AMREP Corporation (AXR): A Net/Net That I'm Passing On
I spent several months researching AMREP Corporation as a deep value investment. The idea is compelling on the surface — a fortress balance sheet, an activist insider, and 16,000 acres of land next to Intel’s $4 billion semiconductor campus. After scraping Sandoval County’s entire assessment database, verifying water rights claims against primary sources, and stress-testing the valuation through multiple frameworks, I’m passing. The reason is simple: water. Not because water rights would add upside, but because the lack of available water may permanently cap the value of the land that the entire idea depends on. Below is the full research.
The One-Paragraph Version
AMREP Corporation (AXR, ~$26/share, ~$133M market cap) is a zero-debt land developer in Sandoval County, New Mexico. It owns approximately 16,600 acres adjacent to Intel’s Rio Rancho semiconductor campus, plus 55,000 acres of mineral rights. Cash and real estate inventory total $113M — 85% of the market cap — meaning the market assigns roughly $20M to the land, minerals, and the entire operating business. An activist investor (James Dahl, 21%+ ownership) is running a documented playbook to force a sale at NAV. Intel is spending $3.5-4B next door. On paper, this is a classic net/net with a catalyst. In practice, Rio Rancho’s water supply is legally constrained, the county’s own plan says full buildout is impossible, and the Pueblo of Sandia ruling capped the city’s water permits. Without water, 16,000 acres of desert land stays desert land — and the idea breaks.
Company Overview
AMREP was founded in 1961 as the American Realty and Petroleum Corporation. Between 1960 and 1975, the company purchased approximately 91,000 acres in Sandoval County, New Mexico and subdivided them into tens of thousands of half-acre and one-acre lots, sold via mass-market advertising campaigns to buyers worldwide. By 1977, AMREP had sold 75,000 of these lots. The remaining land — plus parcels reacquired over the decades — constitutes today’s ~16,600-acre land bank.
The company operates two segments:
Land Development (~52% gross margins): The core value engine. AMREP takes raw land through entitlement, installs infrastructure (water, sewer, roads, utilities), and sells developed lots to national and regional homebuilders. This segment generates the company’s best margins but operates at a glacial pace — roughly 20 acres per year of residential lot development.
Homebuilding (via Amreston, ~10-12% margins): A newer subsidiary building single-family homes directly. As of early 2026, Amreston had 21 homes under lease and was actively constructing spec homes. This is Dahl’s “pivot to earnings” strategy — transforming AMREP from a passive land holder into an active homebuilder to support a higher earnings multiple.
Subsidiaries
AMREP operates through a web of wholly-owned subsidiaries, all verified via SEC Exhibit 21:
| Entity | Role |
|---|---|
| AMREP Southwest Inc. | Primary operating subsidiary |
| Outer Rim Investments Inc. | Land holding (largest by parcel count) |
| Resurrection Land Company LLC | Land holding |
| Amreston Homes LLC | Homebuilding |
| Arroyos LLC | Land holding |
| Lomas Encantadas Development Co LLC | Subdivision development |
| Mountain Hawk Development Co LLC | Subdivision development |
| Hawksite 27 Development Co LLC | Subdivision development |
| Las Fuentes Village LLC | Subdivision development |
| Clean Slate Properties LLC | Land holding |
| Corner Office LLC | Land holding |
A notable discovery from the county assessment data: AMREP Energy LLC — not listed in the 10-K’s subsidiary exhibit — holds 8 parcels totaling 550 acres in “Lands of Outer Rim” subdivisions, assessed at $459/acre. This is likely a mineral rights or energy leasing entity.
