Investing in Water Rights

In September 2022 the town council of Queen Creek, Arizona gave final approval to a $27 million purchase of Colorado River water. The water came from GSC Farm in Cibola, a strip of irrigated ground on the river in La Paz County, about 200 miles away. The seller was Greenstone Resource Partners LLC, whose backers include MassMutual and its subsidiary Barings and a number of public pension funds; through a subsidiary it had bought 485 acres there in 2013 and 2014 for about $9.8 million. Queen Creek got 2,033 acre-feet a year of fourth-priority Colorado River water — enough, the town said, for nearly 6,100 households.

Greenstone, by the reporting of KJZZ, made about $14 million — which is the $24 million reported for the 2018 entitlement agreement less the $9.8 million the land cost. The $24 million and the $27 million are not reconciled anywhere in the public record.

Nothing was manufactured, mined or grown. What Greenstone owned was the legal right to divert water onto 485 acres of desert alfalfa, and what it sold was permission to divert the same water into a suburban water system instead. The physical water never changed. Only its authorised place and purpose of use did, and that change carried the entire price. Three counties — La Paz, Mohave and Yuma — sued the Bureau of Reclamation over the approval, and on February 21, 2024 Judge Michael Liburdi of the US District Court for the District of Arizona ordered the bureau to replace its finding of no significant impact with a full environmental impact statement. The water kept flowing to Queen Creek while they argued.

That is the shape of this market. There is no exchange where an outsider buys a barrel of water and waits. There is a patchwork of state-law property rights, priority dates going back to the 1860s, irrigation-district bylaws, county ordinances and federal contracts, and the money is made in the gap between what a right is worth on a hayfield and what it is worth in a subdivision — minus a decade of permitting and litigation.

This guide covers the doctrines, the Colorado River compact now being rewritten by federal decree, the groundwater regimes in Texas, Arizona and California, the one index and the one futures contract that exist, the Australian market that works, who is making money and how, and whether you can join them. Investing in Farmland, the hub’s flagship, covers the farmland asset itself, and the sister guides are Investing in Farmland Through REITs and Funds, Buying Farmland Directly and Investing in Timberland. This one is about the thing that makes an acre of Western dirt worth $20,000 or worth $400.

A water right is not ownership of water. It is a usufructuary right: a right to use a public resource, granted by the state, defined by a set of parameters that all have to survive a transfer for the right to be worth anything to a buyer.

Six parameters define an appropriative water right in the West, and a purchase agreement that does not state all six is not a purchase agreement.

• Priority date. When the water was first put to beneficial use, or when the application that ripened into the right was filed. This is the whole ballgame in a shortage: an 1872 right is served in full while a 1955 right beside it gets nothing. Priority is not a tie-break, it is an absolute queue.
• Quantity. A diversion rate in cubic feet per second, or an annual volume in acre-feet. One acre-foot covers an acre a foot deep: 325,851 gallons, conventionally a year’s supply for two or three suburban households.
• Source. The named stream, ditch, well, aquifer or project the water may be taken from.
• Point of diversion. Where it comes out. Moving it requires state approval.
• Place of use. The specific parcels the water may irrigate or serve.
• Purpose of use. Irrigation, municipal, industrial, stock, recreation, environmental. Changing this was the transaction at Queen Creek, and it is the hardest of the six to move.

Beneficial use and the “use it or lose it” trap

Every prior-appropriation state conditions the right on continued beneficial use. Stop using the water and the right can be lost by abandonment (intent inferred from a long period of non-use) or by statutory forfeiture (a fixed run of consecutive non-use years, commonly five — Utah, Oregon and Washington all use five — while Colorado runs a ten-year abandonment presumption through its water courts; verify the statute where you are buying).

This is the most counter-intuitive feature of the asset: idle capacity is not a stored option, it is a decaying claim. An owner who buys a right and lets the field go dry while waiting for a city to want it can find the right shrunk to the quantity actually used, or gone. Serious holders keep the water in use — often leasing it back to a farmer at a rent nowhere near the cost of capital — purely to preserve the right.

The no-injury rule, and why “consumptive use” is the number that matters

Every transfer in a prior-appropriation state runs into the no-injury rule: the change may not injure other rights on the stream. An irrigator diverts more water than the crop consumes, and the remainder returns to the river as tailwater and deep percolation, on which downstream users have come to depend. So when a right moves, the transferable quantity is not the paper diversion but the historical consumptive use — what the crop consumed and never returned. A 1,000 acre-foot decree can transfer as 400 acre-feet, and the engineering study that fixes that number is the most consequential document in the deal.

Paper water and wet water

The oldest disappointment in this market is the gap between a decree on file and water in a ditch. Paper water is a right that has not been exercised in years, is junior enough that it is rarely in priority, sits on a source that no longer delivers, or is trapped by a district’s internal rules. Wet water is a right with a delivery history: gauge records, district billing statements, crop records, a pump log. Ask for eight to ten years of those before you ask for anything else.

American water law splits at roughly the 100th meridian, and the split determines whether a water right can be an investment at all.

Riparian: rights that cannot leave the riverbank

In the humid states, water rights derive from land ownership. A riparian right belongs to whoever owns land touching a watercourse and entitles the owner to reasonable use of the water on that land. It carries no priority date, is not quantified, and cannot generally be severed from the land and sold. In a drought, riparians share the shortage proportionally rather than by seniority. That makes a riparian right a use privilege attached to a deed, not a tradeable asset. There is essentially no investable riparian water market in the United States, and any pitch built on Eastern water scarcity is selling utility equity or infrastructure, not water rights.

Prior appropriation: first in time, first in right

The mining camps of the 1850s and 1860s needed water away from the streambank, so the West built a different system. Under prior appropriation, a right is created by diverting water and putting it to beneficial use; the earlier the use, the more senior the right; and in a shortage the most senior right is filled completely before the next gets a drop. The National Agricultural Law Center’s summary is the standard statement of it: rights are granted based on when a person puts water to beneficial use, and continue as long as that use continues.

Because the right is in principle severed from the land — a property interest in its own right — it can be sold, leased, moved and changed. That severability is what makes an investable asset possible. It is also constrained in practice by the no-injury rule, by state engineers, by irrigation-district covenants and by area-of-origin protections, which is why the market is far thinner than the theory suggests.

The hybrid states, and why the label matters

A group of states adopted riparian rights first and then layered prior appropriation on top, converting or grandfathering the riparian claims. The National Agricultural Law Center lists them as California, Kansas, Nebraska, North Dakota, South Dakota, Oklahoma, Oregon, Texas and Washington. The remaining western states — Alaska, Arizona, Colorado, Idaho, Montana, Nevada, New Mexico, Utah and Wyoming, nine by our count on that list — operate pure prior appropriation.

The label matters because in a hybrid state the seniority queue can be interrupted by pre-code riparian claims that were never quantified. California is the extreme case: rights initiated before the Water Commission Act of 1913 took effect, on December 19, 1914, sit outside the state permit system entirely, and the State Water Resources Control Board’s jurisdiction over them is narrower than over post-1914 appropriative rights — a limit California appellate courts have reaffirmed. Pre-1914 rights on the Sacramento–San Joaquin system are among the most valuable water assets in America precisely because they sit above the state’s permitting machinery — and they are almost never for sale.

