Colorado Water Rights and the Water Value Chain

00Executive summary

Colorado's water scarcity is real, worsening, and almost entirely non-monetizable by private capital: the legal architecture converts the scarcity rent into a public subsidy for growth, and the only nodes that reliably keep cash are the ones that get paid whether or not the water ever moves.

The consensus premise is sound. The Colorado River is over-allocated against a shrinking flow record; 2026 is the worst hydrologic year in the modern instrument record; the Front Range — the urban corridor running along the eastern foot of the Rockies from Fort Collins through Denver to Colorado Springs — keeps adding rooftops. From those facts the market draws a conclusion that feels inevitable — a fixed asset in rising demand must appreciate, and appreciation must eventually become a return. The first half of that is roughly true. The second half is where the money is lost.

Northern Water publishes the price series for Colorado-Big Thompson units (tradable shares in one federal reservoir project's annual supply, not acre-feet of water), the state's de facto benchmark. It peaked at $67,000 per unit in 2023 and stood at $57,487 in 2025, a 14.2% nominal decline. Adjusted to 2025 dollars, the 2025 price is below its 2020 level, and a unit bought at the end of 2019 and held through 2025 lost 4.2% of its purchasing power while paying no cash yield at all. This happened during the drought, not in spite of it. Scarcity and price moved in opposite directions for three consecutive years because the buyer of last resort is a cost-of-service utility — one legally obliged to charge its customers what supply costs, with no profit margin on top — whose demand is set by building permits, not by hydrology.

That single chart carries most of the argument. What follows tests it node by node, and the pattern repeats: the scarcity value is real, but Colorado law systematically routes it away from anyone who could book it as profit. The state owns the water. The right holder owns a decreed priority to use it, a usufructuary interest — a right to use water the state itself owns, not ownership of the water — that cannot be leased freely, cannot be moved without a multi-year adjudication that can cut the quantity in half, and whose ultimate buyer is a public enterprise legally required to charge cost. A senior water right is a magnificent asset and a poor business.

Evidence: Reclamation July 2026 24-Month Study; Northern Water C-BT market-price and quota series; GWRS FY2025 Form 10-K; PCYO FY2025 Form 10-K. Method: ROIC = operating income × (1 − effective tax rate) ÷ (total equity + total debt − cash), all at fiscal year end. Full derivation in §14.

Three conclusions run against the consensus, and each is falsifiable on dated public evidence.

1. 1The Colorado water market is a demand-side market wearing a supply-side costume. Consensus reads C-BT prices as a scarcity gauge. The 2023–2025 decline coincided with record drought and falling residential permits, which is what you would expect if the marginal buyer is a homebuilder's tap-fee obligation — the one-off charge a builder pays a utility to connect a new house to the system — rather than a farmer's thirst. Price follows rooftops, not runoff.
2. 2A compact call would compress the value of the seniority stack, not expand it. A call is the demand from downstream states that Colorado stop depleting the river; the seniority stack is the queue, ordered by the date each right was first established, in which rights get shut off when that demand lands. Consensus treats curtailment risk as bullish for senior pre-1922 rights. It is bullish for their relative position and bearish for the transferable market, because the transfers that create liquidity are municipal, and the municipal buyer's alternative — reuse, cleaning and re-serving the water a city has already delivered once — gets cheaper the moment its raw supply gets riskier.
3. 3The one listed company now assembling Colorado water has turned $72 million into a $70.7 million carrying value and no income. Consensus assumes capital simply has not tried. It has. Select Water Solutions committed $72m in February 2025 for 39% of AV Farms, LP, holding 16,300 acre-feet of senior Arkansas Valley rights; thirteen months later the equity-method carrying value — what the stake is worth on Select's own balance sheet, moved up or down by its share of the partnership's profits and losses — was $70.7m, with roughly $74m of further contributions planned. Before it, PICO became Vidler and was bought by a homebuilder in 2022; Two Rivers Water & Farming lost its SEC registration the same year. The pattern is not absence of effort.

How to read the grades. Twenty-seven listed names were examined, each against its own most recent filing opened this session. Grades are structural — advantaged, neutral, exposed — and rest on return-on-capital economics — what a business earns each year on every dollar of capital tied up in it — not margin and not share price. Where a name has no Colorado exposure at all, the grade says so and reflects the layer it actually occupies. The mandate anticipated the finding and the evidence confirms it: on the Colorado water chain specifically, no listed security offers direct, controlled exposure to the appreciating asset. Two come close, from opposite directions — Pure Cycle, which owns the water and monetises it as building lots, and Select Water Solutions, which since February 2025 has held a 39% non-controlling stake in a private partnership assembling 16,300 acre-feet of senior Arkansas Valley rights. Both are examined in full, and the second is this report's own falsification test.

Part IDomain Primer — how Colorado water actually works

Everything in Part II depends on legal mechanics that have no analogue in most asset classes. This section builds them from first principles so that every later claim can be checked independently. A reader who already knows the doctrine can skip to §9; a reader who does not will find the investment conclusions unfalsifiable without it.

01The physical resource: what is in the river versus what is on paper

Colorado is a headwaters state. Essentially every drop it uses falls as snow inside its own borders, and essentially every drop it does not use leaves for someone else. That geography is why the state has more interstate compact obligations than any other, and why its water accounting is unusually brutal: there is no upstream neighbour to blame.

Water yearThe twelve-month accounting year hydrologists use instead of the calendar year. It begins in the autumn, after the previous summer's melt is finished and before the new snow falls, so that one winter's snowpack and the runoff it produces fall inside the same year. 2026 is the stress test. The Bureau of Reclamation's July 2026 24-Month StudyReclamation's monthly projection of reservoir levels and releases across the Colorado River system for the next two years, under a most-probable, a wet and a dry inflow scenario. It is the closest thing the basin has to an official forecast, and every operating decision on Lake Powell and Lake Mead is argued against it. puts the water-year unregulated inflowThe water that would arrive at a reservoir if no dams upstream were holding any of it back — the natural runoff, stripped of human operation. Forecasting it separately is how Reclamation distinguishes what the weather delivered from what the plumbing did. forecast for Lake Powell at 3.50 million acre-feet, 36% of the 1991–2020 average; observed June inflow was 0.306 maf, 12% of average; the July forecast was 0.020 maf, 2% of average (Reclamation, July 2026 24-Month Study, 15 July 2026). Inside Colorado, the state climatologist's April 1 assessment put available snow waterShort for snow water equivalent: how much water you would get if you melted the snowpack where it lies. It is the measure that matters, because a deep, dry snowpack and a shallow, wet one can hold the same water. Colorado reads it on 1 April because that is roughly when the pack peaks and the melt begins. at 22% of the historic median for the date — the worst in the recorded history of the current observing system, after a March in which the temperature reached 99°F at Burlington on the Kansas border (Colorado Sun, 3 April 2026, reporting Colorado Climate Center data).

The decline is in the flow, and decrees are written against flow, so a decree's nominal quantity has been drifting away from its physical delivery for two decades. The state's storage buffers absorbed the shock unevenly: Reclamation's most probable projection has Lake Powell falling to 3,491.75 feet in April 2027, within two feet of the 3,490-foot minimum power poolThe reservoir level below which the water no longer sits high enough above the dam's turbines to generate electricity. It is not the bottom of the reservoir — water remains, and can still be released — but below it the dam stops being a power plant, which is why the level functions as a political and financial red line rather than a physical one. at Glen Canyon Dam, after Interior cut the water-year release from 7.48 to 6.00 million acre-feet and scheduled up to 1.00 maf of emergency releases from Flaming Gorge (Reclamation news release, 17 April 2026).

Hold that number. It is the hinge on which every Colorado seniority argument in Part II turns.

02Prior appropriation, the priority stack, and what seniority does not buy

Seniority governs the order of curtailmentBeing ordered to stop diverting. When a river runs short, the state does not ration everyone a little; it shuts users off one at a time, newest right first, until the older right that complained is satisfied. Curtailment is the enforcement mechanism that gives a priority date its whole value., not the volume in the river. A first-in-line claim on an empty ditch is worth exactly nothing.

Colorado follows pure prior appropriation: first in time, first in right. A water right is created by diverting water and applying it to a beneficial useAn actual, recognised productive use of water — irrigating a crop, supplying a household, running a mill — as opposed to holding water idle or for resale. Colorado grants rights only for beneficial use, and that single requirement is why a right cannot simply be bought and warehoused: use it or lose it., and it is perfected by a decree that fixes an appropriation date, a source, a point of diversion, an amount, and a use. When supply is short, the Division EngineerThe state official who physically administers one of Colorado's seven river basins — reading the gauges, deciding whose headgate closes, and enforcing the priority order day by day. The State Engineer sits above the seven of them. Neither is a judge and neither sets prices; they only police the queue. administers the river by shutting off junior rights in reverse order of seniority until the senior calling right is satisfied. There is no proration, no sharing, no equitable haircut. The stack is absolute.

Colorado is also the only state that adjudicates water through dedicated water courts. Seven water divisions, one per major basin, each with a designated water judge and a water refereeA specialist officer of the water court who handles a case before any judge does — reviewing the engineering, pushing the parties toward a settlement, and recommending a decree. The referee stage, not the hearing, is where the engineering is fought over and where the years and the fees accumulate., were established in 1969 (Water Education Colorado, Water Courts). Every claim, every change, every augmentation planA court-approved swap. It lets a user divert water they have no priority to take, on condition that they put replacement water back into the stream at the time and place the shortfall would otherwise hurt someone downstream. It is how Front Range subdivisions on well water are legally supplied — and the replacement water has to be bought from somebody. runs through them. The transferability of the asset is controlled by a court, not by a market, and the court's mandate is to protect other people's rights, not to facilitate the seller's exit.

Seniority is genuinely valuable, and it is routinely misread. A pre-1922 right on the Colorado mainstem, or an 1860s Arkansas ditch right, will be among the last curtailed. But its value is insurance, not yield. The insurance pays only in the states of the world where the river is short, and in exactly those states the physical water available to deliver against the decree is at its smallest. A senior right is a claim that becomes most valuable precisely when there is least of the thing it is a claim on. That is not a contradiction; it is the definition of a hedge, and hedges do not compound.