Land Projects
| Project | Developed Inventory | Under Development | Undeveloped | Use |
|---|---|---|---|---|
| Lomas Encantadas | 120 lots | 126 acres | — | Residential |
| Commerce Center | — | 29 acres | — | Commercial/Industrial |
| Paseo Gateway | — | — | 327 acres | Long-term hold |
| Hawk Site | 2 lots | 35+93 acres | — | Residential |
| High Contiguous Areas | — | — | ~16,000 acres | Strategic hold |
Source: AMREP 10-K FY2025, Item 2
The Fortress Balance Sheet
This is the single most compelling element of the idea. As of October 31, 2025:
| Item | Amount |
|---|---|
| Cash & Equivalents | $44.6M |
| Real Estate Inventory (at cost) | $68.6M |
| Total Liquid Assets | $113.2M |
| Total Debt | $23,000 |
| Total Liabilities | $3.8M |
| Shareholders’ Equity | $130.0M |
| Book Value Per Share | ~$25.81 |
Key Financial Metrics:
| Metric | Value |
|---|---|
| Debt-to-Equity | 0.02% |
| Current Ratio | 27.98 |
| FY2025 Operating Cash Flow | $10.2M |
| FY2025 Total Revenue | $49.7M |
| Land Sale Gross Margin | ~52% |
| Cash as % of Market Cap | ~33% |
The balance sheet is genuinely extraordinary. Zero debt for a land developer is almost unheard of. AMREP can’t go bankrupt, can’t be forced to sell assets at distressed prices, and can wait out any downturn. The $113M in liquid assets against a $133M market cap means the market is implicitly valuing 16,000 acres of land, 55,000 acres of mineral rights, and the entire operating business at approximately $20M.
Sources: AMREP 10-K FY2025, 10-Q Q2 FY2026
The Land: What the County Assessor Says
To verify the land holdings, I scraped Sandoval County’s EagleWeb assessment database for every parcel owned by AMREP-related entities. The result: 19,313 parcels across 12 entities.
However, the raw data requires a critical adjustment. One parcel — “Lomas Encantadas Unit 1E Phase 1, Block 2, Lot 10” — shows as 11,653 acres assessed at $87,000. This is a master tract / remainder parcel representing the entire unsubdivided boundary of the platted unit, not a separately owned piece of land. It overlaps with thousands of individual lots already counted in the data. At $7.47/acre assessed value, the county is effectively saying this land has near-zero standalone value.
Excluding the double-counted master tract:
| Entity | Parcels | Acres | County FMV | $/Acre |
|---|---|---|---|---|
| AMREP Southwest Inc | 6,197 | 4,312 | $53.7M | $12,449 |
| Outer Rim Investments Inc | 12,609 | 10,543 | $40.0M | $3,793 |
| Lomas Encantadas Development | 88 | 92 | $6.8M | $74,432 |
| Amreston Homes LLC | 62 | 14 | $6.7M | $478,255 |
| Corner Office LLC | 1 | 1 | $1.3M | $1,713,289 |
| AMREP Energy LLC | 8 | 550 | $0.3M | $459 |
| Arroyos LLC | 52 | 76 | $0.1M | $1,361 |
| Clean Slate Properties LLC | 1 | 0.2 | $0.03M | $111,500 |
| TOTAL | 19,018 | 15,587 | $108.9M | $6,988 |
The 15,587 acres closely aligns with the 10-K’s ~16,600 figure (the gap likely reflects parcels sold between the assessment date and reporting date, or parcels classified differently).
What the Two-Tier Structure Reveals
The data breaks cleanly into two tiers:
AMREP Southwest (4,312 acres at $12,449/acre): These are the parcels closer to the city limits — entitled, platted, infrastructure-adjacent. The county’s $12,449/acre assessment reflects development-stage land. These parcels cluster in Rio Rancho Estates units near Lomas Encantadas, Las Fuentes, and the Intel campus. This is the land that’s actually developable in the near term.
Outer Rim Investments (10,543 acres at $3,793/acre): This is the raw, unentitled land farther west. Over 12,500 individual parcels — mostly half-acre and one-acre lots from AMREP’s original 1960s subdivisions — with no infrastructure, no water service, and no near-term development path. The county’s $3,793/acre assessment is the honest floor on bulk undeveloped desert land.