Groundwater is a separate legal system again

None of this necessarily applies underground. Groundwater runs on doctrines of its own — absolute dominion in Texas, correlative rights for California overlying landowners, appropriation with dated priorities in Colorado’s designated basins — with permit-and-management regimes layered on top. An investor who buys “an irrigated farm with good water” in Texas and one who buys the same sentence in Colorado have bought two entirely different legal animals.

The Colorado River is the central fact of Western water, and its allocation has been arithmetically impossible since the day it was written.

The Colorado River Compact of 1922 divided the river between an Upper Basin and a Lower Basin, guaranteeing each the beneficial use of 7.5 million acre-feet a year in perpetuity. The Lower Basin got an additional 1 maf to cover Gila River diversions, taking the compact’s own allocation to 16 maf; a 1944 treaty then gave Mexico 1.5 maf. Total commitments: 17.5 maf a year.

The river has never carried that. The negotiators were working from records of the 1900s and 1910s, an anomalously wet stretch in which annual flows ran as high as about 20 maf. The long-term mean is roughly 15 maf, and as of 2022 the 20-year running average had fallen below 13 maf. Set the commitments against the water and the gap is the whole story of Western water: 17.5 maf promised against a long-run mean of 15 leaves 2.5 maf of paper with nothing behind it, and against the 20-year average to 2022 of below 13 it leaves more than 4.5 maf — better than a quarter of everything allocated. That is our arithmetic on the compact, the treaty and the published flow figures. Tree-ring reconstructions suggest decade-long stretches below 13–14 maf are normal, not exceptional.

Layered on the compact is what practitioners call the Law of the River: the 1928 Boulder Canyon Project Act, Arizona v. California, 373 U.S. 546 (1963), which fixed the Lower Basin mainstream apportionments at 2.8 maf for Arizona, 4.4 maf for California and 300,000 acre-feet for Nevada and made the Secretary of the Interior the river’s watermaster; the 1944 Mexican treaty; tribal reserved rights under Winters v. United States, 207 U.S. 564 (1908), which held that reserving land for a reservation impliedly reserves the water needed to make it a homeland; and the 2007 Interim Guidelines that set shortage-sharing rules for Lake Mead and Lake Powell. About 5.5 million acres of irrigated farmland depend on the river.

2026: the year the rules ran out

The 2007 Interim Guidelines and the agreements built on them expire at the end of 2026. Seven states spent four years failing to agree on what replaces them. On August 21, 2026, the Department of the Interior stopped waiting and adopted a decision framework for operating Lake Powell and Lake Mead through 2036, together with operating guidelines for 2027 and 2028. The Record of Decision cites continuing drought, historically low reservoir levels, concerns about the reliability of infrastructure at Glen Canyon Dam, and the states’ inability to reach a cooperative agreement.

The framework sets general operating principles and “sideboards” and then issues guidelines in two-year intervals through 2036 — which is itself the important design choice, because it converts a decade-long rulebook into a rolling series of short-term administrative decisions.

The 2027–28 guidelines allocate consumptive use of 2.04 maf to Arizona, 3.96 maf to California and 250,000 acre-feet to Nevada: reductions of 27.1%, 10% and 16.7% against the original allocations, which are the 2.8 maf, 4.4 maf and 300,000 acre-feet decreed in Arizona v. California. On top of about 1.25 maf a year of reductions, the Lower Basin states are to voluntarily conserve and store at least 700,000 acre-feet across the two years, and a federally managed pool in Lake Mead, capped at 450,000 acre-feet, offsets up to 25% of reductions to Tribes in the Lower Basin.

The litigation that follows

Three days later, on August 24, 2026, Nevada sued. The State of Nevada, the Colorado River Commission of Nevada and the Southern Nevada Water Authority filed in the US District Court for the District of Nevada against Interior’s Record of Decision, arguing that it would take 71% of the 300,000 acre-feet Nevada now receives while the four Upper Basin states face no mandatory cuts at all. The complaint pleads violations of the Administrative Procedure Act, the National Environmental Policy Act and the Law of the River, and asks the court to set the Record of Decision aside.

The Upper Basin’s position, stated repeatedly by its negotiators, is that Interior lacks statutory authority to override the compact or dictate state-level cuts, and that the 1922 compact does not oblige the Upper Basin to guarantee a delivery volume when hydrology fails to provide it.

For an investor the ROD is the most important repricing event in Western water in fifty years, and it is unresolved. A federal framework that reallocates by administrative decision every two years, under litigation throughout, is the opposite of the stable, adjudicated, priority-dated property system that makes a water right financeable. It also means every deal underwritten before August 2026 on Colorado River supply carried the wrong assumption.

Three American groundwater regimes produce three different assets and three different ways to lose money: Texas’s rule of capture, where the water under the land belongs to the landowner; Arizona’s Active Management Areas, where pumping is proved up inside the cities and unlimited outside them; and the essentially unmanaged High Plains, where the resource is simply being mined.

Texas: the ground beneath your feet is yours

Texas surface water belongs to the state and is appropriated by permit. Groundwater is different. Under the rule of capture, inherited from English common law, a Texas landowner may pump as much as he beneficially uses, even if that drains the neighbour’s well. Groundwater conservation districts, created by the legislature district by district, may impose spacing and production limits — and that is where the conflict arose.

On February 24, 2012, in Edwards Aquifer Authority v. Day, the Texas Supreme Court held unanimously, in an opinion by Justice Nathan Hecht, that a landowner has a constitutionally protected property interest in groundwater in place, before it is pumped. The court reasoned by analogy to oil and gas: groundwater is “an exclusive and private property right … inhering in virtue of [the landowner’s] proprietorship of the land, and of which he may not be deprived without a taking of private property.” The practical consequence, flagged immediately by every Texas water lawyer who wrote about it, is that a district’s production limit is now exposed to a regulatory-takings claim.

For an investor this cuts both ways. It makes Texas groundwater the closest thing in America to a fee-simple water asset, which is why speculative groundwater plays cluster in the Panhandle and West Texas. It also means the resource is being drawn down by owners who each have a private incentive to pump first, and Day makes the political fix — regulation — more expensive to enact. You are buying a well-defined right to a shrinking pool.

Arizona: managed where the people are, open where they are not

Arizona passed its Groundwater Management Act in 1980 under pressure from falling water tables and the risk of losing federal authorisation for the Central Arizona Project. It created four Active Management Areas — Prescott, Phoenix, Pinal and Tucson — and a fifth, Santa Cruz, was split from Tucson in 1995. Together they cover less than a quarter of the state’s land but more than 75% of its population. Inside an AMA, new subdivisions must demonstrate an assured water supply sufficient for 100 years, and four of the five AMAs were required to reach “safe yield,” where withdrawal roughly matches recharge, by January 1, 2025. That deadline passed unmet.

On June 1, 2023, the Arizona Department of Water Resources published a new 100-year groundwater model for the Phoenix AMA showing demand exceeding physically available supply by about 4% by 2121, concentrated on the urban fringe. Governor Katie Hobbs announced simultaneously that ADWR would stop processing assured-water-supply applications for new subdivisions in the Phoenix AMA that relied on groundwater. On April 22, 2026 a Maricopa County Superior Court judge, Scott Blaney, struck that policy down, holding that the department had no legal authority for the “unmet demand” rule it had used to decide whether a subdivision had a hundred years of groundwater. Reporting through mid-2026 described what follows as unsettled.