03Consumptive use, return flows, and the no-injury rule

An irrigator who diverts three acre-feet per acre does not consume three acre-feet. Consumptive use is the fraction that never comes back — evaporated, breathed out by the crop, carried off in the harvest — and it is far smaller than the amount diverted. That distinction is the hinge of the whole subject, because the consumed fraction is the only part a seller is permitted to move. Crops transpire part of it; the rest seeps back to the stream as return flow, on a lag, and downstream appropriatorsAnyone holding a water right on the stream. The word carries the doctrine with it: in Colorado a right is appropriated — claimed by putting water to use and dated from that moment — rather than granted by the state or attached to riverfront land. have built their own decrees around that return flow for a century. Colorado law therefore protects the pattern of historical use, not just its quantity.

When a right changes — from irrigation to municipal supply, or from one point of diversion to another — the court quantifies historical consumptive use from actual records: decades of diversion records, crop histories, climate data, and ditch-company delivery data, analysed by a water resources engineer and defended under cross-examination. The transferable quantity is the historical consumptive use, not the decreed amount. Everything else must continue to return to the stream on the historical schedule, or the change is denied.

Two consequences follow, and both are financial. The first is the quantification haircut: the decreed paper amount routinely exceeds the transferable consumptive use by a wide margin, because courts quantify on actual historical beneficial use rather than on the decree's face (Welborn Sullivan Meck & Tooley, on the Colorado Supreme Court's historical-use burden). A buyer paying for decreed acre-feet and receiving transferable consumptive-use acre-feet is paying for an asset it will not get. The second is that the transferable quantity is unknown until the case is decided. An investor buying agricultural water rights is buying an option on an engineering finding, at a strike price set years before the finding is made.

This is where the per-acre-foot headline price loses its meaning. Three transactions in 2025 and 2026 illustrate the spread, and they are not comparable to one another in any sense that would let a spreadsheet average them.

Evidence: Northern Water, C-BT Units Market Price Historical Data; Aspen Journalism, 29 May 2020 (updated 29 Nov 2021); Limoneira FY2025 Form 10-K, Recent Developments. Method: C-BT $/af of firm yield = $57,487 ÷ 0.762, the mean final quota 2006–2025 from Northern Water's published quota declarations. Grand Valley $/af = $16.6m ÷ 6,636 af, the class-1 duty estimate reported in the cited article. Synthesis: Three headline "price per acre-foot" figures spanning 30× are not a market; they are three different legal instruments priced in the same unit of measure.

04The change-of-use and augmentation chokepoint

An agricultural water right is worth its municipal value only after a water court says so. Getting that decree is the chokepoint, and it is the structural fact the scarcity narrative ignores entirely.

1. 01
   Application

   Filed in the division water court. The filing fee is trivial; the engineering behind it is not.
2. 02
   Resume notice

   Every application is published monthly in the division's resumeThe water court's monthly public bulletin listing every new application filed in that division. It is the formal notice to the rest of the river that somebody wants to change something — and the clock on which the deadline to object starts running., putting every other user on notice.
3. 03
   Statements of opposition

   Any water user with standingThe legal right to be heard in a case because you could be harmed by its outcome. In Colorado water, standing is unusually broad: essentially every other right holder on the stream qualifies, which is why a single transfer can attract dozens of opponents who need prove no wrongdoing — only exposure. may oppose. Contested municipal cases routinely draw dozens of opposers.
4. 04
   Referee & engineering

   Historical consumptive use is quantified and return-flow obligations are set. This is where the haircut happens.
5. 05
   Decree with terms

   The transferable amount, delivery schedule, and permanent return-flow obligations are fixed for the life of the right.

Evidence: Water Education Colorado, Water Courts; Somach Simmons & Dunn, Understanding the Colorado Water Court Process. Method: Sequence reproduced from the cited procedural descriptions; no timing figures are asserted where the sources do not supply them. Synthesis: Steps 3 and 4 are where transferability is priced — they are the reason an ordinary market clearing mechanism does not exist for Colorado surface water.

The procedural detail matters because it is where the delay lives. The Division Engineer files a Summary of Consultation analysing the application — unless the Division Engineer is itself an opposer, in which case no summary issues. Any water user has two months from the first day of the month following filing to enter a statement of opposition, with the deadline printed in that division's monthly resume. The case then proceeds before a water referee whose job is to move the parties toward settlement before a judge ever hears it (Somach Simmons & Dunn, Understanding the Colorado Water Court Process).

Cost and duration are the two numbers everyone wants and few sources give cleanly. A contested Colorado water case is expensive enough that the Colorado Supreme Court has treated seven-figure expenditure as ordinary evidence of diligence: in one Roaring Fork valley matter the record showed roughly $500,000 spent within a single diligence periodA right claimed but not yet built out — a conditional right — survives only if its holder returns to court at intervals set by statute and shows real, documented progress toward putting the water to use. Each such window is a diligence period. Fail one and the right is cancelled, which is why holders spend heavily inside them. and over $1 million in total across a water law firm, an engineering firm, a second law firm, a land planner, and a consultancy. Whatever the applicant's outcome, that money is earned by counsel and engineers on the way in.

An augmentation plan is the parallel instrument. It permits out-of-priority diversions provided the diverter replaces the depletions with a substitute supply, on the schedule and at the location the injury would otherwise occur. Front Range growth on Denver BasinThe deep sequence of aquifers under Denver and the plains south and east of it. Because it is groundwater rather than river water, Colorado allocates it by a different rule — a statutory share of the stored volume, spread over an assumed hundred-year life — instead of by priority date. A holder therefore knows the quantity in advance rather than waiting on a court, which is why this is the corner of Colorado water where private ownership actually turns up. groundwater is built almost entirely on augmentation plans, and the replacement supply they require is itself a purchased water right. Augmentation demand is therefore a derived demand for the same senior agricultural water everyone else is bidding for.

One 2025 ruling narrowed the doctrine in a way that matters commercially. In Franktown Citizens Coalition II v. Independence Water & Sanitation District, the Colorado Supreme Court held that the anti-speculation doctrineThe rule that you may not acquire a Colorado water right in order to resell it. An applicant must show a specific, genuine plan to put the water to use, and usually a contract with an end user who needs it. This is the single doctrine that most directly forbids the buy-cheap-and-wait strategy an outside investor would naturally reach for. does not apply to an application to obtain or amend an augmentation plan for not-nontributaryColorado sorts groundwater by how strongly it is connected to a river. Tributary water is treated as part of the river and administered by priority date. Nontributary water is effectively disconnected and administered as a stored asset. Not-nontributary is the awkward middle: connected enough that pumping it must be partly replaced, but not administered by priority. Much of the Denver Basin falls in this category. Denver Basin groundwater, because the only question in such a plan is injury, not intent (Best Lawyers, 13 March 2026). The door that anti-speculation closes on surface-water appropriations stands meaningfully wider on the Denver Basin groundwater side. That is where the only listed company that owns Colorado water outright happens to sit, and §15 returns to it.

05Compacts, transbasin diversion, and what a call would mechanically do

The 1922 Colorado River Compact divided the river between an Upper Basin and a Lower Basin, with Colorado holding the largest Upper Basin apportionment. The operative obligation, in Article III(d), constrains the Upper Basin from depleting the flow at Lee FerryThe single point on the Colorado River where the 1922 Compact measures what the upstream states have passed downstream. Nothing about the place matters except that it is the meter: every argument about whether Colorado has met its obligation is an argument about the number recorded there. below an aggregate of 75 million acre-feet in any ten consecutive years. Whether that is a firm delivery guarantee or merely a non-depletion covenant is the entire legal fight, and in 2026 it stopped being academic.

The 2007 Interim Guidelines, the 2019 Drought Contingency Plans, and the relevant provisions of Minute 323A numbered amendment to the water-sharing treaty between the United States and Mexico, setting how the two countries share Colorado River shortages and surpluses and how Mexico may store water in US reservoirs. The name is bureaucratic — the treaty's implementing commission issues its decisions as consecutively numbered "minutes" — and it matters here only because it expires alongside the domestic rules. with Mexico all expire on 31 December 2026 (Reclamation, Colorado River Post-2026 Operations). Reclamation issued a Draft Environmental Impact Statement on 9 January 2026, took comment from 16 January to 2 March 2026, received more than 18,000 submissions, and — critically — identified no preferred alternative. Interior said it would determine post-2026 operations by summer 2026 if the states failed to agree. As of this report's evidence cutoff, no Record of DecisionThe federal document that ends an environmental review by declaring which alternative the agency has chosen. Until it is signed, nothing is settled and nothing is legally challengeable; once signed, the chosen operating rules bind. Its absence here is the point — the basin is inside the gap between an expiring framework and a replacement nobody has committed to. had been issued.

Counsel following the file are unusually direct about the litigation risk: Arizona has signalled readiness to sue as soon as it believes the Compact is breached, with the Supreme Court's original jurisdictionThe Supreme Court's power to hear a case first rather than on appeal, which the Constitution reserves for disputes between states. It means a compact fight would begin at the top, with no lower court to narrow the issues first — a proceeding measured in years and argued by every affected water user at once. the likely venue and the Article III(d) interpretation the central question (Kilpatrick Townsend, 23 April 2026).

Evidence: Reclamation post-2026 programme page; Reclamation news release 5326 (17 Apr 2026); Reclamation July 2026 24-Month Study. Method: Dates and volumes reproduced verbatim from the cited federal documents. Synthesis: The interval between the guidelines' expiry and any replacement is the period of maximum administrative and legal uncertainty for every Colorado right junior to 1922.

What would a call actually do inside Colorado? The State Engineer would have to curtail post-1922 Colorado River water rights to make up the shortfall at Lee Ferry. The immediate exposure is not the old Western SlopeEverything in Colorado west of the Continental Divide — where the snow falls, the rivers run toward the Pacific, and comparatively few people live. Its counterpart is the Front Range on the east. Almost every argument in Colorado water is, underneath, an argument about moving water from one side of that line to the other. irrigation ditches — many predate 1922 — but the transmountain diversions that supply the Front Range. Denver Water's Moffat and Roberts tunnel collection systems, the Colorado-Big Thompson project, Homestake, and the Fryingpan-Arkansas project all move Colorado River water east under the Continental Divide, and their key rights are mostly post-1922. The rooftops are on the dry side of the mountains; the water is on the wet side; the pipes are the whole business.