The Valuation Gradient
The development pipeline reveals a 100x value creation arc:
| Stage | Entity | $/Acre |
|---|---|---|
| Raw unentitled land | Outer Rim | $3,793 |
| Entitled/platted lots | AMREP Southwest | $12,449 |
| Development-stage lots | Lomas Encantadas | $74,432 |
| Near-finished homes | Amreston | $478,255 |
| Commercial/office | Corner Office | $1,713,289 |
This gradient is real — it’s how land development creates value. The question is: how fast can AMREP push raw Outer Rim land through the pipeline? At 20 acres per year, the answer is: not fast enough to matter.
Total 2025 property tax across all entities: $1,348,523 — a ~1% annual carrying cost on $108.9M in assessed assets. Not onerous, but a real drag on a company holding 10,543 acres of raw land that generates zero income.
Source: Sandoval County EagleWeb Assessment Database, scraped March 2026
Sum-of-the-Parts Valuation
Using county assessment data as the primary anchor, not the theoretical valuations from my earlier analysis:
County-Anchored Valuation
| Component | Bear | Base | Bull |
|---|---|---|---|
| Cash | $44.6M | $44.6M | $44.6M |
| Outer Rim raw land (10,543 ac) | $31.6M ($3K/ac) | $47.4M ($4.5K/ac) | $63.3M ($6K/ac) |
| AMREP SW entitled land (4,312 ac) | $53.7M (county) | $53.7M (county) | $64.4M (+20%) |
| Amreston homes/lots | $6.7M | $6.7M | $6.7M |
| Lomas Encantadas dev lots | $6.8M | $6.8M | $6.8M |
| AMREP Energy (minerals) | $0 | $5.5M | $27.5M |
| Other entities | $1.4M | $1.4M | $1.4M |
| Less: overhead/liabilities | ($5.0M) | ($5.0M) | ($5.0M) |
| Total | $139.9M | $161.2M | $209.8M |
| Per Share (5.27M shares) | $26.55 | $30.59 | $39.81 |
| vs. $26 current | +2% | +18% | +53% |
Why This Is More Conservative Than My Original Model
My earlier analysis valued all ~16,000 acres at a blended $6K-$10K/acre, producing a Bear/Base/Bull of $27/$45/$68 per share. The county data forced a cleaner and more honest split:
- Outer Rim’s 10,543 acres are raw scattered lots, not contiguous developable land. The county values them at $3,793/acre. Applying $10K/acre to this land was wrong — it conflated entitled, infrastructure-adjacent parcels with raw desert.
- AMREP Southwest’s 4,312 acres are the genuinely valuable parcels, and the county’s $12,449/acre confirms this. These are the acres that actually support higher valuations.
- The bull case compresses from $68 to $40/share because the original model was applying strategic-premium pricing to land that doesn’t warrant it.
Comparable Sales Data
Actual closed transactions for land in Rio Rancho (from LandAndFarm.com):
| Metric | Value |
|---|---|
| Median $/acre (individual lots, 2023-2026) | $4,999 |
| Range | $3,225 - $6,000 |
| Sample | 9 closed sales |
AMREP’s own bulk transactions:
| Transaction | Acres | $/Acre | Date |
|---|---|---|---|
| FY2023 bulk sale | 1,196 | $5,942 | Calendar 2022 |
| Project Ranger/Castelion | 590.8 | $4,237 | Late 2024/early 2025 |
| Castelion site | ~549 | $4,562 | Late 2025 |
Critical caveat (Buffett’s Texas Pacific Land Trust problem): These are sales of 600-1,200 acres. You can’t extrapolate small-parcel pricing to 16,000 acres. Bulk sales at scale trade at significant discounts. AMREP has never demonstrated a $10K/acre transaction on any meaningful acreage.