The legislature had already built a partial bypass around the same problem. Arizona’s Ag-to-Urban law, SB1611, effective September 26, 2025, lets holders of irrigation grandfathered rights in the Phoenix and Pinal AMAs retire those rights in exchange for groundwater savings credits a developer can apply against the physical-availability test for an assured water supply. That is the farm-to-city trade written into statute and confined to the two AMAs where the houses are going. Between a struck-down rule and a new statutory credit, treat the Phoenix position as contested rather than settled.

Outside the AMAs, most of rural Arizona has essentially no pumping limit, which is why the water-grab stories cluster there. That regulatory asymmetry — unlimited pumping in the countryside, a 100-year proof requirement in the cities — is the arbitrage the private buyers described later in this guide are running.

The Ogallala: a slow, arithmetically certain drawdown

The High Plains (Ogallala) Aquifer underlies about 112 million acres across eight states — Colorado, Kansas, Nebraska, New Mexico, Oklahoma, South Dakota, Texas and Wyoming — and supplies roughly 30% of the groundwater used for US irrigation. Recharge is negligible relative to pumping across most of its southern extent, so the water level is a one-way series. Water tables have fallen more than 200 feet in parts of Texas and Kansas.

The Kansas Geological Survey’s annual January measurements are the cleanest public read on the rate. Measured in January 2025, the change over calendar 2024 was a decline of roughly a foot across the western Kansas High Plains region: northwest Kansas down 1.34 feet, against 0.47 feet the year before, and the southwest Kansas management district down 1.52 feet, against 1.43 feet the year before. The 1996–2024 regional average is −0.42 feet a year. The following year broke the run: measured in January 2026, western Kansas gained 0.18 feet over 2025 and south-central Kansas nearly 2.5 feet, on above-average rainfall — the first overall increase across the Kansas High Plains aquifer since 2019, and a reminder that a wet year interrupts the series without changing its direction.

There is no Ogallala water market to speak of, and that is the point. Where groundwater is owned by the overlying landowner and cannot practically be moved to a city hundreds of miles away, depletion does not create a tradeable asset — it destroys the value of the land above it and reprices the land somewhere else. An investor’s exposure to the Ogallala is best expressed as an avoidance rule, not a position.

California did not regulate groundwater extraction at the state level until 2014, and the law that ended that is now redrawing the value of every irrigated acre in the San Joaquin Valley.

The Sustainable Groundwater Management Act of 2014 requires locally formed groundwater sustainability agencies to write plans that bring their basins into balance — pumping equal to recharge, without “undesirable results” such as land subsidence or dry domestic wells. Critically overdrafted basins must reach sustainability by 2040; the rest by 2042. That is not a distant deadline in an asset with a 30-year hold; it is inside the horizon of anything you buy today.

What balance costs the San Joaquin Valley

The Public Policy Institute of California has done the arithmetic on what balance costs. Its September 2023 report on managing water and farmland transitions puts the decline in average annual farm water availability in the Valley at 3.15 million acre-feet by 2040, roughly 20% of current supplies, with SGMA accounting for 2.7 maf of that and climate change and Delta environmental flows accounting for the rest. Even under optimistic assumptions about new supply, about 500,000 acres of irrigated farmland — around 10% of the Valley’s irrigated footprint — must come out of full irrigation. Without new supplies, PPIC puts the figure at 900,000 acres. The cost it estimates is a $4.5 billion decline in farm GDP and 50,000 lost jobs.

Probation, and the fight over enforcement

SGMA has teeth: if a basin’s plan is inadequate, the State Water Resources Control Board can put it on probation, collect extraction reporting directly, and charge pumping fees. On April 16, 2024 the board designated the Tulare Lake Subbasin probationary, the first basin in the state to be so designated, after the Department of Water Resources found in 2023 that plans for six critically overdrafted basins would not achieve sustainability by 2040.

The Kings County Farm Bureau sued. In September 2024 it won a preliminary injunction that stopped the board from taking any action in the subbasin. On October 29, 2025 the Fifth District Court of Appeal reversed that injunction in companion opinions and ordered the trial court to sustain the state’s demurrer, gutting the Farm Bureau’s core civil claims.

On August 12, 2026 Judge R. Shane Burns of the Kings County Superior Court paused enforcement again, this time against two of the five groundwater sustainability agencies in the subbasin. The merits are set to be heard on October 6, 2026, a date still standing as of September 10, 2026, and the Farm Bureau has said it will take the case to the California Supreme Court if it loses.

Read the litigation the way a buyer should: SGMA’s substance is not seriously in doubt after the appellate ruling, but its pace is being fought basin by basin, and the pace is what determines whether a given orchard has water in 2032. The flagship farmland guide covers what this has done to California permanent-crop returns; the point for a water investor is the mirror image. Every acre-foot of pumping that SGMA extinguishes raises the value of a secure surface right or an adjudicated groundwater allocation in the same valley. That is the clearest structural bid in US water, and it has a date on it.

There is no total-return index for US water rights, and any pitch that implies otherwise is selling you something. What exists is a set of local price series, none of them investable, and one foreign market with a real record.

The one public US price

The Nasdaq Veles California Water Index (NQH2O) began on October 31, 2018. It tracks the volume-weighted average price of water-rights lease and sale transactions across the five largest and most actively traded regions of California: the surface water market and four adjudicated groundwater basins. It is a modified volume-weighted average, published in dollars per acre-foot and disseminated once a week, after the close on Wednesday.

Its record is a drought instrument. In the 2021–22 drought the index ran to $1,144.14 an acre-foot on June 27, 2022, up 56% in six months, which Bloomberg Law reported at the time as an all-time high. Wet years take it back down. What the series demonstrates is not a return but a distribution: California water is a volatile, weather-driven spot price with a wide range and no carry, which is the profile of a commodity rather than of an income asset.

Our tape

We store the index daily. On September 1, 2026 our tape read 535.91; on September 2 it read 566.37, and it read 566.37 again on September 3, 4, 5 and 7. Those five identical prints are not a still market — they are the index’s own weekly publication cadence showing through a daily collection: Nasdaq disseminates NQH2O once a week after Wednesday’s close, and September 2, 2026 was a Wednesday. Read that as the first practical lesson about the asset: the only public price for US water refreshes about once a week, which tells you how few transactions there are to average. It is also why our water sleeve is still marked awaiting in the composite at a 0.3% target weight, the smallest of the three land sleeves on this hub, against farmland’s 0.6% and timberland’s 0.8%.

Against the 2022 peak of $1,144.14, the September 7, 2026 print of $566.37 is just under half — 49.5% of it. Do not read that as a trend: the index is a spot lease price and its level is mostly a function of the water year. It is a stress gauge, not a valuation.

The permanent-rights markets

Leases price a year; permanent rights price the perpetuity, and the numbers are an order of magnitude apart. The Colorado-Big Thompson market on the northern Front Range is the deepest municipal water market in the United States. WestWater Research, which tracks Western water transactions, puts CBT units at about $85,000 an acre-foot in 2025, down from a high of $101,000 in 2022 — a fall of about 16% in three years.