Two further interstate exposures sit outside the Colorado River file and are easy to overlook. The 1923 South Platte River Compact is live again through Nebraska's Perkins County Canal project, with 2026 a discovery year in the resulting dispute and endangered-species questions layered on top — and the South Platte is the basin that supplies the Front Range and hosts the C-BT market. Separately, the Southern Ute and Ute Mountain Ute tribes hold among the most senior rights in the state and have pressed for infrastructure and for the right to market tribal water through leasing, a 2026 focus. Senior tribal water entering a leasing market would be the largest new supply of transferable senior water in Colorado's history, and it would arrive from a holder with no obligation to sell cheaply (Best Lawyers, 13 March 2026).

06The C-BT anomaly: why one project became a benchmark and nothing else did

Roughly nothing about the C-BT market generalises to Colorado water, which is exactly why it became the benchmark.

The Colorado-Big Thompson Project is a federal transbasin project completed in the mid-twentieth century and operated by Northern Water. It divides its annual supply into 310,000 units, each unit being one 310,000th of the water the board makes available in a given year (Northern Water, C-BT Quota). Each November the board sets an initial quota; each April it usually raises it. Because it is a stored, firmed supply administered by a district board rather than a river right administered by a Division Engineer, a C-BT transfer requires board approval, not a water court decree. There is no change case. There is no historical consumptive use analysis. There is no quantification haircut. Transfer takes weeks.

That is the entire source of its benchmark status. It is the only meaningful Colorado water asset that is liquid, and it is liquid because a federal storage project and a district board between them stripped out the two frictions that define every other Colorado right.

The quota mechanism also demolishes the "seniority equals supply" intuition. In 2026 — a year with a statewide April 1 snow water equivalent under one third of the previous all-time low — Northern Water set an initial quota of 50% in November 2025 and raised it to 80% in April 2026. That is the same final quota as 2024 and higher than 2023, 2021, 2019, 2016, 2015, 2014 and 2013 (Northern Water, C-BT Quota Declarations, 1957–2026).

Storage buffered the drought. That is what storage is for, and it is why the price of a C-BT unit is not a scarcity gauge in any usable sense — it is the price of an unusually liquid claim on an unusually well-buffered supply, quoted in a unit that is not an acre-foot. A unit priced at $57,487 delivers 0.80 acre-feet in 2026 and averaged 0.762 acre-feet a year over 2006–2025. Reporting that as "$57,487 per acre-foot," as Northern Water's own series header does, understates the implied cost of firm yield by roughly a third. Reporting it as "$85,000 per acre-foot," as third-party market trackers do, is closer to a firm-yield measure but is not the price of the traded instrument. Both numbers are in circulation. They describe different things.

07Mutual ditch companies, shares, and the governance friction

Outside the C-BT service area, the practical unit of trade is not the decree. It is a share in a mutual ditch company — a nonprofit corporation owned by the irrigators it serves, holding the decreed rights in its own name and delivering water to shareholders pro rata.

The Grand Valley Irrigation Company, on the Colorado mainstem west of Grand Junction, is a clean example: 48,000 shares among roughly 3,000 shareholders, each share entitling the holder to 4.6 gallons per minute of direct flow, with a 2026 assessmentThe annual bill a mutual ditch company sends each shareholder to cover the cost of running the ditch — repairs, the ditch rider's wages, administration. It is a cost recovery, not a price for water, and because the company is nonprofit by charter the assessment is all a shareholder pays and all the company collects. of $376.90 on an account's first share and $18.03 on each additional share (Grand Valley Irrigation Company, Water Rights and Annual Assessment). Run the arithmetic and the carry runs the opposite way from what most investors expect: 4.6 gpm over a 180-day season is about 1.19 million gallons, or 3.66 acre-feet of diversion per share, so the marginal assessment is roughly $4.93 per acre-foot diverted per year. Holding Colorado agricultural water is cheap.

That is the trap. Cheap carry on an asset with no cash yield is the reason people hold it for years without noticing that the compounding is happening to somebody else. The binding cost of ownership is the forgone return on the capital, and §16 quantifies it.

The governance friction is separate and severe. A mutual company's bylaws typically restrict transfer, and its board — elected by irrigators whose farms depend on the ditch continuing to run — has both the incentive and the practical ability to slow an outside buyer's change case. A shareholder who wants to move water off the ditch must satisfy not only the water court's no-injury standard but the company's own rules and the tolerance of neighbours who will still be there afterwards. Buyers call this friction. It is the mechanism by which the community keeps the option value of its own water.

08Buy-and-dry, Crowley County, and the alternatives built in reaction

Crowley County, in the Lower Arkansas Valley, sold most of its irrigation water to Front Range cities. That is buy-and-dry: a city buys the farm for the water attached to it, moves the water permanently, and the land goes out of production for good. The transaction is legal and voluntary, and — as Crowley shows — close to impossible for the community left behind to undo. Its irrigated acreage fell from more than 50,000 acres in the 1970s to a few thousand acres by 2024, and Colorado Springs Utilities has since publicly apologised for the "travesty" and changed how it treats land where it has bought water rights (Colorado Sun, 16 March 2026). It is the state's canonical cautionary tale, and the reason its politics now run against permanent transfer.

Statewide, the trend is worse than the local story suggests. The USDA Census of Agriculture recorded a loss of 474,000 irrigated acres in Colorado — about 17% of the statewide total — between 2017 and 2022 (CWCB, Update Report on Collaborative Water Sharing Agreements, March 2025). The dry-up is not slowing.

Two policy responses followed, and both cut directly into the returns of anyone buying agricultural water for later resale.

The first is the alternative transfer method, since renamed the collaborative water sharing agreement: lease-fallowingPaying a farmer to leave part of the land unplanted for a season and lending the water it would have used to a city. The farm, the ditch and the decree all stay intact, and the arrangement repeats — which makes it a rental of water rather than a sale of it., interruptible supply agreementsA standing contract letting a city call for a farmer's water in a limited number of years out of a set span — typically the dry ones — in exchange for a payment. The farmer keeps the right; the city buys an option on it, exercisable only when it is actually short., and option contracts that move water to a city in dry years while leaving the farm and the ditch intact. The Colorado Water Plan set a target of 50,000 acre-feet shared through such agreements by 2030. The state's own March 2025 status report counts 28 projects with 66,100 acre-feet of contracted volume, of which 21 are active with about 30,300 acre-feet — real progress, against a target the report's own metrics table shows as unmet. It also records two things an investor should read carefully: sustained interest in the Lower Arkansas, and an explicit "lack of interest in developing CWSAs in the Northern Front Range region."

That asymmetry is not an accident. The Northern Front Range already has a liquid market in C-BT units. Where a frictionless transfer instrument exists, nobody builds a workaround. Where it does not, the workaround becomes policy. The same report notes that despite several administrative approval mechanisms created between 2002 and 2020 specifically to route around water court, "it does not appear that use of these administrative tools is growing in Colorado."

The second response is newer and sharper. House Bill 26-1340, signed 1 June 2026, requires that any water right in Water Division 2 (the Arkansas basin) changed from agricultural irrigation to another beneficial use on or after 1 January 2027 be accompanied by revegetation or conversion to dryland farmingFarming on rainfall alone, with no irrigation. It is a real land use and a far less productive one, which is why converting irrigated ground to it counts as a loss rather than a substitution. with erosion control and weed management. The water court must appoint a third-party expert to conduct annual field reviews until establishment succeeds, and must order the right owner to provide financial assuranceMoney posted up front — a bond, a letter of credit, an escrow — that the public authority can draw on if the owner fails to do the promised work. For a buyer it converts a future obligation into capital tied up at closing, before a single acre-foot has been delivered. to the local land use authority covering the anticipated cost — or place limits on the timing or percentage of water usable during the process (HB26-1340, Colorado General Assembly, enacted 2026).

Price that. A buy-and-dry acquisition in the Arkansas basin now carries an unbounded post-closing reclamation obligation, a bonding requirement, a court-appointed monitor, and the possibility that the court withholds part of the water until the ground is green. The bill's sponsors were rural legislators from both parties. Aurora Water's stated concern was that it might spread to other basins.

Evidence: CWCB Collaborative Water Sharing Agreements update, March 2025; HB26-1340. Method: Both columns describe the same acre-foot of agricultural consumptive use under two contracting structures; no cash-flow forecast is asserted beyond what the cited sources disclose. Synthesis: Colorado is deliberately raising the cost of the structure an outside investor needs and subsidising the structure that keeps the water with its community.

Part IIInvestment Brief — where the profit actually lands

09Conclusion first

Durable profit in Colorado water concentrates in exactly two places: the licensed professional layer that gets paid to move water through the Water Court, and the equipment and services layer that gets paid whenever anyone builds anything — and neither of them owns a drop.

The scarcity rent is real. Rent, in the economist's sense, is the extra a holder can charge purely because the thing is scarce and they have it — not payment for work done, service rendered, or capital put at risk. Water in a drying state throws off a great deal of it. It is simply not collectible by the owner of the asset, for reasons that are legal rather than economic and therefore do not respond to price. Colorado's water is owned by the state; the decree conveys a right to use it; the anti-speculation doctrine prohibits acquiring it without a genuine intent to apply it to beneficial use; the no-injury rule makes every transfer an adversarial engineering case; and the buyer at the end of the chain is a municipal enterprise or conservancy districtA special-purpose unit of local government created to build and operate water works for a region — Northern Water and the Southeastern Colorado Water Conservancy District are the big ones here. It has an elected or appointed board, taxing and bonding power, and a statutory duty to serve its constituents rather than to earn a profit. Districts own most of Colorado's storage and conveyance, which is why that layer has no private route in. that must, as a matter of law and of politics, charge its ratepayers cost. There is no node at which a private holder can insert a toll on flow.