Sources: LandAndFarm.com, AMREP 10-K FY2025, SEC filings
The Catalysts: Intel, Castelion, and Dahl
Intel’s Rio Rancho Expansion
Intel has operated in Rio Rancho since 1980 — a 45-year presence. The current expansion is its largest:
| Detail | Value |
|---|---|
| Investment | $3.5-4.0B |
| Facilities | Fab 9 and Fab 11x |
| Technology | Foveros 3D advanced packaging |
| Direct jobs | Hundreds permanent + 3,000+ construction |
| Economic multiplier | ~3,500 additional jobs statewide |
| GRT rebates | ~$14M |
This is the primary demand catalyst. Intel’s workforce needs housing, and AMREP owns the land adjacent to the campus. The bull case assumes Intel-driven demand pushes bulk land prices from $4-5K/acre toward $10K+.
Risk: Intel is under structural competitive pressure from TSMC and Samsung. Its CHIPS Act funding status is uncertain. If Intel scales back Rio Rancho operations, the demand catalyst evaporates. The land doesn’t disappear, but without a demand driver, 16,000 acres of desert land stays cheap.
Castelion (Project Ranger)
The defense technology firm Castelion broke ground in January 2026 on a $220M manufacturing campus in Sandoval County, on land purchased from AMREP. The project is expected to create 300 high-wage jobs (average $100K salary) and generate $650M in economic output.
This is a genuine second demand source — and AMREP already monetized part of it through the land sale at $4,237-$4,562/acre.
James H. Dahl — The Activist
Dahl is the catalyst this idea depends on. As of early 2026:
- Controls 21%+ of AMREP through a 13D group with associate Rainey Lancaster
- Purchased 13,000+ shares in February 2026 alone at $24.52-$26.23
- Board expanded from 4 to 5 members (January 2026)
- 30,000-share director ownership requirement eliminated
- Timothy McNaney (former co-president of a New Mexico homebuilder) appointed to the board
Sources: SEC EDGAR Form 4 and Schedule 13D filings, AMREP 8-K January 2026
Dahl’s Activist Playbook
Dahl’s approach at AMREP follows a pattern documented across his prior investments. Studying his interventions at Pope Resources (POPE) and St. Joe Paper Co. (JOE) reveals a four-phase playbook:
Phase 1: Aggressive Accumulation
The model: At Pope Resources, Dahl spent years buying shares despite an archaic MLP structure. At St. Joe, he bought until it became his largest holding.
At AMREP: Dahl and Lancaster have collectively accumulated 21%+ ownership, with aggressive open-market purchases continuing through early 2026. The pace of buying — 13,000 shares in a single month — signals he’s approaching critical mass.
Phase 2: Board Modernization
The model: At St. Joe, pressure from Dahl and outside investors forced retirement of the old guard and installation of “new blood” — including Peter Rummell, a former Disney executive.
At AMREP: The board expansion, elimination of the 30,000-share director requirement, and appointment of McNaney (a homebuilding executive, not a legacy board member) follow the same script. The board is being reconstituted for action.
Phase 3: Pivot to Earnings
The model: St. Joe shifted from passive timber harvesting to active development of master-planned communities.
At AMREP: The Amreston homebuilding subsidiary and the 21-home leasing strategy represent the earnings pivot. Dahl wants recurring cash flow to support a higher valuation — or to make the company more attractive to an acquirer.
Phase 4: NAV Gap Closure
The model: When Pope’s board refused to convert to a REIT, Dahl penned open letters in 2018-2019 demanding a merger. This forced a sale to Rayonier in 2020 at a 30% premium to the pre-announcement price.
Prediction for AMREP: If the board doesn’t announce a major transaction or significant development acceleration following the Intel and Castelion groundbreakings, Dahl is expected to move from passive buyer to active agitator — likely through an amended 13D calling for a sale to a national homebuilder (Lennar, D.R. Horton, Meritage) that needs AMREP’s land to serve Intel-driven housing demand.
The critical question Buffett would ask: Would you own AMREP if Dahl sold his entire position tomorrow? If the answer is no — if you’re buying because of Dahl, not because of the business — you’re making a bet on a person, not an investment.
The Water Problem — And Why I’m Passing
This is where it breaks.