Read that direction carefully, because it is the most useful single fact in this guide about how the asset prices: the deepest permanent water market in the country fell by a sixth during the drought that took the California lease index to its record. Scarcity headlines and permanent-rights prices are not the same series. That is a consultancy’s transaction tabulation reported in the press rather than an audited index; individual trades are reported across a considerably wider range.

Queen Creek gives a second permanent benchmark: $27 million for 2,033 acre-feet a year is about $13,300 per acre-foot of annual entitlement — about $11,800 if you use the $24 million entitlement price reported for the 2018 agreement rather than the $27 million purchase the council approved in 2022. A third: Limoneira, the listed California citrus grower, said in January 2025 that it had monetised part of its roughly 21,000 acre-feet of senior and adjudicated water rights at $30,000 an acre-foot — a company-reported figure for adjudicated California water, and a reminder that “the price of water” means nothing without the basin and the seniority attached.

A fourth number is not a water price at all, and is worth keeping beside the other three for that reason. The San Diego County Water Authority’s June 2025 settlement with Metropolitan, ending a fifteen-year rate fight, fixes a price starting at $671 per acre-foot in 2026, escalating with a stated consumer price index, to move the 227,000 acre-feet a year of water the Authority bought from Imperial Irrigation District in 2003 — a transportation price, not a water price, and a reminder that conveyance is a separate and sometimes larger cost than the water.

What a real record looks like: Australia

Australia is the only country with a mature, unbundled, liquid water market and therefore the only place a genuine return series exists. The Aither southern Murray–Darling Basin Water Entitlement Index has tracked entitlement prices since 2008, and over the period to June 2019 water entitlements generated an average yield of approximately 4%, according to a CFA Institute analysis of the asset class. The index has since passed to Ricardo, which acquired Aither; on its reporting the entitlement index fell about 4% in 2023-24, a second consecutive annual fall, then rose 5.7% in 2024-25, its first annual increase in three years.

Four percent of yield, from the deepest water market on earth, with capital values that go down as well as up, is the honest benchmark. Anyone quoting you a double-digit water yield is quoting a capital gain from a repricing event, not a yield.

CME Group listed the first exchange-traded water futures in the world, and the contract has never worked. Understanding why is the most efficient way to understand what water is not.

The NQH2O futures are cash-settled against the Nasdaq Veles California Water Index. Each contract represents 10 acre-feet of water. CME announced the launch in September 2020 and the contract began trading on Globex on December 7, 2020. CME planned a market-maker programme to keep two-sided quotes. The contract received extraordinary press at launch — Pedro Arrojo-Agudo, the UN special rapporteur on the human rights to safe drinking water and sanitation, objected publicly within days that water “belongs to everyone and is a public good” and should not be exposed to financial speculation — and then went quiet.

The academic literature is blunt about the outcome. Jingjing Wang and Xiaoyang Wang, writing in Applied Economic Perspectives and Policy in 2022 under the title Why is water illiquid? The NQH2O water index futures, document a market with scarce trading volume from the start — daily volume in the low tens of contracts in the first months, and a futures market amounting to a single-digit percentage of the physical short-term water trade it was written on. Their diagnosis is a lack of effective hedging demand, given the heterogeneity and small size of the spot markets, compounded by the opacity of the settlement index. The reasons are structural, and all of them are lessons about the underlying asset.

• Nothing is delivered. A cash-settled contract on an index of local transactions has no arbitrage anchor. You cannot buy the index, so a mispricing cannot be squeezed out.
• The index is an average of five separate markets. A farmer in the Central Valley hedging a lease price faces basis risk against a benchmark that blends four adjudicated groundwater basins and the surface market. If the hedge does not track your exposure, you do not put it on.
• The natural hedgers do not have a hedgeable exposure. A city’s water cost is set by district rates and long-term contracts, not by spot leases. A farmer’s water cost is set by his district’s allocation. Very few participants have a mark-to-market spot water exposure that a financial hedge would smooth.
• The index refreshes weekly. A weekly print cannot support a daily-margined derivative with any confidence, as our own tape shows in miniature.
• The population of participants is tiny. The parties to the transactions behind the index are irrigation districts, farmers and municipalities, none of which has a mandate to trade futures.

The failure is not a marketing problem. It is the market telling you that water is not a commodity in the financial sense: it is a bundle of local, legally distinct, non-fungible entitlements that cannot be transported, standardised or delivered. Every scheme to securitise water runs into that same wall.

Australia did what the United States has not: it separated the water right from the land title, made both tradeable, and built a public registry. The result is the only water market in the world that an outside investor can enter on ordinary terms.

Entitlements and allocations

The southern Murray-Darling Basin runs a two-tier system that is worth learning even if you never trade it, because it is the vocabulary every serious water conversation uses.

• An entitlement (or water access right, or in Victoria a water share) is the perpetual claim: a share of whatever the resource yields, held on a public register, separable from land, mortgageable. This is the capital asset.
• An allocation is the water actually announced against that entitlement in a given season, expressed as a percentage of the entitlement volume. This is the annual crop, and it is traded separately, in megalitres.

Entitlements come in reliability classes. In Victoria, high-reliability water shares (HRWS) are announced first and most fully; low-reliability shares (LRWS) get water only in good years. The price gap between them is the market’s estimate of the value of seniority, and it is enormous: above the Barmah Choke, high-reliability entitlement averaged around A$4,900 per megalitre in the 2025-26 water year against about A$950 per megalitre for low-reliability, per Integra Water Services’ price pages, updated July 20, 2026. That is a 5.2× premium for reliability on the same river. Read across to the United States: that ratio is what a priority date is worth when a market is allowed to price it.

Allocation prices, the annual rent, are far more volatile. Aither’s annual water markets reporting put the average 2023-24 allocation price at A$76 per megalitre against a long-term average of A$166. Turnover is real money: Australian water markets had an estimated turnover of over A$4 billion in 2021-22, down from around A$6 billion in 2020-21, per the Bureau of Meteorology, whose regular national reporting was subsequently paused for water market reforms. The 2025 Ricardo Water Markets Report found that Commonwealth environmental buybacks drove an increase in market turnover in 2024-25, the first in five years, identifying at least 22 of 198 gigalitres of southern Basin trade volumes as buyback-related, and noted the first annual increase in groundwater entitlement activity since 2019-20.

Why it works, and why the US cannot copy it

Two features make it work, and neither exists in the US. First, unbundling: the entitlement is a registered, standalone, perpetual title, so a buyer does not have to buy a farm to buy water. Second, a single administering authority per system with public registers and defined trade rules, so a transfer is a registry entry rather than a change-of-use adjudication. American water is administered by fifty state engineers, thousands of irrigation districts, dozens of federal contracts and a river compact under litigation. That is why an Australian entitlement transfers in weeks and a US change of use takes years.

Australia has its own politics: Commonwealth buybacks move prices directly, foreign-ownership registers and periodic parliamentary inquiries into “water barons” are a standing feature, and the Millennium Drought of 1997 to 2009 is the reference stress event. But a 4% average yield with a public register and a working transfer process is a better asset than a 10% story with no way in.

Four business models have been tried on US water at scale. Three have produced real returns and none of them is “buy water and wait” — each is a permitting business wearing a commodity costume. The fourth has produced thirty years of losses, and is the most instructive of the lot.

