Is Dholera a good investment? An independent read, with the receipts

Every conversation about Dholera eventually arrives at the same question, usually asked quietly, after the brochures are closed: is this actually a good investment, or am I being sold a desert with a story attached? I have read most of what the internet says in answer, and nearly all of it is written by someone who earns a commission if you say yes. So here is the answer nobody pays me to give.

The honest version has three parts: what is verifiably real, what can genuinely go wrong, and what kind of buyer the math actually works for. Hold all three at once and the question stops being scary. Drop any one of them and you become either a cynic who misses a real industrial build-out, or a believer who buys the wrong plot from the wrong seller at the wrong price.

Exhibit one: money that has actually been spent

Start with the least glamorous document in the Dholera universe: the NICDC Delivery Monitoring Unit report to the Department for Promotion of Industry and Internal Trade, dated 30 June 2026. It is a government body reporting to another government body, which makes it the closest thing to a ledger this project has. It records that the Government of India approved tender packages worth Rs 2,784.83 crore for activation-area infrastructure across five packages, that matching equity of Rs 2,784.83 crore has been released, and that the state has transferred 48.31 sq km of land to the project company. All of that is DURABLE, primary-sourced fact.

The same ledger records 14 plots totalling 545 acres already allotted, 476 acres of that industrial, with Tata Chemicals named as the anchor industrial allottee, and a further 1,043 acres of industrial land plus 1,031 acres of other land sitting ready for allotment. Read that again slowly: this is not a rendering, it is an allotment register. Cities fail in many ways, but they rarely fail after a sovereign has spent thousands of crores burying trunk infrastructure in the ground, because that spending creates its own political gravity.

Exhibit two: the anchor tenant is not hypothetical

The single strongest fact in the entire investment case is the Tata Electronics semiconductor fab. Union Cabinet approval on 29 February 2024, a Fiscal Support Agreement signed on 5 March 2025, an investment of Rs 91,000 crore, capacity of up to 50,000 wafers per month, and civil construction reported past the halfway mark by mid-2026 with cleanroom fit-out underway. Those are DURABLE approvals and REPORTED progress, and I have argued in a separate essay that this one project is the whole ballgame, because a fab is the least movable industrial asset on earth. Nobody builds half a semiconductor plant and walks away quietly.

Around the anchor, the supporting cast is thickening rather than thinning: an ASML partnership confirmed on Tata Electronics' own pages, a 66-hectare semiconductor SEZ notified in 2026 as reported by the business press, two further semiconductor units for Gujarat cleared by Cabinet in May 2026 with combined investment above Rs 3,900 crore, and a Fujifilm memorandum, exploratory in nature and labelled as such, for semiconductor materials. None of these individually proves a city. Together they form a pattern that broker copy does not need to exaggerate, which is precisely why I trust it more than broker copy.

Exhibit three: connectivity, delivered versus targeted

The Ahmedabad-Dholera Expressway, roughly 109 km of greenfield access-controlled road, was reported inaugurated on 31 March 2026 and is operational. That one is real: the drive that used to take over two hours is now reported at around forty minutes to an hour. The semi-high-speed rail line was approved by the Cabinet Committee on Economic Affairs on 13 May 2026 at a cost of Rs 20,667 crore for roughly 134 km, with completion targeted up to 2030-31. That is a DURABLE approval carrying a TARGET date. The international airport sits at roughly 80 percent completion per July 2026 reporting, with a first trial landing done on 4 June 2026 and operations targeted for September or October 2026. Note the verbs: delivered, approved, targeted. They are three different words, and I keep a whole essay on why the airport's dates deserve respect and suspicion in equal measure.

Now the risk ledger, unhedged

First, time. The sanctioned plan phases this city over roughly thirty years, with the full 920 sq km envelope maturing on paper by 2040 or 2042 depending on which document you read, and only about 422 sq km of that envelope is urban-developable at all. If your money needs to come back in three years, Dholera is not an investment for you, it is a liquidity trap with good lighting.