What the County’s Own Plan Says
The Rio Rancho Estates Area Plan (RREAP), adopted by Sandoval County in September 2014, is the most authoritative government document on development constraints in AMREP’s territory. It states:
“Preliminary findings indicate that approximately 16,000 acre-feet of water annually would be needed to support a full residential build-out scenario on the 40,000+ lots in the Estates, and that only 8,000 acre-feet of water annually is available underground on a renewable basis.”
“Given the water availability in the Estates, a full residential build out of the Estates is not possible. Under this scenario, water availability to both County and City residents would be jeopardized.”
The county itself says full buildout is impossible. The math: 16,000 AF needed, 8,000 AF available. That’s a 50% shortfall — structural and permanent, not cyclical.
What the Pueblo Ruling Did
In August 2025, Judge Eichwald of the Office of the State Engineer denied Rio Rancho’s request to expand its water permits from ~12,020 AF to 24,000 AF. The Pueblo of Sandia — whose water rights predate and legally override all state permits under the Prior and Paramount doctrine — successfully blocked the expansion.
| Element | Detail |
|---|---|
| Ruling | Denied Rio Rancho’s permit expansion |
| Current cap | ~12,020 AF |
| Requested | 24,000 AF |
| Legal doctrine | Prior and Paramount (Pueblo rights predate all state permits) |
| Appeal filed | Yes — by Rio Rancho and the State Engineer (25 alleged errors) |
| Expected resolution | 2027-2030 |
If the ruling stands: the city’s water supply for new development is permanently capped. Every lot AMREP sells requires a water availability letter from Rio Rancho Utilities. If the city can’t provide water, AMREP can’t develop.
The Water Rights I Couldn’t Verify
Earlier research cited a September 2020 Sub-File Order confirming 2,797.11 AFY of water rights held by AMREP. I spent considerable time trying to verify this from primary sources:
- The New Mexico Water Rights Reporting System (NMWRRS) shows only 6 acre-feet of verifiable water rights for AMREP entities
- The 2,797 AF figure does not appear in any AMREP 10-K filing
- The OSE link in the original research pointed to the Santa Cruz/Truchas adjudication — which covers northern New Mexico, not Sandoval County. AMREP’s land is in the Middle Rio Grande basin, a completely different case
- New Mexico’s groundwater basins have been fully appropriated per the State Engineer — no new rights are available
I couldn’t verify the 2,797 AF claim from any primary source. The 10-K mentions “ensuring the availability of water service” as a development activity, but never quantifies AMREP’s own water rights. Without verifiable water rights, I can’t include them in the valuation. And without water rights, the development economics of 16,000 acres of desert land become fundamentally questionable.
What This Means for the Land
The two-tier assessment structure now makes sense through the water lens:
AMREP Southwest (4,312 acres, $12,449/acre): These parcels are near the city limits, within reach of Rio Rancho’s existing water infrastructure. They can be developed — slowly — under the city’s current water allocation. This land is genuinely valuable.
Outer Rim (10,543 acres, $3,793/acre): This is the land that needs water it may never get. Outside city limits, no municipal water service, and private wells cost $30,000 each and are limited to 1 AF/year. The RREAP documents only 63 individual private wells and 30 shared wells in the entire 43,629-acre plan area. At $3,793/acre, the county is telling you this land has a low probability of near-term development.
The county is right. Without water, 10,543 acres of Outer Rim land is a permanently impaired asset. Not worthless — someone will buy desert land for $3-4K/acre — but structurally limited in its upside. The development-stage AMREP Southwest land is real value. The Outer Rim land is a hope.