Farm-to-city arbitrage

Buy irrigated farmland cheaply for its water, then spend years converting the water’s purpose of use to municipal and selling it to a growing city at a municipal price. This is the Greenstone trade at Queen Creek, and it is the model Water Asset Management LLC has industrialised. The New York firm bought 12,793 acres — nearly 20 square miles — in La Paz County’s McMullen Valley Basin for about $100 million in cash, a purchase that closed on July 19, 2024 in the name of a Delaware entity, Emporia III LLC, and drew immediate “water grab” coverage from Arizona outlets and the Water Education Foundation.

The firm’s own materials describe roughly 40,000 acres of farmland with senior sustainable water rights across its private-equity strategy; that is a self-reported figure and we label it as such. It has also bought land near the Colorado River in Mohave County, and it attempted to sell water rights from there to the Central Arizona Project’s governing agency — a deal the agency killed after public protest.

The legislature has been the battleground since. In the 2026 session two bills, HB2757 and HB2758, would have opened a transfer path out of the McMullen and Butler valleys; HB2758 passed the Arizona House on February 19 and cleared a Senate committee 4–3 on March 17, then died. By August 2026 a local coalition was gathering signatures to create a groundwater protection zone over the McMullen Valley aquifer instead.

That killed deal is the model’s defining risk. The asset is only worth the municipal price if a public body will approve the transfer, and public bodies are politically responsive to exactly the constituency that objects to the trade.

Water for houses, bought by the housebuilder

The cleanest exit in the sector’s history was not to a city but to a homebuilder. Vidler Water Resources — formerly PICO Holdings, incorporated 1981 and based in Carson City, Nevada, which renamed itself in March 2021 — spent decades assembling water rights and storage in Nevada, Arizona, Colorado and New Mexico. On April 14, 2022, D.R. Horton agreed to acquire it for $15.75 a share in cash, an equity value of about $291 million, a 19% premium to the 90-day volume-weighted average price and a 39% premium to book value. The tender offer expired on May 24 and the merger closed before the market opened on May 25, 2022.

Two readings of that price are both correct. It proves that patiently assembled Western water rights are worth a premium to book to a strategic buyer who needs them. It also shows how narrow the buyer universe is: after decades of work the trade cleared at 1.39× book to the one category of acquirer that cannot build without an assured water supply, with no competing bid visible in the public record. That is not a liquid asset with a clearing price; it is a bilateral negotiation with one counterparty.

The water utility, which is a regulated business

The third model is the least glamorous and the only one with a public market: own the pipes and the regulated rate base. It is what the water ETFs actually hold, as the routes section shows — a utility investment with a utility’s return profile, and almost nothing to do with the price of a water right.

The fourth model: a resource that has never paid

Cadiz Inc. owns land over an aquifer at the base of a 2,000-square-mile Mojave watershed. Since the 1990s it has proposed conveying that water to Southern California cities; it now describes the Mojave Groundwater Bank as offering 2.5 million acre-feet of new water supply against an estimated 30 million acre-feet in storage — the company’s own characterisation of its own project. It has been opposed in turn by environmental groups, the Metropolitan Water District, federal land managers over the right-of-way, and litigation, with the Pacific Institute maintaining a standing critique of the plan.

Read the financial record first. For the six months ended June 30, 2026, Cadiz reported revenue of $2.6 million and a net loss applicable to common stock of $22.6 million. The comparable 2025 half was a $19.9 million loss, so the deficit is widening, and revenue fell from $4.1 million in the second quarter of 2025 to $979,000 in the second quarter of 2026 as filter sales dropped away.

The balance sheet is the harder read. At June 30, 2026 it held $5.3 million in cash and working capital of $1.6 million against long-term debt of $87.0 million, on which its effective interest rate for the half was 15.4%. Project financing arrives in tranches: a definitive agreement for up to $51 million from the Lytton Rancheria of California, described by Cadiz as the first tranche of an equity raise of up to roughly $450 million.

None of that makes Cadiz a fraud. It makes it an accurate picture of the cost structure: a resource of genuine physical scale, a coherent commercial plan, a thirty-year permitting fight, and a loss profile that would have exhausted almost any private investor long ago. The asset is real. The option on it has been very expensive to hold.

Most of what is marketed as water investing is not water. Here is the full menu, ranked by how much genuine water-right exposure it delivers per dollar, with the honest verdict on each.

1. Farmland with a senior, transferable, wet right

This is the only route that gives a retail buyer direct, unlevered exposure to the value of a water right, and it is the route the professionals use. The water is bought inside a farm; the farm produces rent that pays the carry and keeps the right in beneficial use; and the water’s optionality is free. The sister guide on buying farmland directly covers the acquisition mechanics; what a water buyer adds to that process is a title search on the water right itself, an engineer’s consumptive-use opinion, and a delivery history. Minimums are the price of a farm.

2. The two listed farmland REITs, one of which reports its water

Gladstone Land (LAND) is the closest thing to a listed water-rights vehicle in the US market: as of its second-quarter 2026 reporting it held about 98,000 acres across 142 farms in 14 states and roughly 56,000 acre-feet of banked water and water assets, largely in California, which it discloses as an asset in its own right. Farmland Partners (FPI) held about 70,100 acres in 11 states at June 30, 2026 with a much smaller water story. On our tape, FPI closed at $10.65 and LAND at $9.83 on September 4, 2026. Both are equities with leverage, management fees and REIT tax treatment, and the water is a minority of the asset value — but LAND is the one place a retail account can own California water inside a listed wrapper. The sister guide on farmland REITs and funds covers the vehicles properly.

3. Listed companies whose business is water rights

A very short list, getting shorter: Vidler was acquired in 2022 and is gone, and Cadiz (CDZI) remains listed with the financial profile above. Pure Cycle (PCYO) holds water and land serving Front Range development east of Denver, carrying its water rights at about $32.9 million and describing them as supporting up to 60,000 connections, and it added 1,635 adjudicated acre-feet from the Box Elder Creek alluvium in a December 2025 water-court settlement. Limoneira (LMNR) farms roughly 7,000 acres in California, Arizona and Argentina and holds about 21,000 acre-feet of senior and adjudicated water rights, part of which it monetised in January 2025 at $30,000 an acre-foot. Those are company-reported figures; read the filings rather than our characterisation. The category is small enough that single-company risk dominates any thesis about water.

4. Water ETFs — which do not hold water

This is the most common mistake in the category. Invesco Water Resources (PHO) tracks the Nasdaq OMX US Water Index, charges 0.59% and held roughly $1.99 billion in assets across about 38 holdings as of July 2026, led by Ferguson, Xylem and Ecolab. First Trust Water (FIW) tracks the ISE Clean Edge Water Index — the top 36 industry names by market capitalisation, equal-weighted within five tiers — charges about 0.50% on roughly $1.8 billion, and its largest positions in 2026 are Waters Corp (5.40%), Agilent Technologies (5.15%), Roper Technologies (5.04%), American Water Works (4.32%) and Veralto (4.19%). Invesco also runs a global version, CGW, at about $1.0 billion and 0.58%.

Look at that holdings list. Roper is a diversified software and industrial-technology company. Waters and Agilent are analytical-instrument makers. Veralto sells water-quality and product-identification instruments. American Water Works is a regulated utility; Ferguson, PHO’s largest position, distributes plumbing supplies. These are good businesses, and a fund of them is a perfectly reasonable industrial-equity holding. It is not water-price exposure, it will not rise when a drought lifts a lease price, and it correlates with the equity market, not with hydrology.