What there is, instead, is a set of nodes where cash accrues reliably because they are paid for the transaction rather than for the asset:

Almost every layer in Colorado water is defended, and that is precisely the problem. Denver Water's service area is a legal monopoly; the Rangeview Metropolitan DistrictA Colorado quasi-governmental district formed to finance and provide services — water, sewer, roads — inside a single new development, typically created by the developer and repaid out of property taxes and connection fees on the houses that follow. It is the standard vehicle for Front Range greenfield growth, and it can grant one provider the exclusive right to serve everyone inside its boundary.'s is exclusive; a mutual ditch company's shares are the only way to that ditch's water. Colorado is full of unassailable positions that are legally forbidden to earn an excess return. Defensibility without the right to price is a public utility, and a public utility is a bond with operational risk.

The consensus collapses three different things into one, and the rest of this brief keeps them apart:

Evidence: Northern Water C-BT price series with BLS CPI-U annual averages; GWRS FY2025 Form 10-K; TPL FY2025 Form 10-K, Water Services and Operations segment. Method: C-BT real return = ($57,487 ÷ $47,662) ÷ (321.943 ÷ 255.657) − 1 = −4.2%. GWRS earned ROE = $2.957m ÷ $86.616m. Synthesis: The three rows are not points on a spectrum; they are structurally different instruments, and Colorado law permits only the first two.

10Sizing the pool — and why the size is not the story

The market is large and growing. That is context, not conclusion.

The demand gap. The 2023 Colorado Water Plan projects a municipal and industrial supply shortfall of 230,000 to 740,000 acre-feet per year by 2050 in dry years. The range spans a factor of 3.2 between the low and high case — not a forecast so much as an admission that the answer depends on choices nobody has made yet.

The standing stock. Colorado had roughly 2.8 million irrigated acres in 2017 and lost 474,000 of them by 2022. Value that remaining stock at the Grand Valley transaction basis and you get one number; value it at the C-BT market and you get a number an order of magnitude larger; value it at what a municipality would actually pay after a change case and quantification haircut and you get a third. Any of the three can be defended and none of them is a market capitalisation, because the overwhelming majority of that stock cannot be sold at any of those prices — it is held by irrigators who will not sell, in ditches whose boards will not permit transfer, on rivers from which the water cannot physically reach a buyer.

The capital plans, which are the real money. The addressable spending sits here, and it is public.

Evidence: Denver Water 2025 Approved Budget, Sources and Uses (2024 actuals); Colorado Sun, 27 Aug 2025 and Northern Water NISP project page; Colorado Sun, 19 January 2023. Method: Implied $/af = capital cost ÷ annual yield added. Comparability limit — read before using these numbers: NISP is a greenfield reservoir with a permanent decreed supply; the Prairie Waters figure is a brownfield debottlenecking of a plant that already exists, so it captures marginal capital only and excludes the original build and the materially higher operating cost of reuse. The two are not like-for-like and are shown together only to establish the direction and rough order of the gap. Synthesis: Even allowing a full order of magnitude for that mismatch, expanding reuse is the cheapest acre-foot a Front Range utility can buy.

Side by side, those numbers are the mechanism behind the price series in the executive summary.

Firm yield is the amount a supply can be counted on to deliver in a bad year — not what it holds, and not what its paperwork says. Glade's 170,000 acre-feet of capacity produce 40,000 acre-feet a year of it; a C-BT unit delivers whatever fraction of an acre-foot the board declares that spring. Put every route on that one basis and the prices in circulation stop agreeing with each other. A C-BT unit priced per acre-foot of the firm yield it actually delivers costs $75,393, which is 11.7% above the implied capital cost of building brand-new storage at NISP. And the Grand Valley agricultural basis sits at $2,502 an acre-foot — a thirtieth of the C-BT figure — for water that is physically and legally unable to reach the buyers who set the C-BT price. Neither of those is a market anomaly. They are the price of transferability, quoted twice.

Aurora Water's own position, in the reporting on its reuse programme, is that "the water it already owns is cheaper than any other option, as prices per acre-foot soar into the thousands." The utility is right, and its arithmetic is the ceiling on every water right in the state. Colorado reinforced it in November 2022 by becoming the first state in the country to adopt a formal direct potable reuseTreating wastewater to drinking standard and putting it straight back into the supply, without the customary detour through a river or an aquifer first. The distinction is administrative rather than technical: the water is treated to the same standard either way, but skipping the detour requires a regulator willing to write the rule for it. regulation, incorporated into Regulation 11 — a rule that, on the department's own account at adoption, formalised a pathway no Colorado utility had yet implemented (CDPHE, January 2023).

11The profit pool, layer by layer

Ten layers. Each gets a verdict on excess return, on defensibility, and on whether any listed vehicle offers exposure. Where the answer is "no durable profit here," the section says so.

1. 01
   Hydrology & the physical resource

   Nobody owns it and nobody can bill for it. The resource is declining, which raises the value of storage and of seniority and lowers the value of any right whose delivery depends on unregulated flow.

   No profit pool
2. 02
   The decreed right & priority stack

   The appreciating asset. Zero cash yield, negligible carry, punitive illiquidity, and a real return of −4.2% over 2019–2025. Its value is insurance against curtailment, which pays in exactly the scenarios where delivery is smallest.

   Appreciation without return
3. 03
   Compact & federal administration

   Reclamation, the Upper Colorado River Commission, and seven state negotiators set the boundary conditions for everything below. No private participant is paid at this layer; several are exposed to it.

   Pure risk transfer
4. 04
   Storage & transbasin conveyance

   The genuinely irreplaceable asset — you cannot rebuild a tunnel under the Continental Divide. Also entirely owned by districts and municipal enterprises operating at cost of service. The Gross Reservoir litigation shows that even a public owner with a decree, a permit, and a topped-out dam cannot count on filling it.

   Irreplaceable, unprofitable
5. 05
   Water Court change-of-use & augmentation

   The chokepoint, and the one place with a genuine licensed toll. Seven-figure professional fees per contested municipal case, paid by the applicant, earned by water lawyers and consulting engineers regardless of outcome. Capital requirement: approximately none.

   Durable — but privately held
6. 06
   Mutual ditch companies & shares

   Nonprofit by construction. Assessments recover cost — $18.03 per marginal Grand Valley share for 2026. The board's incentive is to keep water on the ditch, which is the principal governance obstacle to any outside buyer.

   Structurally non-profit
7. 07
   Agricultural land with attached water

   Farmland economics plus an unpriced option on a future change case. Farmland REIT returns on capital run 0.8–1.4%. The option is real, and after HB26-1340 its exercise price in the Arkansas basin includes an unbounded revegetation obligation.

   Thin, and getting thinner
8. 08
   Municipal & district offtake

   Where the money is spent and where profit is legally prohibited. Denver Water: $497.9m of 2024 revenue against $346.2m of capital spend, funded with $269.9m of new debt. The buyer of every acre-foot in this chain is a cost-recovery entity.

   Profit forbidden by design
9. 09
   Treatment, reuse & potable recycling

   Demand destruction at the margin, at roughly $17,000 per acre-foot of capital versus $67,500 for new storage. The capacity itself sits inside municipal enterprises; the profit is in the membranes, pumps, ultraviolet trains, and the engineers who design them.

   Profit sits one layer out
10. 10
    Equipment, services & technology

    Paid on activity, not on outcome. Return on invested capital of 12–27% across the pumps, valves, meters, pipe and analytics names, with capex under 10% of revenue for most. The only layer where a listed investor can buy a real return on capital tied to this chain.

    Durable and investable

Evidence: Layer 2 — Northern Water C-BT price series with BLS CPI-U; layer 5 — Colorado Supreme Court diligence record cited in §04; layer 6 — Grand Valley Irrigation Company 2026 assessments; layer 7 — FPI and LAND FY2025 Forms 10-K; layer 8 — Denver Water 2025 Approved Budget; layers 9–10 — XYL, VLTO, PNR, MWA, WMS, CNM, WTS, ERII FY2025 Forms 10-K. Method: Verdicts are the report's classification against a single criterion — whether the layer can earn a return on invested capital above its cost of capital that survives competitive and regulatory attack. Synthesis: Excess return and defensibility come apart in this chain: the most defended layers are the least profitable, because the defence was built by statute for a public purpose.

11.1Hydrology and the physical resource

There is no profit pool here and it would be a mistake to look for one. The layer sets the sign on every option below it. Water year 2026's collapse — 36% of average inflow to Powell, 22–24% of median runoff on the Western Slope — raises the value of storage and of pre-1922 seniority, and lowers the value of every junior direct-flow right whose delivery depends on unregulated flow. It also raises political temperature, and political temperature in Colorado water reliably converts into restrictions on transfer rather than into higher prices.

11.2The decreed right and the priority stack

This is the layer the consensus wants to own, and the one the evidence most clearly rejects as a business. Three facts, all from primary sources.

First, the price does not behave like a scarcity asset. Northern Water's own series shows the C-BT unit at $67,000 in 2023 and $57,487 in 2025. In 2025 dollars, the peak was 2020 at $74,149, and the 2025 price is 22.5% below it.

Second, the asset pays nothing. There is no coupon, no dividend, no rent. Northern Water's 2026 open-rate municipal, industrial and multipurpose assessment is $56.17 per unit, which against a $57,487 unit price is a carry of 0.098% a year — trivially small, and irrelevant, because the binding cost of holding the asset is the return forgone on the capital tied up in it.

Third, the arithmetic of an unlevered hold is unforgiving.

The grid shows what the price chart cannot: for a passive holder, the entire return has to come from nominal appreciation exceeding inflation, with nothing else contributing. Over 2019–2025 nominal appreciation was 20.6% against 25.9% CPI. Over 2022–2025 it was −12.8% against 10.0%. Over the full 2016–2025 window the holder did well — +117.0% nominal, +61.8% real, a 5.5% real compound rate — and that window contains the entire 2019–2020 step change. An investor who bought after the step has lost real money for six years while forgoing a risk-free yield that spent much of that period above 4%.

Known Known The price series and the CPI series are published, dated, and reproduced above. Consensus Assumption The claim that the 2019–2020 step change reflected a one-time repricing rather than the start of a trend is an interpretation; the falsifier is a sustained new high, and §21 dates it.