RREAP: The Moat and Its Limits
The RREAP does confirm several bullish points:
- AMREP owns approximately one-third of all parcels in the plan area — the county explicitly names them as the dominant landowner
- 30,000 parcels are held by absentee owners worldwide, making competitive land assembly effectively impossible
- The “antiquated platting” from pre-1973 subdivisions means no infrastructure exists — creating a barrier to entry that only a scaled operator like AMREP can overcome
- The Paseo del Volcan highway under construction through the plan area is free infrastructure AMREP didn’t pay for
- The county plan itself calls for “Private Sector Limited Redevelopment” and “Large-Scale Government Redevelopment Areas” — both structurally favorable for the largest landowner
But the same RREAP that confirms the moat also confirms the constraint: “a full residential build out of the Estates is not possible.” The moat protects AMREP’s competitive position, but the water ceiling caps the value of that position.
What Buffett Would Say
I applied Buffett’s investment framework to this, and his verdict is unflattering:
“This is a cigar butt, and I don’t smoke cigar butts anymore.”
This is a textbook cigar butt: a stock trading below the replacement cost of its assets, with a hope that the gap closes. The “one puff” is the NAV gap. But if the gap doesn’t close for 10 years, your capital sits idle earning a 7.7% earnings yield in an illiquid stock.
The Texas Pacific Land Trust Problem
Buffett addressed land holding companies directly in 1998:
“People sometimes get very confused about… they’ll look at some huge Land Company… take Texas Pacific Land Trust… and they’ll sell 1% of their land every year and they’ll take that as applying to everything and come up with some huge value compared to the market value but that’s nonsense.”
This is exactly what the bull case does. It takes AMREP’s $5-6K/acre transactions on 600-1,200 acres and extrapolates to 16,000 acres. But you can’t sell 16,000 acres at the price you sell 600 acres. The county data confirms this — Outer Rim’s 10,543 acres are assessed at $3,793/acre, well below the small-transaction comps.
The One Positive: Fortress Balance Sheet
Buffett would respect the balance sheet unreservedly. Zero debt. $44.6M in cash. $23,000 in total liabilities. AMREP can’t go bankrupt and can’t be forced to sell. But: “A fortress balance sheet keeps you alive. It doesn’t make you rich.”
Productive vs. Unproductive Assets
Most of AMREP’s value rests on Buffett’s “Category 2” assets — things that produce nothing today but that you hope will be worth more tomorrow. The 15,500 acres of undeveloped land and 55,000 acres of mineral rights are idle. The land requires decades of entitlement work and hundreds of millions in infrastructure investment before it produces anything. Buffett would rather own productive assets that compound.
His Position Sizing
Buffett wouldn’t buy AMREP in 2026. If running a small fund, having done all the scuttlebutt work, he might put 2-3% in as a special situation with a strict time limit: “If nothing happens in 3 years, I’m out.” He would not put 15-20% in an asset play dependent on an activist.
What Munger Would Say
”You’ve built a beautiful map. Now stop admiring it and go check the territory.”
Munger would respect the analysis but demand field verification: calls to the county assessor, local land brokers, the planning office, Intel IR. Paper stress-testing is not territory work.
The Confirmation Bias Warning
Every file in the research builds the bull case and then stress-tests it on paper. Munger’s Darwin Rule applies: disconfirming evidence must be actively sought and immediately recorded, because your brain will try to forget it.
The Reverse Lollapalooza
Multiple converging catalysts sound great (a Lollapalooza effect). But Munger warns that the Lollapalooza works in reverse too. If Intel delays AND rates re-spike AND Dahl gets distracted AND water entitlements stall — each manageable alone, devastating in combination — the stock freezes at $26 for a decade.
The Survivorship Bias Question
“Who else owned 17,000 acres near a major employer and still lost money?”
Cases to study:
- Vail Ranch (Temecula, CA): Massive land bank near defense/tech employers, took 30+ years to monetize
- Land developers near military base closures (BRAC rounds)
- Land plays near auto plants that shuttered (Detroit suburbs)
- Developers who held through 2008 with “fortress” balance sheets and still saw land marked down 60-70%
The answer might be “those situations are different.” But do the work to prove it’s different rather than assume.