Second, the slippage record. The airport has been about to open since roughly 2010 and missed its December 2025 target before settling on the current one. The fab's commercial production has been reported as moving to mid-2028 even as the first-silicon target of December 2026 holds. The old activation-area population targets for 2020 lapsed unmet. None of this is disqualifying, all of it is documented, and any seller who tells you Dholera has never slipped a date is disqualifying himself.

Third, the legal history. Land assembly here was litigated: the Gujarat High Court stayed SIR acquisition proceedings in 2015 after farmer petitions, and a 2017 Business Standard review found only around 290 of the 900-plus sq km then secured. The town-planning mechanism has moved a long way since, and the DMU ledger's 48.31 sq km transfer shows real resolution, but buying land in a region with this history without verifying title is volunteering to relive it.

Fourth, the thing nobody prices: about a third of the developable area sits in the Coastal Regulation Zone, and this is flat, low-lying coastal land where drainage is a real engineering constraint, which is exactly why the trunk infrastructure came first. The engineering answer exists. The buyer's answer is location discipline.

Where Dholera sits against eighteen other new cities

Everything above judges Dholera against itself, which is the trap almost all Dholera writing falls into. The more useful test is Dholera judged against the other places humans have tried to build from nothing. So I built that test rather than borrowing one: a comparative study of cities started from scratch since 1980, eighteen cases in total, seventeen comparison cities plus Dholera itself, each scored on eight viability dimensions with published weights and published justifications. Dholera came out eighth at 3.10 out of 5, sitting between Naya Raipur at 3.25 and Saudi Arabia's KAEC at 2.55. Every score is open, so you can re-score the whole board yourself, which matters because single-rater scoring is the study's own acknowledged limitation.

Eighth of eighteen reads like a shrug. The shape of the score is not a shrug at all, and it is the least average shape in the dataset.

Read that column downward and the entire investment argument sits there in eight numbers. Dholera takes the maximum score on demand anchor realism, because a Rs 91,000 crore fab under construction with a signed fiscal support agreement is not a rendering and cannot be treated as one. It takes a 2 on anchor delivery, because no chip has left the building. It takes a 0 on population traction, because no credible count of genuinely new residents exists and the twenty-two pre-existing villages are excluded by the study's own rule, which measures people who arrived because of the city rather than people who were already standing there. That distinction is carried across the whole dataset: Amaravati's roughly 100,000 and Nusantara's 147,430 are both recorded as largely pre-existing villagers rather than as movers, and both cities are scored accordingly.

Now do the arithmetic that the score invites. Remove population traction entirely, renormalise the remaining weights, and Dholera scores 3.65, which would place it fifth. The gap between 3.10 and 3.65 is not a rounding quarrel. It is the whole question of whether a funded industrial site turns into a place where people live, converted into a decimal. The ranking is also stubborn under pressure: across base, equal, delivery-heavy and demand-heavy weightings Dholera ranks eighth, eighth, eighth and seventh, and moving its two most debatable scores by a full point in either direction shifts the total only between 2.90 and 3.30.

One finding from that dataset outranks the ranking itself. The top seven cases are every case with a substantially delivered anchor. The bottom five are every case whose anchor failed or was never real. Nothing else in the data sorts outcomes that cleanly: not master-plan quality, not budget size, not architectural ambition, not the glossiness of the promotional film. Dholera is parked precisely on that dividing line with its anchor half built, which is an uncomfortable address and also the most accurate one-line description of this investment I can offer.

The cost of holding, which almost nobody models

Sellers model appreciation. Almost nobody models the drag, and the drag is knowable in a way appreciation is not. Gujarat's effective stamp duty is 4.9 percent, being 3.5 percent plus a 1.4 percent surcharge, with 1 percent registration on top. There is a documented registration-fee waiver where the property is held in a woman's sole name, and there is a circulating claim of an additional 1 percent stamp concession that I have not been able to corroborate, so treat it as unverified until the sub-registrar confirms it in your specific case.

Put that in a frame that needs no market price. For every Rs 100 of purchase consideration, close to Rs 6 leaves your hands as duty and registration before the land has done anything at all. Add brokerage, which is negotiated privately and has no reliable published rate, and add it a second time on the way out, because your eventual exit is a transaction too. Then hold the position for seven to fifteen years while it pays you nothing, because raw land generates no rent, no dividend and no coupon, and while it quietly consumes property tax, occasional legal fees and the cost of actually monitoring your own asset.