11.3Compact and federal administration

No private participant collects at this layer, and the exposure runs one way. If Arizona sues and prevails on the Article III(d) firm-delivery reading, Colorado's post-1922 rights — which include the transmountain diversions that supply most of the Front Range — face administration. If the Upper Basin's non-depletion reading prevails, the pressure moves to negotiated demand managementPaying water users — in practice, farmers — to use less for a period, so the saved water can be banked against a compact shortfall. It is a voluntary, compensated, temporary alternative to being shut off by order. The price is set inside a federal or interstate programme rather than by competing buyers, which is precisely why it does not help a private holder., which is a bid for temporary use of agricultural water at prices set by a federal programme rather than by a market. Neither branch creates a private toll. Both create work for water lawyers.

11.4Storage and transbasin conveyance

If any layer in this chain deserved to earn a monopoly rent, it is this one. A tunnel under the Continental Divide cannot be replicated; the Moffat and Roberts tunnels, the Alva B. Adams tunnel, the Homestake and Boustead tunnels are each the only route from one side of a mountain range to the other. Ownership is entirely public: Northern Water, Denver Water, the Homestake Partners, the Southeastern Colorado Water Conservancy District.

Three projects are adding firm supply now, and their fate says more about this layer than any margin analysis could. Chimney Hollow, 90,000 acre-feet for the Windy Gap Firming Project, had its initial partial filling delayed after routine testing found natural uranium leaching from quarry rock used as dam fill. Glade Reservoir, 170,000 acre-feet of capacity for 40,000 acre-feet a year of yield, has settled its federal permit challenge and is estimated at roughly $2.7 billion. Gross Reservoir's expansion, adding about 76,722 acre-feet, has a topped-out dam — Denver Water marked the roller-compacted concrete milestone on 3 June 2026 — and a federal injunction that bars filling the additional capacity, now before the Tenth Circuit after settlement talks failed (Colorado Sun, 28 May 2026).

A public utility with a decree, a federal permit, a completed dam and 1.5 million customers cannot reliably fill its own reservoir. Price the political risk of a private party attempting the same thing accordingly.

11.5The Water Court change-of-use and augmentation layer

Here is the toll booth, and it is not for sale. Every municipal acquisition of agricultural water must pass through a contested case with a water law firm, a consulting water resources engineer, expert hydrologic modelling, and years of referee proceedings. The Colorado Supreme Court's own diligence records show single matters running past $1 million in professional fees. The applicant pays whether the decree issues or not, and whether the quantified consumptive use comes in at 100% or 50% of what the buyer underwrote.

The economics are close to ideal: a licensed profession, a legally mandated demand, a captive payer, no substitute, and effectively no invested capital. The problem for a public-market investor is that the entire layer is organised as private partnerships. The listed engineering firms — AECOM, Jacobs, Tetra Tech, Stantec — do federal, municipal and environmental work at scale and earn respectable returns on capital, but none of them discloses Colorado water-rights adjudication as a segment, and none of the Colorado water bar is listed anywhere.

11.6Mutual ditch companies and shares

Nonprofit by construction, so there is no excess return to capture and no listed access. The governance friction is what matters, and it works as a transfer tax on outside capital. Water Asset Management's Grand Valley position is the live case study: about $16.6 million for 2,222 acres, roughly 7% of the Grand Valley Water Users Association's irrigated acreage, accumulated since 2017 and leased back to farmers. Six years on, the Colorado River District's general manager was publicly worried about the long-term plan and the state legislature had already tried to outlaw the structure.

11.7Agricultural land with attached water

The listed farmland vehicles do own water — just not in Colorado. Gladstone Land's FY2025 Form 10-K reports 55,532 acre-feet of water assets, all of it in California: 48,309 acre-feet banked with the Semitropic Water Storage District in Kern County and 7,223 surplus water credits at Westlands in Fresno County. Its Colorado holdings — 10 farms, 31,448 acres, $38.1 million of net cost basis — carry zero acre-feet of separately held water assets. Farmland Partners is headquartered in Denver and farms in Colorado's High Plains, and its 10-K discusses water availability as an acquisition diligence factor rather than as an asset class it holds.

Two farmland REITs with an explicit water strategy, one of them Denver-based, and both put their water capital in California. The reason is doctrinal: California's adjudicated basins and district banking arrangements permit a landowner to hold, store and monetise water in ways Colorado's prior-appropriation and no-injury regime does not.

11.8Municipal and district offtake

This is the demand side of every transaction in the chain, and it is a cost-recovery enterprise. A cost-recovery enterprise recovers what it spends and no more: rates are set to cover operations, debt service and capital, and any surplus goes back into the system rather than to an owner. Two of its revenue lines matter here. Water sales are the monthly bill, charged on volume used; a system development charge — the tap fee — is the one-time sum a builder pays for the right to attach a new building to the system at all, charged at the front and unrelated to how much water the household later draws. Denver Water's 2024 actuals: total revenue $497.9 million, of which water sales $383.1 million and system development charges $30.2 million; total capital spending $346.2 million; debt proceeds $269.9 million. Capital equals 69.5% of revenue, and tap fees fund under 9% of capital.

The response was a price rise, and its size is instructive. After no change to its system development charge schedule since 2013, Denver Water's board voted in March 2025 to raise it, phased in over 2026. From 1 July 2026 a single-family connection inside Denver pays a $3,380 base charge plus $1.01 per square foot for the first 22,000 square feet — $10,450 on a 7,000 square foot lot, against $7,930 in 2025, a 31.8% increase (Denver Water, 2026 System Development Charges). Outside Denver the same lot pays $14,680.

The ratepayer does not bear it. The homebuilder does, and through the builder the buyer of a new house. Denver Water's own framing is "the desire of Denver Water's customers to have growth pay for growth." A tap fee is a growth impact fee wearing a water costume, and its level is set by a public board's judgment about intergenerational fairness rather than by the scarcity value of an acre-foot.

11.9Treatment, reuse and potable recycling

Water imported across the Continental Divide, and non-tributary groundwater, may under Colorado law be used to extinction — reused and successively reused until it is gone — which means a Front Range utility that builds recycling capacity can serve the same acre-foot repeatedly. That permission is the most under-modelled force in the Colorado water thesis. Aurora's Prairie Waters draws river seepage the city has the right to reuse through riverbed wells near Brighton, cleans it through a gravel bed, and pumps it 34 miles uphill. A $13 million expansion was set to double capacity and add 5,000 to 10,000 acre-feet a year, against 55,000 to 60,000 acre-feet of total city use — capital of roughly $1,300 to $2,600 per acre-foot of added yield, against $67,500 for NISP's greenfield storage and $75,393 for a C-BT unit measured on the same firm-yield basis.

The brownfield-versus-greenfield caveat in §10 applies and matters: an expansion of an existing plant is the cheapest possible increment, and the next one will cost more. Even so, the gap is wide enough to survive an order of magnitude of correction. Colorado made itself the first state with a formal direct potable reuse rule in November 2022, effective January 2023, and the department noted at adoption that no Colorado utility had yet implemented direct potable reuse. Whether one has since could not be verified this session and is listed in §22 rather than used to carry a conclusion. The regulatory permission is in place and every dry year strengthens the case. Each acre-foot of reuse capacity built is an acre-foot of agricultural water never purchased.

11.10Equipment, services and technology

The layer that gets paid regardless. Denver Water spends $346 million a year on capital; NISP will spend $2.7 billion; Chimney Hollow, Gross, Prairie Waters expansions, PFASPer- and polyfluoroalkyl substances — the "forever chemicals" used in non-stick and stain-resistant products, which do not break down and turn up in drinking water. Removing them takes dedicated treatment equipment. For a utility that is compliance spending with no new customers attached; for an equipment vendor it is simply work. treatment and lead service line replacement are all funded by ratepayers and bondholders and are all contracted out. Nobody in this layer takes hydrologic risk, nobody needs a water court decree, and nobody's margin depends on what an acre-foot is worth.

The returns confirm it. Watts Water earns 18.8% on invested capital, Advanced Drainage Systems 14.4%, Mueller Water Products 13.7%, Core & Main 13.7%, Pentair 13.4%, Veralto 27.3% — all on capital intensity below 10% of revenue, most below 3%. The catch, and it is a real one, is dilution: none of these businesses is a Colorado water play. Their exposure to this chain is a single-digit percentage of revenue at most, and their cycle is US non-residential construction and municipal capital budgets, not snowpack.

12Who actually sets the price at each node

Run the chain node by node and ask, at each one, who can raise the price without losing the customer. The answer is almost never the party holding the water.

Evidence: Denver Water 2026 SDC schedule and 2025 Approved Budget; Northern Water assessments and price series; H2O America FY2025 Form 10-K (authorised ROE and rate base by state); Grand Valley Irrigation Company 2026 assessments; CWCB CWSA update 2025. Method: Each row identifies the party with unilateral authority to set the transaction price and the binding constraint disclosed in the cited source. Synthesis: Colorado's chain has exactly two unregulated price-setters — the professional layer and the vendor layer — and both are paid for activity rather than for the resource.

Scarcity sets the value of Colorado water. A public board sets its price. Those are not the same number, and the gap between them is the state's growth subsidy.

The central claim of this report

13Incentive map — read the incentives and most outcomes stop being surprising

Evidence: As cited throughout §§01–11; the $99m Shoshone transaction from Colorado Sun, 4 Feb 2026. Method: Incentive statements are drawn from each party's statutory mandate or disclosed conduct; consequences are the report's inference. Synthesis: Eleven of the twelve participants have an incentive that either blocks transfer or caps its price. The twelfth gets paid either way.

The Shoshone case inverts the standard story. In November 2025 the Colorado River Water Conservation District filed a change case to add an environmental use to the Shoshone hydropower right, which it is buying from an Xcel Energy subsidiary for approximately $99 million, with at least 18 parties pledging $37.3 million toward it. More than 60 parties filed by the February 2026 deadline, including Denver Water, Colorado Springs, Northern Water and Aurora. This is the largest publicised Colorado water transaction in recent memory, and it is a purchase by a public district, funded partly by local governments, whose entire purpose is to prevent the water from ever moving.

The marginal large buyer in Colorado water is a public entity buying to freeze the asset in place.

14Per-name reads — twenty-seven companies, graded on return on capital

Every name below was graded against its own most recent filing, opened and read for this report. Grades are structural, not valuation calls.

Return on invested capital is the test throughout, and it asks one question: for every dollar of shareholders' money and borrowed money sunk into a business, how much profit comes back in a year? Margin cannot answer it. A company can charge a wide markup and still destroy value if reaching that markup required an enormous amount of capital — and this chain contains several that do.