His Bottom Line
“The probabilistic framework is right. The margin of safety on liquid assets is genuine. But you’re sitting in your library when you should be in the field. Make the phone calls. Get closed bulk land transaction data. And if you find disconfirming evidence, write it down immediately.”
Probabilistic Outcome Table
| Outcome | Probability | Value/Share | Weighted |
|---|---|---|---|
| Bear: Stagnation, no catalyst 5+ years | 25% | $27 | $6.75 |
| Base: Steady monetization over 3-5 years | 45% | $31 | $13.95 |
| Bull: Acquisition/forced sale within 2-3 years | 20% | $40 | $8.00 |
| Catastrophe: Intel pulls out, water permanently constrained | 10% | $18 | $1.80 |
| Expected Value | $30.50 |
County-anchored expected value of ~$30.50/share vs. current ~$26. That’s a 17% expected return — before discounting for duration, illiquidity, and the 2.92M share float. After adjusting for a 3-5 year hold period and opportunity cost, this doesn’t clear the bar.
Why I’m Passing
The idea is real. The math works on paper. But three things keep me out:
1. Water Is a Structural Constraint, Not a Cyclical One
The county’s own plan says full buildout is impossible. The Pueblo ruling capped the city’s water permits. New Mexico’s groundwater basins are fully appropriated — no new rights are available. This isn’t a headwind that resolves with time; it’s a ceiling on the ultimate value of the land. The 10,543 acres of Outer Rim land — two-thirds of AMREP’s total holdings — may never have a development path.
2. The Margin of Safety Is Thinner Than It Appears
The county-anchored Bear case is $26.55/share — essentially flat with the current price. The old analysis showed a Bear case of $27 with water rights providing a floor. Without water, and with the county data forcing a more honest split between entitled and raw land, the downside protection narrows to roughly breakeven. You’re not getting paid enough for the illiquidity and duration risk.
3. The Upside Depends on Things I Can’t Verify
The bull case requires: (a) Dahl successfully forcing a sale, (b) Intel sustaining its Rio Rancho expansion, (c) a national homebuilder paying a strategic premium for land that may not have water access, and (d) the Pueblo ruling being overturned on appeal (resolution expected 2027-2030). That’s a lot of dominoes. The county-anchored bull case of $40/share is +53% upside, but over a 3-5 year period with low confidence in the probability of each catalyst.
The Water Framing
Here’s the cleanest way to think about it: AMREP is a land development company. Land development requires water. The county says there isn’t enough water. The courts agree. Until that changes, the 16,000-acre land bank is significantly impaired — not worthless, but capped in a way that limits the upside this idea needs to work.
It’s a net/net. The floor is real. But the upside — the reason to tie up capital for years in an illiquid stock — requires water that may not exist. I’d rather pass and be wrong than invest and be right about the assets but wrong about the constraint.
What Would Change My Mind
- Pueblo ruling overturned on appeal — If Rio Rancho’s water permits expand from 12,020 AF to 24,000 AF, the development constraint loosens significantly
- Verifiable water rights for AMREP — If the 2,797 AF Sub-File Order can be confirmed from primary sources, the idea strengthens materially
- Bulk land transaction above $6K/acre — Any demonstrated transaction at scale (1,000+ acres) at $6K+ would invalidate the Texas Pacific Land Trust critique
- Dahl files an amended 13D calling for a sale — This would signal Phase 4 is beginning and compress the timeline
- Stock drops below $20 — At that price, cash alone covers 85%+ of market cap and you’re getting all land for effectively free, water or not
Sources
SEC Filings
- AMREP 10-K FY2025
- All AMREP SEC filings (CIK: 0000006207)
Government Records
- Sandoval County EagleWeb Assessments
- Rio Rancho Estates Area Plan (RREAP, 2014)
- Rio Rancho Water Resource Management Plan (2004)
Market Data
- LandAndFarm.com — Sold Listings, Sandoval County
- LandWatch.com — Sandoval County
County Assessment Data
- Sandoval County EagleWeb, scraped March 2026: 19,466 parcels across all AMREP entities