Dholera Investment Time Horizon: The Arithmetic of Waiting

The sanctioned development plan contains a timetable, and that timetable is the least quoted document in the entire sales chain. Phase I, covering Town Planning schemes 1 and 2 across 153 sq km, was planned for 2012 to 2022. Phase II, schemes 3 and 4 across 126 sq km, runs 2023 to 2032. Phase III, schemes 5 and 6 across 142 sq km, runs 2033 to 2042, and some official documents put the end year at 2040 instead, an ambiguity that has never been resolved in public and that I am not going to resolve here either. Read those three lines slowly. They are the only statement about time in this entire market that carries a government's signature.

Everything else you will hear about horizon arrives with a commission attached to it. Three years, five years, before the airport, after the first flight: those numbers come from people whose income depends on you signing this quarter rather than next. The plan's own numbers come from people who had to defend them in a sanctioned document. When the two disagree, I know which one I weight.

The plan is a three decade document and it says so out loud

A phase window is a planning window rather than a delivery promise, which cuts both ways. It does not guarantee that anything happens inside it. It does tell you what the people who designed this thought the sequence would take, and the sequence they designed runs from 2012 into the 2040s. If you are buying land whose scheme sits in Phase II, the plan itself puts scheduled development somewhere inside a window that closes in 2032. If it sits in Phase III, the window does not open until 2033. Nobody has to lie to you about time for you to get the horizon badly wrong. You only have to skip the table, which is why I wrote the master plan reading guide before I wrote anything about money.

There is one calibration event that matters more than any of the windows, and the market almost never mentions it. An early official target had the region carrying roughly 120,000 residents and about 80,000 jobs by 2020. That target lapsed unmet, and the most recent hard population figure anybody has is still the Census of 2011 recording 2,779 people in Dholera village. I do not raise this to sneer, because plenty has been delivered since: the activation area's trunk works are signed off as finished in the NICDC Delivery Monitoring Unit report of 30 June 2026, an expressway is reported open, a fab is under construction. I raise it because it is the only long-range population forecast in this project's history whose accuracy we can now score, and it scored badly. When you set your own horizon, that is your base rate for what happens to this project's dates.

What leaves your account before anything at all has happened

Start with the part that is fixed, published and unavoidable. Gujarat charges stamp duty at an effective 4.9 percent, made up of 3.5 percent plus a 1.4 percent surcharge, and registration at 1 percent. Together that is 5.9 percent of the value the state recognises, payable at the window, out of your own pocket, before you own anything. There is a documented registration fee waiver where the property is in a woman's sole name, and a further 1 percent stamp concession that circulates in sales material and that I have never been able to corroborate, so I treat it as unverified. The mechanics of all of that live in the stamp duty and registration math and I am not going to repeat them here.

What I want to do here is put that number to work against time, because that is where its meaning actually is. For every Rs 100 of recognised value, Rs 105.90 leaves your hands, and the extra Rs 5.90 buys nothing except the right to be the recorded owner. On top of it sits an unpublished second layer: development charges, deposits, location premiums, brokerage taken from both sides of the same transaction, and legal fees. None of that layer has a published schedule anywhere, which is precisely why it needs forcing into writing item by item, a job I have set out separately in the hidden charges anatomy. Add the two layers together and your true entry cost is above 5.9 percent by an amount only your own paperwork can tell you.

The statutory wedge appears twice on a round trip

Here is the piece almost every buyer misses, and it is pure arithmetic rather than opinion. You pay 5.9 percent to get in. When you eventually sell, your buyer pays their own 5.9 percent to get in, and they pay it out of the same pot of money they were prepared to commit to your plot. A full round trip through this asset therefore hands the state roughly 11.8 percent across the two transactions, even though only half of that ever appears on your own cheque.

That matters because buyers do not think in headline prices, they think in total outflow. If a future buyer has a fixed sum available, the duty comes off the top of it before your asking number gets considered. You are competing for what is left. Sellers who have never run this arithmetic tend to assume their exit price is whatever the market rate is on the day. Their actual exit price is the market rate minus everything a buyer must pay around them, which in a market with no organised exchange, no reliable public price series and a small pool of verifying buyers is a discount you do not control and cannot forecast.

Zero yield is what converts time itself into a cost

A plot produces nothing. No rent, no dividend, no coupon, no distribution, not one rupee in any year you hold it. In a place where the census baseline is 2,779 people and there is no resident wave yet, there is also no rental market to argue about. This is the structural difference between land here and almost every other asset a first-time investor has held, and it is what makes horizon the central variable rather than a footnote.