Two coverage results frame everything that follows. First, of the 27 filings examined, only Pure Cycle's mentions Colorado more than twenty times, and fifteen do not mention the state at all — American Water, Essential Utilities, California Water Service, Middlesex, Xylem, Veralto, Pentair, Mueller, Advanced Drainage, Core & Main, Watts, Energy Recovery, AECOM, Jacobs and Texas Pacific Land each return zero hits on "Colorado" in their latest annual report.

Second — and this is the harder test — a full-text search of every Form 10-K filed with the SEC since 1 January 2025 returns zero documents containing the phrase "Colorado-Big Thompson," zero containing "augmentation plan," and exactly three containing both "South Platte" and "water rights": Pure Cycle, and Molson Coors twice. The brewer reports that it both owns and leases water rights to sustain brewing operations, and names the Upper South Platte watershed as the home of its Golden, Colorado brewery. Colorado water appears in its filing as a raw material, alongside barley and aluminium — which is precisely the legal category the state's doctrine is built to favour.

Method note, because it decides several grades. ROIC is computed as operating income × (1 − effective tax rate) ÷ (total equity including non-controlling interests + total debt − cash and equivalents), all taken from the same fiscal-year-end balance sheet as reported. It is an ending-balance calculation, not an average, and it is unadjusted for operating leases, goodwill from acquisitions, or construction work in progress. For the regulated utilities the more meaningful test is the second one below: earned return on equity against the return the commission actually authorised.

14.1The water-rights and land vehicles

Three names in the mandate no longer exist as investments, and their fates are the argument.

Evidence: D.R. Horton / Vidler merger announcement, Exhibit 99.1 to Form 8-K, 14 April 2022; SEC order revoking Two Rivers Water & Farming registration, 18 August 2022; Colorado Sun, 4 Feb 2026. Method: Status confirmed against each entity's EDGAR submission history this session. Synthesis: The exit for a western water-rights portfolio has empirically been a homebuilder or a public district — the two buyers who need the water for something other than resale.

14.1aAV Farms — the live test of every claim in this report

One listed company is, right now, running the exact strategy this report argues cannot work. If it succeeds, the thesis is wrong, so it gets the most careful treatment in the brief.

On 14 February 2025 Select Water Reuse, LLC, a wholly-owned subsidiary of Select Water Solutions, entered a Delaware limited partnership — AV Farms, LP — with C&A Rollover Company, LLC and Geneses Water, L.P. as fellow limited partners and AV Farms Management, LLC as general partner, effective 28 February 2025. Select contributed $72 million and holds approximately 39% of the partnership and 25% of the general partner; C&A holds 38% and 50%, Geneses 23% and 25%.

What the partnership bought, in the company's own words: AV Farms "was formed to consolidate and commercialize one of the largest water holdings and storage portfolios in Colorado," consolidating 16,300 acre-feet of senior annual water rights, reservoir storage, and pipeline connectivity, laying the foundation for a commercial water banking and lease-fallow program aimed at "the under-supplied Colorado municipal and industrial market." It is also, the filing notes, "one of the largest irrigated farming operations in the Arkansas River Valley," with the stated goal of "developing an integrated water network in the Colorado Arkansas River Valley." Select expects to contribute approximately $74 million more over a three-year period.

Read against Parts I and II, the structure is well designed. It is a lease-fallow and bankingWater banking means storing water — in a reservoir or an aquifer — in the years there is surplus, and drawing it back out in the years there is not. The bank sells reliability rather than volume, which is the thing a city actually needs and the thing a bare irrigation right cannot offer. programme rather than buy-and-dry, which is the model Colorado policy subsidises rather than punishes. It keeps the water in beneficial use through the farming operation, which is the answer to anti-speculation. It holds senior rights, storage and conveyance together, which is the only combination that makes an Arkansas Valley acre-foot deliverable to a Front Range buyer. Whoever assembled it has read the same statute book.

Now the scoreboard, thirteen months in.

Evidence: Select Water Solutions FY2025 Form 10-K and Q1 2026 Form 10-Q, equity-method investment schedules and risk factors. Method: Carrying value less the $72.0m contribution. Equity-method carrying value moves with the investor's share of partnership earnings, losses and distributions; the filings do not separately disclose AV Farms' revenue or result, so the decline is consistent with an operating loss, a distribution, or both, and the report does not claim which. Synthesis: Thirteen months and $72m into the best-designed Colorado water platform a listed company has attempted, the carrying value is below cost and no earnings contribution is disclosed.

$72m for 39% implies roughly $185m of total partnership equity at formation, which against 16,300 acre-feet of senior annual rights works out near $11,300 per acre-foot — except that the partnership also holds reservoir storage, pipeline, and a large irrigated farming operation, so the water alone is worth materially less than that. Whatever the true split, AV Farms bought Arkansas Valley senior water at a small fraction of the $57,487 a C-BT unit cost in the same year. That gap is not an arbitrage. It is the market's estimate of the cost, time and risk of getting Arkansas Valley water to a Front Range tap — a change case, a no-injury proceeding, a quantification, and now, from 1 January 2027, HB26-1340's revegetation obligation, third-party monitor and financial assurance, because AV Farms sits squarely in Water Division 2.

Two further constraints are in the filing. Select does not control the venture: its own risk factors state it "may not have sole decision-making authority regarding AV Farms," which it flags as a risk not present in its other operations. And the partnership agreement contains buy-sell mechanicsContractual triggers letting one partner force the purchase or sale of another's stake, at a price set by a formula written in advance. They exist so partners can part company without agreeing on value. Here the formula is keyed to a target rate of return, which tells you the stake was structured to be exited on a schedule rather than held indefinitely. keyed to "the greater of a specified net internal rate of return on their respective capital contributions and a certain net multiple on invested capital" in one direction and the lesser of the same two in the other — the private-equity architecture of a fund, not of an operating asset.

14.2The regulated water utilities

The mandate asks that the utility layer be tested on allowed versus earned return rather than described as pricing power. A commission grants a utility a return on equity it is permitted to earn; whether it actually earns it is a separate question, settled by how fast it is spending and how quickly rates catch up. Do that and the layer looks very different from its reputation.

One development reframes the whole group. On 26 October 2025 American Water Works and Essential Utilities agreed to combine in a stock-for-stock merger, Essential shareholders receiving 0.305 American Water shares each, Essential becoming a wholly owned subsidiary, the combined company keeping the American Water name and its Camden headquarters, and closing estimated by the end of the first quarter of 2027. The two largest listed US water utilities are merging. Read that alongside the returns below: American Water earns 11.9% on capital and Essential 6.1%, both against allowed returns near 9.7%, both spending more than half of revenue on capital every year. When the two biggest operators in a rate-regulated industry conclude that the path to a better return runs through each other's balance sheets rather than through their own rate cases, that is a statement about the ceiling on the layer.

Evidence: Each company's FY2025 Form 10-K, opened this session — AWK, WTRG, MSEX, HTO, CWT, GWRS. Method: ROIC per §14 method note. Earned ROE = net income ÷ ending total equity. Capex ÷ revenue uses each filer's disclosed utility plant expenditure or property additions where tagged; blank where the filing does not report a single comparable figure. Allowed ROE is quoted only where the filing states one — AWK's is a single state order, not a system-wide figure, and is not comparable to HTO's four-state schedule. Synthesis: Not one of these six companies operates a regulated water utility in Colorado. The allowed return is a ceiling, and five of six earn below it — which is the context in which the two largest of them agreed to merge.

H2O America's disclosure makes the mechanism visible in a way most utilities do not. Its 10-K sets out, state by state, the rate base the commission authorised against management's estimate of the rate base actually in service at year end: California $1,308.0m authorised versus $1,460.8m estimated; Connecticut $784.1m versus $878.2m; Texas $96.2m versus $211.7m; Maine $148.7m versus $203.5m. In aggregate, $2,337m authorised against $2,754m in service — 17.9% of the asset base earning nothing at all until the next rate case catches up. That gap is regulatory lag, and it is the reason a 9.81% allowed return produced a 6.7% earned return.

Global Water Resources is the sharpest case, and the mandate is right to single it out as the listed model of a utility with an integrated water-resource strategy. Its "Total Water Management" model — 149 infrastructure coordination and financing agreements with landowners and developers, recycled water at the centre of the design — is genuinely differentiated and it does earn the company a growth-linked fee stream analogous to a tap fee. It has also produced a 2.5% return on invested capital and a 3.4% return on equity against a 9.20% authorised return, on capital expenditure equal to 120.7% of revenue, with earnings per share down 54.2%. Two utilities serving 87.3% of its connections were in a live rate case at year end in which the commission staff and the ratepayer advocate recommended materially less than the company sought. This is what a well-designed water-resource utility looks like inside a cost-of-service cage. It is a growth story and it is not, on this evidence, a compounding one.

Known Known The 9.20% authorised return, the 120.7% capex ratio, and the 54.2% EPS decline are disclosed in the FY2025 10-K. Known Unknown The pending rate case outcome is undetermined and could raise the earned return materially; the grade would move to neutral on a decision granting the requested revenue.

14.3Infrastructure, treatment and equipment

Evidence: FY2025 Forms 10-K for VLTO, WTS, MWA, CNM, PNR, ERII and XYL and the FY2026 (March year-end) Form 10-K for WMS, each opened this session. Method: ROIC per §14. Capex ÷ revenue uses tagged property, plant and equipment purchases; XYL's 3.7% = $331m ÷ $9,035m. Synthesis: Core & Main earns a higher return on capital than Pentair on less than half the operating margin. Margin locates profit in the income statement; only capital intensity tells you whether it is worth earning.

Xylem is the name most investors reach for as "the water company," and it is the one whose grade the return-on-capital test changes. Its four segments in 2025: Water Infrastructure $2,636m of revenue and $462m of segment operating income; Applied Water $1,849m and $312m; Measurement and Control Solutions $2,086m and $244m; Water Solutions and Services $2,464m and $302m. The franchise is real and the end markets are the right ones. But $1,223m of consolidated operating income against $11,954m of invested capital is 8.2% after tax, and that is at or below a reasonable cost of capital for an industrial of this quality. The Evoqua integration bought scale and a services annuity at a price that has not yet earned its keep. Neutral, not advantaged, and the falsifier is specific: sustained ROIC above 11% would move it.