5. Private water funds

Water Asset Management and a handful of peers run private vehicles buying farmland-with-water in the West. These are the real thing, and they are institutional: accredited or qualified-purchaser only, closed-end, long-dated, with the political risk described above and, in this sector, an unusual amount of reputational risk. We could not verify current minimums, fee terms or fund-level returns for any US private water strategy, and we are not going to characterise terms we have not seen. If you are shown one, ask for the realised gross and net IRR of every prior fund, the number of change-of-use applications filed and the number approved, and the average months from filing to approval. Those three answers describe the business.

6. Australian water entitlement funds

The one country with an unbundled, registered, liquid market also has listed and unlisted funds holding entitlements. For a US taxable investor that brings currency risk, foreign-ownership registration, non-US tax reporting and a policy variable — Commonwealth buybacks — that moves prices directly. Cleanest exposure available anywhere, most operationally awkward for an American. We have not verified any specific vehicle or its terms, and will not describe what we have not seen.

7. Futures, tokens and anything promising a water yield

NQH2O futures are available and effectively untradeable at retail size, for the liquidity reasons above. And there is no securitised US water right paying a yield: if a product claims one, the yield comes from something else — a farm lease, a loan, or the sponsor’s balance sheet — and you should identify which before you buy.

The cost stack in water rights is unusual in two ways: the round-trip friction is dominated by professional fees rather than commissions, and the carrying cost has to be paid whether or not the asset produces anything.

Acquisition friction

A purchase carries a brokerage commission (Front Range brokers quote low single digits; we model 3% and label that our assumption), title and closing work on the right itself, and — the item with no analogue in other assets — an engineering and hydrology study to establish historical consumptive use. That study plus water-rights counsel is routinely a five-figure cost on a modest transaction and six figures on a large one, and it is not optional: without it you do not know what quantity is transferable, which is to say you do not know what you bought.

The change-of-use application

Where the value is created, so is the cost. Changing a purpose or place of use requires an application to the state engineer or, in Colorado, a decree from a water court, with published notice, a statutory opposition period and objectors who will appear. Legal and expert costs run through the whole proceeding, which takes years. The Queen Creek transfer needed federal environmental review and still drew a three-county lawsuit that ended, in February 2024, in an order to replace it with a full environmental impact statement. Budget the calendar, not just the fee.

Annual carry

Rights held in a project or district carry an annual assessment whether or not water is taken. Northern Water levies one per Colorado-Big Thompson unit and publishes the rate each year: for 2026 it is $36.07 a unit at the open irrigation rate and $56.17 a unit for municipal, industrial and multipurpose use. Districts elsewhere charge per acre-foot of allotment or per assessed acre, and property tax may apply where the state treats the right as real property. Against that a leased-out right earns rent, usually small relative to capital value and taken partly to keep the right in beneficial use rather than for the income.

Fee drag on the wrappers

The listed routes carry ordinary wrapper costs — 0.59% for PHO, about 0.50% for FIW, 0.58% for CGW, plus the internal cost structure and leverage of the farmland REITs. Private water funds carry private-equity economics; we have not verified specific terms and do not quote them.

US tax treatment of water rights turns on two questions. The first is whether state law calls the right real property, which decides both the exchange treatment and the character of the gain. The second is whether the water itself is being physically exhausted, which decides whether the owner gets a depletion deduction — and over one American aquifer the answer to that has been yes since 1965.

The rate on a sale

A water right is a capital asset, not a collectible. That matters: the 28% maximum rate under IRC §1(h)(4) applies to collectibles such as art, coins, wine and cars, and it does not apply here. A long-held water right sold at a gain is taxed at ordinary long-term capital gains rates — 0%, 15% or 20% depending on income — plus the 3.8% net investment income tax under IRC §1411 where the thresholds are met. If the right was held as part of a trade or business rather than for investment, §1231 treatment and depreciation recapture questions arise; take advice on that.

Section 1031, and the state-law hinge

The 2017 Tax Cuts and Jobs Act limited like-kind exchanges under IRC §1031 to real property. Whether a water right qualifies therefore depends on whether the state treats it as a real property interest. The final regulations under Treas. Reg. §1.1031(a)-3, issued in 2020, define real property for this purpose to include certain perpetual interests in land. The principle practitioners work from is that a perpetual water right, in a state whose law treats appropriative rights as real property, is generally like-kind to a fee interest in land and can be exchanged for other real estate; a right to a specific quantity of water, or to water for a limited period, generally cannot. Colorado and California are the jurisdictions most often cited as treating appropriative rights as real property.

The authority behind the March 2023 Grant Thornton note is PLR 202309007, in which the IRS concluded that a licence to divert and use water in perpetuity was real property and like-kind to a fee simple interest, reasoning from Rev. Rul. 55-749 and turning on the fact that the right was not to a specific total volume or to water for a limited period. Note what a private letter ruling is: under IRC §6110(k)(3) it binds the IRS only as to the taxpayer who requested it and may not be cited as precedent by anyone else. It tells you how the Service thinks; it is not authority you can rely on. Have counsel confirm the treatment for your right in your state.

This is genuinely valuable. A farmland owner selling a farm can exchange into water rights, and vice versa, deferring the gain — which is one reason water rights trade at prices that look disconnected from their income. The hub’s 1031 guide covers the mechanics of the exchange itself.

Depletion, and the aquifer where it is allowed

The general rule is that water is not a depletable deposit. Percentage depletion under IRC §§611–613 is for mines, oil and gas wells and similar natural deposits, and water is not one of them. The exception sits directly under the aquifer this guide gives a section to. In United States v. Shurbet, 347 F.2d 103 (5th Cir. 1965), the Fifth Circuit allowed cost depletion to irrigators drawing from the Ogallala formation beneath the Southern High Plains of Texas, holding that groundwater being mined out of a formation with no meaningful recharge is an exhaustible natural deposit like any other. The IRS acquiesced in Rev. Rul. 65-296, set out the mechanics in Rev. Proc. 66-11, and has extended the treatment to other parts of the Ogallala where a taxpayer can show that the water is being depleted and that recharge is negligible.

Who can use it is the practical question, and the answer is narrower than the authority sounds. The deduction is computed from the saturated thickness under the land at acquisition and the share of the purchase price attributable to water, so it belongs to the buyer of irrigated High Plains farmland who owns the ground above a measurably shrinking aquifer and pumps it.

It does not belong to the buyer of a transferable surface right, a Colorado-Big Thompson unit or a Colorado River entitlement: none of those is a deposit being mined out from under anything. If you are buying irrigated ground over the Ogallala — and the groundwater section above says why you should think hard before you do — commission the saturated-thickness measurement and the purchase-price allocation at closing, because the deduction is measured from conditions at acquisition and the evidence has to be gathered then. It is a tax offset against a depletion you are living through, not a reason to buy. Outside those facts, there is no depletion for water.

No depreciation either, and the profile that leaves

A perpetual water right is not depreciable, because it has no determinable useful life. So for the holder outside the Ogallala exception, the asset produces little income, cannot be depreciated, and cannot be depleted: a pure capital-appreciation holding with an annual cash cost. That is a demanding profile, and it is why so much water is held inside farms, where the dirt and the crop carry the arithmetic.

