acres = $1.85B — is roughly 49% of the $3.8B market cap, against a published 2026 transaction comp four miles down the road.

What the filing says

Q1 2026 MPC EBT was $84M, +33% YoY. Bridgeland (Houston) sold 62 acres at $688K/acre — volume +68% YoY, price +14% YoY, with in-community new home sales +12%. Summerlin (Las Vegas) custom lots cleared $7.2M/acre. The CFO said Q1 was "ahead of expectations" and that guidance would have been raised absent the Vantage acquisition distraction.

The Vantage Holdings insurance acquisition has its Delaware regulatory hearing May 19; Ackman: "I think we'll beat our quarter end estimate absent something unexpected happening, but I don't expect the unexpected here." Marc Grandisson — former CEO of Arch Capital (≈20 years, ≈20% CAGR) — joined the board on May 7, the day before the call. Annual guidance was withdrawn in full, replaced by a three-KPI framework: MPC residual land value (current per-acre prices, undiscounted), adjusted maintenance free cash flow, and remaining condo gross profit. Management's five-year IV target is $211/share, implying a 16% IV CAGR off the $104 floor.

What the market thinks

The tape did not reward the print: -2.3% on May 8, 0.7× average volume, near the $61 52-week low. Sell-side coverage is thin (three analysts on the Q1 call). Options are thinly traded. Forward P/E 63 is the optical metric the new KPI framework was designed to retire — Ackman and CFO O'Reilly explicitly told the call that quarterly EPS "is not the right metric" for a land company.

Trailing 250-day regression: orthogonal α = -32.8% annualized vs SPY/QQQ/XLRE, idio variance 73.3%, β_SPY = 1.40. The stock has been bleeding a third of its factor-explained return per year. Mean-reversion of that residual to zero, on its own, is +33%/yr before any of the idio thesis plays through.

Why the gap exists

The 38% discount works backward almost exactly to the sum of three standard Ackman-holdco haircuts: execution risk on Vantage (≈10%), the perpetual NAV discount applied to Pershing-controlled vehicles like PSH (≈20-25%), and a Sunbelt residential land headwind (≈10%). The cumulative ≈38% × $104 ≈ $64.50 — matching the tape. The market is internally consistent.

What's mispriced is the compression optionality on each lens. Grandisson's arrival answers the single biggest objection to the mini-Berkshire framework. Q1 Bridgeland data says HHH is not commodity Sunbelt land — though that is a one-quarter print, and FPH (zero residential closings in Q1) and JOE (residential JV equity -66% YoY) do not corroborate the premium-MPC strength. The Tract / West Phoenix arithmetic is in the public domain and unpriced. Each lens has a dated test in the next 90 days.

Risks, ranked by impact

Q2 Bridgeland reverts. Single-quarter MPC strength may have been pull-forward or seasonal. Houston-area homebuilders (MTH, MHO, TPH) have not yet been screened for direct corroboration. Q1 print is genuinely positive but unconfirmed.

Vantage book mix property-cat-heavy. Property cat reinsurance is two years past peak: RNR Jan renewals down low-teens, EG April -14.6%, ACGL "double-digit rate decrease." If Vantage's book skews property-cat, Grandisson's "right after the 2019 turn" framing is stale and ROE upgrade thesis weakens 10-15%.

β = 1.40 to SPY. A 20% market drawdown delivers -28% on factor exposure alone. Naked HHH long is mostly leveraged-market plus REIT beta with an idio kicker; clean expression requires SPY and XLRE hedges.

Ackman holdco discount is structural. PSH has traded at 30%+ NAV discount for years. $104 IV may anchor a floor without compressing the discount, particularly if the Pershing Square USA listing absorbs the marginal Ackman-aware buyer.

Catalysts

Three of four idio disambiguators resolve in a 60-day August window.

Forward EV

State-weighted 12-month expected return: ~+16%. Factor-weighted cross-check: +18-21%. Market-implied (from price action at 11% of 52-week range with dead options): ≈0-5%. Edge ≈12-17 points over the disambiguator window. Idio Sharpe ≈0.59 — decent, not extraordinary.

What would change our mind

Toward more bullish: Ackman / Pershing Form 4 Code P (open-market purchase) post-IV disclosure — currently absent. A binding hyperscaler MOU on Teravalis (8-K Item 1.01). Q2 Bridgeland confirming volume and price.

Toward less bullish: Q2 Bridgeland reversion. Vantage book mix disclosed property-cat-heavy. A direct Houston-area homebuilder (MTH, MHO, TPH) print echoing the DFH/PHM Sunbelt softness signal that has stood unrebutted from Q4 2025.

The trade is the specific 38% discount, the $1.85B West Phoenix delta against a public 2026 comp, and a catalyst stack that forces the market to reprice — in either direction — within 90 days. Not conviction. Just an addressable mispricing with dated resolution.

Evidence