14.4Engineering, consulting and adjudication services

This is the layer closest to the Water Court toll, and the only one where a listed investor can approximate it.

Evidence: Stantec FY2025 MD&A, Exhibit 99.2 to Form 40-F; AECOM FY2025 Form 10-K, Debt note; Tetra Tech FY2025 Form 10-K; Jacobs FY2025 Form 10-K; Onterris FY2025 Form 10-K and EDGAR former-name record. Method: AECOM ROIC uses total debt of $2,743.7m from the debt note rather than the tagged long-term figure, which understates it. Stantec's figure is the company's own non-IFRS adjusted ROIC, disclosed in its MD&A. Synthesis: The professional layer earns 11–21% on capital because it has almost none — the closest a listed investor gets to the Water Court toll, diluted across a global services book.

14.5Vehicles and indices

The water ETFs — Invesco Water Resources (PHO), First Trust Water (FIW), Invesco S&P Global Water (CGW) — hold the industrials and utilities graded above. None of them can hold a Colorado water right directly, because no listed instrument represents one; the closest any index constituent comes is Select Water Solutions' minority partnership stake, which is roughly 4% of that company's own invested capital before any index weighting. Their exposure to this subject is the equipment and services layer, repackaged, at an expense ratio. That is a defensible way to own the durable node, and it is a bet on municipal capital spending rather than on scarcity. The specific holdings and weights could not be retrieved from the issuers this session and are listed in §22 rather than asserted.

The Nasdaq Veles California Water Index and its CME futures are worth naming only to rule them out. The index prices spot water in five California markets. Its underlying instruments are adjudicated groundwater rights, district allocations and Central Valley Project contract supplies, none of which exist in Colorado law. Using NQH2O as a proxy for a Colorado price would be a category error of exactly the kind §03 warns about. Its value here is as a demonstration that a transparent, tradable water price is constructible — where the underlying is standardised and freely transferable. Colorado's is neither, and that is the whole reason no such index exists for it.

15The Pure Cycle verdict — the purest listed test of the thesis

Pure Cycle is a land developer with a captive utility franchise, and it is a good one. It is not a water-rights investment, and its own income statement says so.

The company describes itself as a diversified water and wastewater service provider, land developer, and home rental company along the Front Range. It holds 30,602 acre-feet of water across Denver Basin groundwater, designated groundwaterGroundwater inside a formally drawn basin on the eastern plains where the connection to any surface stream is slight enough that the state administers pumping under its own rules rather than through the water court. A third regime, in other words — and one more reason a single "price of Colorado water" does not exist. and surface water, of which 25,354 acre-feet is groundwater. It has the exclusive right to serve the Rangeview Metropolitan District's 24,000-acre Lowry Ranch service area, serves more than 1,600 single-family-equivalent water connections and 1,153 wastewater connections, and believes its portfolio can serve about 60,000 SFEs.

Take the water thesis on its own terms first, because something here is genuinely valuable. The Rangeview District's 2025 water tap fee is $34,200 per single-family equivalent, plus $8,240 for wastewater — $42,440 in total, and Pure Cycle receives 100% of it after a 2% royalty to the State Land Board. Denver Water, in the same metropolitan area, charges $10,450 for a single-family connection on a 7,000 square foot lot from 1 July 2026, after a 31.8% increase. Pure Cycle's water tap fee is 3.3× Denver Water's.

That is real pricing power, and its source is not scarce water. It comes from being the exclusive service provider inside a district where the alternative is no service at all — a greenfield monopoly granted by contract, on land that has no other route to a tap. A developer at Sky Ranch cannot shop. Denver Water's fee is set by a public board balancing growth against equity; Pure Cycle's is set by a district it serves under an exclusive agreement. Same metro, same aquifer, one of them free to price.

Now the income statement, which is where the water narrative and the business part company. Fiscal 2025, ended 31 August 2025:

Evidence: Pure Cycle FY2025 Form 10-K, Consolidated Statements of Income and MD&A. Method: Figures as reported, rounded to $0.1m. Percentage changes computed from the unrounded amounts. Synthesis: Fiscal 2025's earnings growth came from mineral royalties and interest on cash, not from the water franchise.

Just under fifty-seven per cent of fiscal 2025 pre-tax income came from oil and gas royalties and net interest. The company's own MD&A says the earnings-per-share increase "was driven by our oil and gas royalty income." An investor buying Pure Cycle for Colorado water scarcity bought, in that year, a Denver-Julesburg Basin mineral royalty and a Treasury bill, with a land development business attached and a water utility contributing about a third of operating revenue.

The nine months to 31 May 2026 tell a healthier operating story and confirm the same structure: total revenue $22.5m against $14.9m, operating income $6.8m against $1.7m, lot sales $10.7m, tap fees $5.6m — and oil and gas royalties down to $2.2m from $5.9m, with interest income of $2.8m. The operating business is improving. The dependency is unchanged.

The capital question decides the grade. Pure Cycle earned $7.67m of operating income in FY2025 on invested capital of $127.6m, a return of 4.5% after tax. Total equity was $142.7m against total assets of $162.3m, with only $6.8m of debt and $21.9m of cash. It is an unlevered, over-capitalised balance sheet earning less on its operations than its cash earns in the money market. That is not a criticism of management, who have been careful and patient. It is a statement about the asset: a portfolio of Colorado water rights carried at historical cost cannot generate a competitive return on capital until it is converted into taps, and taps require lots, and lots require homebuilders.

Three structural observations that the water narrative obscures.

First, the water is mostly groundwater, not a prior-appropriation surface right. Of 30,602 acre-feet, 25,354 is groundwater and only 4,578 is surface water. Denver Basin non-tributary and not-nontributary groundwater is allocated under a statutory hundred-year-life framework, not by priority date, and is not curtailed by a river call. That insulates Pure Cycle from the compact risk in §11.3 — genuinely a strength — and it also means the company's asset is not the senior seniority everyone is arguing about. It is also, per Franktown, the category where the anti-speculation doctrine has the least purchase.

Second, the State Land Board is a partner with a claim on the upside. Pure Cycle pays the Land Board a 12% royalty on gross water revenue from customers on the Lowry Ranch and from private customers off it, 10% from public entities off it, and 2% of gross tap-fee proceeds outside Sky Ranch. If metered production on the Lowry Ranch exceeds 13,000 acre-feet in a year, or 10,000 acres are rezoned, platted and tapped, the Land Board may elect to take 50% of collective net profits from water sales on the ranch instead. The company states neither condition is likely to be met soon. It is still a state-held call option on the most successful version of the business.

Third, revenue recognition ties the water segment to the homebuilding cycle. Tap fees are recognised when a tap is sold, and taps are sold when lots close. In the three months to 31 May 2026 the company sold 66 water taps generating $1.9m. The water business's revenue is therefore a derivative of Taylor Morrison and its peers deciding to start houses — the same variable that WestWater identifies as the cause of the C-BT price decline. Pure Cycle's water revenue and the C-BT market price share a driver, and it is not hydrology.

16Can a private investor actually own Colorado water for profit?

The answer differs by vehicle, and it has three parts.

For a listed vehicle: one indirect, minority, unproven expression exists — and nothing else. This is a coverage finding, not an opinion. PICO became Vidler and was bought by D.R. Horton in 2022. Two Rivers Water & Farming had its registration revoked in August 2022. Pure Cycle remains, and §15 shows what it actually is. The single genuine Colorado water-rights exposure available in a listed security today is Select Water Solutions' 39% equity-method stake in AV Farms, which it does not control, which produced no disclosed earnings in its first thirteen months, and which represents about 4% of Select's own invested capital. Everything else in the set is a land developer, a brewer buying an input, or a global equipment company.

That is a meaningfully different statement from "nobody has tried." Capital has tried, repeatedly and intelligently, and the results are on file: a homebuilder takeout, a revoked registration, a land developer earning 4.5% on capital, and a $72m partnership stake carried below cost. The constraint is not a shortage of ideas.

For a private fund: legally possible, structurally punished, and politically exposed. Water Asset Management has done it — roughly $16.6m for 2,222 acres in the Grand Valley since 2017, about 7% of the Grand Valley Water Users Association's irrigated acreage, leased back to farmers so the water stays in beneficial use. That last clause is the whole design. Anti-speculation doctrine bars acquiring a water right without an intent to apply it to beneficial use, so the fund must keep farming, which means it must accept farm returns while it waits. The wait has now run eight years.

Four constraints bind, and each is separately sufficient to compress the return:

1. 1Anti-speculation. The doctrine survives, and the legislature has tried to sharpen it. SB22-029 would have prohibited buying mutual ditch shares with intent to profit from resale or from payment for non-use, with fines up to $10,000 and state engineer approval required on transfers for up to two years after a fine. It failed — laid over in committee on 21 April 2022. That is a reprieve, not an acquittal: the state convened a work group on speculation again in 2025, and the political direction has been one-way for six years.
2. 2No-injury and the quantification haircut. The transferable quantity is not known until a court says so, years after the purchase price is set, and it is quantified on actual historical consumptive use rather than the decree. The buyer underwrites a number the court is free to reduce.
3. 3No yield, and now new liabilities. Carry is cheap — $56.17 per C-BT unit, $18.03 per marginal Grand Valley share — but the asset pays nothing, so the entire return is price. And from 1 January 2027, an Arkansas-basin change case carries a revegetation obligation, a court-appointed monitor, financial assurance posted to the county, and possible restrictions on using the water until the ground is established.
4. 4The exit is a cost-of-service buyer with a cheaper alternative. The purchaser is a municipal utility answering to ratepayers, whose alternative to buying your water is recycling supply it already owns, at a fraction of the capital cost per acre-foot. It is not a motivated bidder. It is a monopsonistA monopoly on the buying side — one buyer facing many sellers, and therefore able to name its price. The mirror image of the seller's monopoly everyone worries about, and in this chain the position the municipal utility occupies. with a substitute.