Income from leasing

Rent received for leasing water is ordinary income. Assessments and management costs incurred to produce it are ordinarily deductible against it. Where the right is held in a passive structure, the passive activity loss rules apply.

Property and transfer taxes

Where a state treats the right as real property it may be assessed for property tax and may trigger transfer tax on sale. Practice varies by state and by whether the right is severed from land; verify locally before modelling.

Water has failure modes that other real assets do not, and most of them are legal or political rather than physical.

Paper water

The dominant loss in this asset is buying a decree that cannot deliver. It happens when the right is too junior to be in priority in most years, when the source has been over-appropriated, when the historical consumptive use turns out to be a fraction of the paper quantity, or when a district’s rules prevent the water leaving. There is no fraud required. The seller shows you a real decree; the decree is simply worth much less than its face quantity.

Forfeiture and abandonment

As set out above, non-use kills. An investor who buys and idles is running down the asset. Colorado’s water courts periodically publish abandonment lists; other states run administrative forfeiture. The remedy is to keep the water in beneficial use, which means owning an operating farm or a lease, which means the “pure water play” is rarely available in practice.

The public trust doctrine

California courts have held that the state retains a public trust interest in navigable waters that can require re-examination of previously granted rights. The canonical case is the Mono Lake decision, National Audubon Society v. Superior Court, 33 Cal. 3d 419 (1983), in which the California Supreme Court integrated the public trust doctrine with the appropriative rights system, holding that the state has an affirmative duty to take the trust into account when planning and allocating water and retains continuing supervision that lets it reconsider past allocations. The investment consequence is that a valid, decreed, long-exercised right is not immune to reduction on environmental grounds. That is a sovereign risk embedded in the title.

Administrative reallocation

The Record of Decision of August 21, 2026 is the live example: Lower Basin allocations cut 27.1%, 16.7% and 10% by federal decision, in two-year increments, with litigation immediately following. No amount of diligence on a fourth-priority Colorado River right would have protected against it.

Political and reputational risk

Farm-to-city transfers are unpopular in the counties losing the water. La Paz, Mohave and Yuma counties sued over Queen Creek. The Central Arizona Project’s governing agency killed a Water Asset Management sale after protest. Rural county boards have adopted export restrictions; state legislatures have taken up area-of-origin bills. An investor buying rural water to sell to a city is buying an asset whose value depends on a political permission that the affected community will fight, and coverage of the trade — “water grab,” “Wall Street is thirsty” — is uniformly hostile. If your capital cannot survive being named in a CNN story, this is not your asset class.

Tribal reserved rights

Federal reserved water rights for tribes carry priority dates from the establishment of the reservation, which is often the most senior date on the river, and many remain unquantified pending settlement. A settlement that quantifies a large senior tribal right changes the value of every junior right on the same system. The Colorado River ROD’s federally managed pool offsetting up to 25% of Lower Basin tribal reductions is one recognition of this; the underlying doctrine comes from Winters v. United States, 207 U.S. 564 (1908).

Basis and physical risk

Aquifers fall, subside and lose storage capacity permanently; wells must be deepened; conveyance capacity is finite. Australia’s Barmah Choke splits an otherwise single market into two price zones on physical throughput alone, and American systems have their equivalents. Water in the wrong place is worth what it is worth in that place.

Two examples, because the professional’s economics and the outsider’s economics are different, and both are instructive.

Example A: the Greenstone trade on documented numbers

This is the deal from the cold open. The land cost is a matter of public record rather than a figure backed out of a reported profit, which makes it the best-documented US farm-to-city water arbitrage anyone can put arithmetic to.

• Bought: 485 acres of irrigated farmland at Cibola, La Paz County, Arizona, in 2013 and 2014, for about $9,800,000, and leased back to farmers for the first four years.
• Sold: 2,033 acre-feet a year of fourth-priority Colorado River water to the Town of Queen Creek. The 2018 agreement price for the entitlement was reported at $24,000,000; the purchase the council gave final approval to on September 21, 2022 was reported at $27,000,000. The two figures are not reconciled in the public record.
• Reported profit: about $14,000,000, per KJZZ — which is $24,000,000 less the $9,800,000 land cost, so we run the arithmetic on the $24,000,000 entitlement price and treat $27,000,000 as the upper bound.
• Price per acre-foot of annual entitlement: $24,000,000 ÷ 2,033 = $11,805; on $27,000,000, $13,281.
• Gross multiple on the land cost: $24,000,000 ÷ $9,800,000 = 2.45×; on $27,000,000, 2.76×.
• Gross annualised return: over the nine years from the 2013 purchases to the 2022 approval, 10.5% a year on the lower price and 11.9% on the higher; measured to the 2018 agreement instead, 11.8% and 13.5%. All Invest Alternative arithmetic, before tax, before carrying costs the reported profit does not itemise, and before crediting the lease income of the first four years or whatever the dried-up 485 acres are still worth.

Call it 10% to 13% a year gross over roughly nine years. The flagship farmland guide puts NCREIF farmland at about 9.8% a year from 1992 to 2025, so on the documented land cost the most celebrated water arbitrage of the decade beat the boring asset — by roughly one to three points a year, depending on which of the two reported sale prices you use.

Then price what that premium bought. A decade of illiquidity. A federal environmental review, a lawsuit from three counties and a judge ordering the review replaced with a full environmental impact statement. Sustained hostile national press. And a band this wide at the end of it, because the land cost is documented, the sale price is reported two ways and the carry is not disclosed at all. Professionals with the right land and the one buyer that wanted it earned a modest premium over an index fund of dirt, and took a decade of political risk to collect it. That is the honest ceiling on this trade, not the floor.

Example B: an outsider buys ten CBT units

Now the trade an accredited retail buyer could actually attempt: a block of Colorado-Big Thompson units, the most liquid municipal water instrument in the United States. Every assumption is stated, and the ones that are ours are labelled ours.

The purchase

• 10 CBT units at $85,000 (WestWater Research’s 2025 level for CBT units, reported April 2026): $850,000.
• Broker commission at 3% (our assumption): $25,500.
• Water counsel, title on the right, Northern Water change-of-ownership processing (our assumption): $6,000.
• Total invested: $881,500.

The carry, years 1 to 10

• Annual assessment: Northern Water’s 2026 open irrigation rate of $36.07 a unit, so $360.70 a year and $3,607 over ten years if the rate never rises, which it will. At the municipal and industrial rate of $56.17 it would be $5,617.
• Lease income: at Northern Water’s 2026 quota of 80%, ten units yield 8 acre-feet; leased at $300 an acre-foot (our assumption), less a 10% agent fee: $2,160 a year, $21,600 over ten years. The board sets the quota annually and it has run far below 80% in dry years, so treat this as the good case.
• Net carry over ten years: +$17,993, taxed as ordinary income. At a 40.8% combined federal rate (37% plus 3.8% NIIT), tax of $7,341 leaves $10,652 after tax.
• On $881,500 invested, that rent is a 0.25% running yield. This asset does not pay you.