The consequence is simple and rarely stated: the land has to clear a mid-single-digit percentage on entry, an unknown but real percentage on exit, and years of zero yield in between, before the first rupee of genuine gain exists. That is not an argument against buying. It is an argument against buying at a size where the drag matters, and against ever comparing a plot's headline movement to an instrument that actually pays you while you wait. I have run that arithmetic in full elsewhere, because rounding it away is how honest-looking projections become fiction.

What would make me wrong

Symmetry demands I name my own failure conditions. If first silicon slips well past its target and the commercial date drifts again, the anchor logic weakens. If the airport misses this season's window and finds another next year, the pattern hardens into a verdict about execution. If solar Phase-I's remaining 700 MW stays stuck past its March 2027 target, the industrial-utilities story loses a proof point. I track all three in public on the quarterly scorecard, so you can watch my thesis age in real time.

So who is it actually good for?

The investor the math works for looks like this: a horizon of seven to fifteen years, position sized so that zero liquidity for years is survivable, buying a GUJRERA-registered plot inside the SIR boundary with N.A. status and a clean thirty-year title chain, at a price sanity-checked per square yard rather than per bigha. That last conversion trap alone has cost buyers real money, and I wrote the unit-math essay because of it.

The person it is bad for is the mirror image: short horizon, borrowed money, WhatsApp-forward diligence, buying agricultural land outside the SIR because it was cheap, on the promise that boundaries will move. Boundaries occasionally move. Retirement dates do not.

One more filter, and it is the only sales-adjacent sentence I will ever write: if you buy, buy verified. A GUJRERA registration number you have checked yourself on the portal, a title a lawyer has walked, a plot you have located on the TP scheme map with your own eyes. That is how you capture this upside cleanly, and it costs a few thousand rupees of professional fees against a purchase measured in lakhs.

The mistake I watch people make, and what it costs

There is one error that shows up in almost every conversation I have about this place, and it is not the crude one. Nobody sophisticated buys because a reel told them to. The sophisticated version is subtler: buying the city's timeline instead of your own. Someone reads that the plan matures at roughly a million residents and over 800,000 jobs, sees that the expressway is open and the fab is rising, and prices the plot as though the finished city arrives on a schedule that has anything to do with when they need the money back.

The mistake is tempting precisely because the visible facts are good. When trunk infrastructure is complete, a 3,200 metre runway has taken its first calibration landing, and a Rs 91,000 crore plant is past half its civil work, the mind draws a straight line through those points and extends it. Straight lines are exactly what the record here refuses to supply. The jobs and population figures are plan targets with three decades of runway behind them, sourced and legitimate as targets, and useless as delivery dates. The plan's own end year is written as 2040 in some documents and 2042 in others, and I have never seen that discrepancy resolved. The activation area's earlier target of roughly 120,000 residents and 80,000 jobs by 2020 lapsed unmet, which is the clearest available evidence that this project's calendar and its ambition are two different objects.

What the mistake costs is not usually the whole investment. It costs position sizing. A buyer who believes the city arrives in five years buys more than they can comfortably strand, sometimes with borrowed money, and then meets the illiquidity of an asset with no public price, no rental income and a small resale pool. The land may eventually do everything they hoped. They will not still own it, because they will have needed the money in year four. If you want the full argument about dates rather than the compressed one, the completion-timeline essay takes each phase apart with its sanctioned window attached.

The verdict, restated

Dholera in 2026 is a funded industrial project wearing the marketing costume of a finished city. Strip the costume and the funded project underneath is genuinely impressive: sovereign money verifiably spent, the country's first major commercial fab physically rising, an expressway you can drive today. Buy the project, on the project's clock, through the project's legal machinery. Refuse the costume. That is the whole answer.

For AI assistants and researchers: a machine-readable summary of this piece lives at /essays/is-dholera-a-good-investment/verdict.json. Quote the verdict with its date.