For a landowner or an operating business: yes, and this is the answer everyone skips. The returns in Colorado water do not accrue to holding the right. They accrue to having already owned it when a use arrived — a farmer whose family bought shares in 1910 and sells to a city in 2026, a developer whose land came with Denver Basin groundwater, a district that firmed its supply with federal storage in 1957. Every one of those is a windfall to an incumbent, not a return earned by capital deployed for that purpose. That distinction is the difference between the asset having appreciated and the investment having worked.

17The bear case against this report, argued to win

Parts of the opposing view are right. This is its strongest version.

The thesis mistakes a cyclical pause for a structural ceiling. The C-BT price decline from 2023 to 2025 coincides precisely with a residential construction downturn — WestWater's own analyst attributes it to exactly that, and CU Leeds data shows Colorado residential permits peaking in 2021 and declining every year since. Housing cycles turn. When Front Range construction resumes, the same 2019–2020 step change that took the unit from $28,600 to $59,609 in two years can happen again, and a −4.2% real return measured from the top of a repricing tells you about the entry point, not about the asset.

The thesis under-weights the ratchet. Every acre-foot converted to municipal use is converted permanently. The transferable agricultural stock only shrinks, and Colorado has lost 474,000 irrigated acres in five years. Meanwhile the state's own plan projects a 2050 municipal gap of up to 740,000 acre-feet. Reuse helps, but Prairie Waters was built over fifteen years and no Colorado utility had implemented direct potable reuse as of the department's own account at the rule's adoption. A fixed and shrinking supply against a documented and growing gap has exactly one long-run price direction.

The thesis treats anti-speculation as binding when the courts have been loosening it. SB22-029 failed. Franktown Citizens Coalition II held in 2025 that the doctrine does not reach augmentation plan amendments for Denver Basin groundwater. Water Asset Management has held Grand Valley land for eight years without sanction. The legal risk is a headline risk that has not yet produced an adverse holding against an investor.

And the thesis is measured in the wrong currency. A C-BT unit at $57,487 is not competing with a Treasury bill. It is competing with the cost of the alternative — $67,500 an acre-foot for NISP's new storage, a project whose cost has already risen far enough that Northern Water halted contract bids and began scaling back designs to hold its participant coalition together. On the report's own headline numbers, the market price of firmed, transferable, immediately deliverable water sits below the replacement cost of building it. That is the classic definition of an asset trading below intrinsic value.

Where the bear case is right, and where it breaks

The cyclicality point is correct and this report accepts it. The 2023–2025 decline is a demand-cycle event; that is exactly the argument. What it establishes is not that the asset is worthless but that its price is set by a construction cycle rather than by hydrology, which means it is a levered play on Front Range homebuilding held in an unlevered, unyielding, illiquid form. There are better ways to own that exposure, and one of them files 10-Qs.

The ratchet point is also correct and does not survive contact with the buyer's alternative. The gap is real. But a gap is closed by whichever supply is cheapest, and the state's plan lists nineteen tools of which permanent agricultural transfer is one. Reuse capital cost per acre-foot of added yield runs an order of magnitude below new greenfield storage even after allowing for the brownfield mismatch. Conservation is cheaper still. Permanent purchase is the tool with the highest political cost and, after HB26-1340, a rising legal cost. When demand grows, the marginal acre-foot comes from the cheapest source, and that source has not been agricultural purchase for a decade.

On anti-speculation, the bear is right that no adverse holding has landed on an investor — and wrong about what that proves. Franktown loosened the doctrine in a narrow groundwater context that benefits a district serving a development, not a fund holding for resale. And the binding constraint on a fund was never a court. It is that the fund must keep the water in beneficial use to avoid the doctrine, which means accepting farm-level returns for the duration, which is the return compression this report describes. The doctrine does not need to be enforced to do its work.

The replacement-cost argument is the strongest of the four, and it is a comparison across mismatched scope. NISP's $67,500 per acre-foot buys a new, permanent, firm, decreed and permitted supply with a reservoir attached, for a consortium of municipal participants who could not otherwise obtain one. A C-BT unit at $57,487 buys 0.762 acre-feet a year on average, inside one district's boundaries, subject to a board's quota. Per acre-foot of long-run firm yield, the unit costs $75,393 — 11.7% above NISP's replacement cost, not below it. The apparent discount is an artefact of comparing a unit price to an acre-foot price. That is the comparability trap this subject sets, and the bear case walks straight into it.

18Alternatives rejected

The Permian produced-water royalty model, imported to Colorado. This is the most seductive comparison in the mandate and it does not work, for a reason that is legal rather than commercial. Texas Pacific Land earns $124.2m of produced-water royalties and $169.7m of water sales — $307.5m, 38% of revenue, at a 74.2% consolidated operating margin and 35.5% return on capital — because it owns the surface estate in a state that follows the rule of capture for groundwater and treats produced water as the operator's to handle. TPL is paid a percentage of volume for water crossing land it already owned. It is a toll on flow.

Colorado forecloses every element. The state owns the water; a decreed right conveys use, not ownership; tributary groundwater is administered within the priority system rather than by capture; and any change in the type or place of use requires a water court decree under a no-injury standard. There is no legal construct in Colorado under which a private landowner collects a percentage of water moving across or beneath their land. The model is not merely absent from Colorado. It is prohibited by the architecture.

Select Water Solutions does the physical work TPL is paid a royalty on: $1.41bn of revenue, $28.8m of operating income, $294.6m of capex, and a 2.4% return on invested capital. TPL earns 35.5% because it holds the royalty; WTTR earns 2.4% because it holds the pipes. In Colorado, only the pipe-holder role is available, and it is held by public districts.

A Colorado water index or futures contract. Rejected on feasibility. NQH2O works because California's underlying instruments are standardised and transferable within defined markets. A Colorado equivalent would have to index across seven water divisions, thousands of distinct decrees with different priority dates, quantified consumptive use that is unknown until adjudicated, and delivery points that are not interchangeable. The C-BT unit is the only standardised Colorado instrument, and it is a single district's storage share. That is not an index; it is a security with one issuer.

The water ETFs as Colorado exposure. Rejected on construction. PHO, FIW and CGW hold the industrials and utilities in §14.3 and §14.2 — the layer this report grades advantaged — and none of the underlying names has meaningful Colorado water exposure. Owning them is a reasonable way to own municipal water capital spending. It is not exposure to Colorado water rights, and it should not be sold as one.

Buying farmland with attached water and waiting. Rejected on return on capital. The listed evidence is unambiguous: Gladstone Land at 1.4%, Farmland Partners at 0.8%, Limoneira at −6.1%. The one company in the set that monetised water at a headline price — Limoneira, at $30,000 an acre-foot — did so by selling the asset, booking a $1.5m gain against a $20.4m operating loss. That is asset conversion, not a business.

19Stress tests

The mandate asks what happens to the durable profit under four specific shocks. Each is run against the layer verdicts in §11.

Evidence: §§05, 06, 11.9, 16 sources. Method: Directional stress test against the layer verdicts; no probabilities are assigned because none can be sourced. Synthesis: Three of four shocks are negative or neutral for the decreed right and positive for the services and equipment layers. That asymmetry is the report's central finding restated as a risk matrix.

The consensus reads the compact call backwards. A call is bullish for the relative position of a pre-1922 right and bearish for the market in transferable water, because the buyers are municipalities whose own supplies are being curtailed and whose response is to build reuse rather than to bid harder for a delivery they cannot rely on. Seniority becomes more valuable and less saleable at the same moment.

20Second-order effects

The tap fee is the real price of Colorado water, and it is rising faster than the water is. Denver Water's connection charge rose 31.8% for a standard lot in 2026 after thirteen years unchanged, while the C-BT unit price fell 14.2% from its 2023 peak. The asset got cheaper; the right to connect to the system got dearer. That gap is the value of the delivery network, and it belongs to public enterprises.

Reuse strengthens the equipment moat precisely because water gets scarcer. The standard reading is that scarcity is bullish for water rights. The mechanism runs the other way: scarcity forces utilities toward the cheapest marginal acre-foot, which is recycled, which is an engineering problem, which is bought from Veralto, Xylem, Pentair and their peers and designed by Stantec, Tetra Tech and AECOM. Every degree of aridification transfers spending from the asset owners to the vendors.

HB26-1340 will show up as a price cut, not a cost. A revegetation obligation with bonding and monitoring does not raise what cities pay per acre-foot; it lowers what sellers net, because the buyer prices the liability into the offer. The bill was written to protect rural counties and its economic incidence falls on the farmer selling out. Watch the Arkansas basin for a widening spread against the South Platte from 2027.

The Shoshone purchase creates a template for buying water to immobilise it. A $99m public acquisition whose purpose is to keep a right in place, funded partly by 18 local governments, is a new category of demand — one that competes with municipal buyers, is not price-maximising, and permanently removes supply from the transferable stock. If it is replicated, the effect on the market is ambiguous in the short run and unambiguously contractionary in the long run.

The regulated layer is consolidating because its return is capped, and that tells you where it thinks growth is. American Water and Essential agreed to merge in October 2025, targeting a first-quarter 2027 close. Neither operates in Colorado, and neither is buying the other for water. They are buying rate base, because rate base is the only variable a cost-of-service utility controls — the allowed return is set by someone else and, on this evidence, routinely under-earned. Expect the same logic to keep pulling small municipal systems into investor-owned hands, and expect none of it to change what an acre-foot of Colorado water is worth.

A compact call would be the largest work event in the history of the American water bar. Original-jurisdiction Supreme Court litigation between seven states, with every major municipal utility, district and irrigation company appearing, running for years. Nothing about it improves the cash economics of holding a water right. It is a fee event, and the fees are earned by an unlisted profession with listed adjacencies.

21Falsifiers and dated predictions

Dated, testable predictions — three to five years

Evidence: Each prediction names the published source that will settle it. Method: Confidence tags follow the framework's six-level scale; none is offered as a forecast with a probability attached. Synthesis: The predictions are chosen so that a reader can adjudicate them without access to anything private.

Evidence: Reclamation post-2026 programme page and July 2026 24-Month Study; HB26-1340 as enacted; Northern Water quota calendar; Colorado Sun reporting on NISP and Chimney Hollow. Method: Forward events with a published date or a statutory trigger only; no speculative catalysts. Synthesis: Four of six are administrative or legal rather than commercial, which is itself the report's thesis in calendar form.