The sale, year 10

Assume the unit price compounds at 3.3% a year, taking $85,000 to about $117,500 a unit. Note what that assumes away: CBT units fell from $101,000 in 2022 to $85,000 in 2025, so a positive forward path is an assumption and not an extrapolation.

• Gross proceeds: $1,175,000.
• Broker at 3%: −$35,250. Net: $1,139,750.
• Capital gain: $1,139,750 − $881,500 = $258,250.
• Federal tax at 20% plus 3.8% NIIT = 23.8%: −$61,464. Colorado’s flat 4.40% income tax applies on top and would take a further $11,363 of the gain; we exclude it below, so the figures flatter the outcome.
• After-tax proceeds: $1,078,286.

The result

• Total after tax: $1,078,286 + $10,652 = $1,088,938.
• On $881,500 over ten years: 2.14% a year after fees and federal tax.

Test that answer against its inputs before trusting it. Run exactly the same arithmetic at $68,500 a unit and a $60 annual assessment — a plausible alternative set, well away from the published figures used above — and it returns the same 2.14% a year. The result is close to insensitive to both, because what sets it is the assumed 3.3% appreciation net of a 3% round trip, a flat carry and a 23.8% federal rate, not the price level you happen to buy at. Better inputs did not rescue this trade and worse ones would not have condemned it.

The arithmetic is the argument. A 3.3% annual appreciation in the most liquid water market in America, after a 3% round-trip commission, a decade of assessments and a 23.8% federal rate, returns about 2.1% a year, and that is before Colorado’s 4.4% takes its share of the gain. Water rights only work as an investment if the appreciation is large, and large appreciation in this asset comes from a repricing event: a change of use approved, a compact rewritten, a basin adjudicated. You are not buying a yield. You are buying a legal option, paying rent on it, and waiting for a government to act.

The order of operations matters more here than in any other real asset, because the diligence that decides the outcome happens before the offer and cannot be redone afterwards. This is the sequence.

1. Decide which exposure you want. Water-price exposure, scarcity-driven capital appreciation and water-industry equity are three different things. Only the second is genuinely available to a private US buyer, and only through land.
2. Rule out the impostors. If your intention is water, do not buy PHO, FIW or CGW; you would be buying industrials and a regulated utility. If you want those, buy them deliberately and cheaply.
3. Pick a legal regime before you pick a property. A pure prior-appropriation state with an active water court and a working transfer process — Colorado is the model — behaves nothing like a Texas rule-of-capture county or an unadjudicated California basin. Learn one regime properly rather than three superficially.
4. Take the listed step first, if you want one. Gladstone Land is the only listed US vehicle reporting meaningful water assets, about 56,000 acre-feet at its Q2 2026 reporting. Own it knowing you are buying a leveraged farmland REIT with a water option inside it, and read the sister guide on farmland REITs and funds first.
5. Hire the water lawyer before you make an offer. In this asset the diligence is the deal: a water-rights attorney in the relevant state and an engineer who can opine on historical consumptive use are the two mandatory hires.
6. Demand the three documents. The decree or permit with all six parameters, the consumptive-use engineering opinion, and eight or more years of delivery records. No exceptions.
7. Check the district’s rules, not just the state’s. Irrigation districts, conservancy districts and mutual ditch companies have bylaws that can forbid water leaving the district at all, whatever the state would permit.
8. Model the carry with zero income. Assume the lease income is nil and the assessments are due. If the position still clears your hurdle on the repricing event alone, it is a real position.
9. Write down the exit before you buy. Who is the buyer, what approval do they need, and how long did the last three comparable applications in that state take? If you cannot answer, the exit is “another speculator”, which is not an exit.
10. Size it as venture, not as real estate. Illiquid, binary on an administrative decision, decade-long, and on the best-documented professional example we have, a gross annualised outcome somewhere in the 10–13% band against about 9.8% a year for NCREIF farmland from 1992 to 2025, with a public record too thin to narrow the band further.

Specific readings and dates, each with an as-of, that would change the view. All figures are as of September 10, 2026 unless stated.

The Colorado River litigation

Nevada’s suit against Interior, filed August 24, 2026 in the District of Nevada by the State, the Colorado River Commission of Nevada and the Southern Nevada Water Authority. Watch for (a) whether an Upper Basin state or California files as well, (b) whether any court enjoins the 2027–28 operating guidelines before they take effect, and (c) whether the seven states produce a consensus agreement that supersedes the ROD before the next two-year guideline cycle. A judicial ruling that Interior lacks authority to impose the cuts would reprice every Lower Basin right; so would a seven-state deal.

The next guideline cycle

The framework issues guidelines in two-year intervals through 2036. The first renegotiation under it tests whether the federal process is durable or a placeholder for a compact fight, and it is the scheduled repricing date for Colorado River exposure.

SGMA enforcement

The Kings County Farm Bureau merits hearing is set for October 6, 2026 in the Kings County Superior Court, and the Farm Bureau has said it will petition the California Supreme Court if it loses. Watch the count of subbasins placed on probation, and watch whether the State Water Board begins collecting pumping fees anywhere. A basin that reaches actual fee collection is the first real-world price signal SGMA has produced.

The NQH2O index level

The index prints once a week, after the close on Wednesday. Reference points: the record $1,144.14 on June 27, 2022, and $566.37 on our tape on September 7, 2026. A sustained move back above $1,000 would mean a drought as severe as 2021–22; a sustained move below $300 would mean a wet cycle that removes the near-term bid under California water. Neither is a valuation signal for permanent rights, but both change the lease economics that pay the carry.

CME NQH2O futures open interest

The contract has never held durable liquidity: the academic work on its first months found daily volume in the low tens of contracts, and we could not obtain a current volume or open-interest figure from CME’s daily bulletin. A persistent four-figure open interest, sustained over a full water year, would be the first evidence that financial water exposure has become possible at all. Until then, the answer to “how do I trade water” is that you cannot.

Colorado-Big Thompson unit prices

WestWater Research’s $85,000 an acre-foot for 2025, against the $101,000 high of 2022, is the benchmark for permanent municipal water on the Front Range, and it fell by about a sixth between 2022 and 2025 — through the same drought that took the California lease index to its record, which is the single most instructive price pair in this guide. A return to the 2022 high on rising volume would say Western municipal demand is outgrowing supply again; another leg down would say Front Range demand, not hydrology, sets this price.

Arizona’s rural groundwater rules

Whether the legislature creates a statutory transfer path — or a pumping limit — outside the five Active Management Areas. The 2025 Ag-to-Urban law, SB1611, opened a path inside the Phoenix and Pinal AMAs; the 2026 McMullen and Butler valley transfer bills, HB2757 and HB2758, died after passing the House, and a local campaign has been gathering signatures since August 2026 to create a groundwater protection zone over McMullen Valley instead. Rural Arizona is where the private capital went precisely because there is no limit; a rule in either direction reprices those holdings at once. Separately, whether the April 22, 2026 ruling against the Phoenix AMA unmet-demand rule is appealed, and what replaces it.

Kansas Geological Survey January measurements

Published annually; the regional average has been −0.42 feet a year since 1996, and the January 2026 reading was the first overall increase since 2019, at +0.18 feet across western Kansas on a wet year. Whether that holds is the question: a return to declines of a foot and a half in southwest Kansas would resume the countdown to the terminal date for irrigated agriculture in the southern High Plains, and therefore for the value of the land above it.