Run the arithmetic in the only units available. That 5.9 percent of statutory entry cost, spread across a ten year hold, is about 0.59 percent of your capital consumed per year purely to stand still, in simple terms before any compounding. Across fifteen years it is about 0.39 percent a year. Those look like small numbers until you remember what sits underneath them: a zero. There is no income arriving to absorb them. Then add whatever local property tax or panchayat demand applies to your specific plot, which you obtain from the local office rather than from a brochure, plus the cost of periodically re-verifying your own records.

Then add the hurdle that nobody prints. Your money had an alternative use, and over ten or fifteen years the general price level moves as well. I will not put a figure on either, because the honest figure is the one that applies over your particular holding period and nobody has it yet, mine included. But the shape of the requirement is not in doubt. To have gained anything real, the land must clear the entry cost, clear the exit friction, clear inflation over the whole period, and then clear whatever your money would have done elsewhere. Only what is left after all four is a return. This is exactly why I refuse to print return expectations, and why the returns essay refuses too.

What seven to fifteen year money is actually required to believe

If the horizon is that long, then the purchase is not a bet on land at all. It is a bet on four propositions, and it is worth writing them down separately because they fail separately.

The first is that the anchor ramps. The Tata Electronics fab with PSMC was approved by the Union Cabinet on 29 February 2024, at Rs 91,000 crore, for up to 50,000 wafers a month on 300 mm, with reporting that it begins at the more mature nodes and moves to 28 nm later. First silicon is targeted around December 2026, and commercial production is reported for mid 2028. No chip has been produced yet. Nothing in the reporting I could find as of late August 2026 moves either date in either direction, so both stand as a target and a report rather than as facts.

The second is that employment converts into households. The full city plan projects roughly 800,000 jobs at maturity, with a published breakdown of about 312,900 direct and about 483,630 indirect. Maturity in that sentence means the 2040s. Jobs also do not automatically become residents, particularly now that the expressway makes it possible to work in the region and sleep in Ahmedabad.

The third is that a buyer exists when you want out, that they want the specific parcel you own, and that they can verify it quickly enough to keep their nerve. The fourth is that your paperwork survives contact with a careful lawyer a decade from now, in a market where the title chain, the survey number to final plot mapping and the encumbrance record are the whole game.

Notice the asymmetry. Three of those four are entirely outside your control, and the fourth is entirely inside it. That is the strongest argument I know for spending your effort at entry on paper rather than on price negotiation: check the scheme's registration number and status on the GUJRERA portal, and buy only against a clear title chain that places the land inside the notified region. Paper quality is the one input in this whole calculation that responds to how careful you are.

Three dated checkpoints worth re-underwriting on

A long horizon does not mean a decade of not looking. It means deciding in advance which dated events would make you revisit the whole case, and then actually revisiting it. Three qualify, and each comes with its own tier.

Airport operations first, targeted for September or October 2026, which is reported and is a target rather than a schedule. The runway is 3,200 m Code 4E, an Airports Authority trial and calibration landing took place on 4 June 2026 with aircraft VT-CNS, and the airport was reported roughly 80 percent complete in July 2026 with the terminal at about 75 percent. I went looking again on 25 August 2026 for anything official from the Airports Authority of India or the state that puts a date on commercial flights, and came back with nothing. That absence is the finding. Targets on this airport have slipped repeatedly since around 2010, including December 2025, so the event to re-underwrite on is a published commercial schedule, not another announcement.

The fab comes second. First silicon around December 2026 is a target; commercial production in mid 2028 is reported. Either milestone changes the character of the place, because a construction site becomes an employer with a payroll, a supply chain and a reason for people to need housing within driving distance. A slip of a quarter or two on a project of that scale is normal and tells you very little. A slip with no revised date attached tells you a great deal.

Rail third, and this is the one with the longest fuse and possibly the largest effect. The CCEA approved the Ahmedabad to Dholera semi high speed line on 13 May 2026, recorded in PIB release 2260624 at Rs 20,667 crore for about 134 km of double line connecting Ahmedabad, the region, the airport and Lothal, with completion targeted up to 2030 to 2031. The faster speed and journey time figures circulating in the press are not in that release, so I leave them out. Road moves individuals; rail is what lets a workforce live somewhere other than where it works, which is the mechanism by which a project like this stops being an industrial estate. Milestones to watch are the boring ones: tendering, land handover, contract awards. Those, plus the solar park's remaining 700 MW with its target of March 2027, are the calendar I would actually keep.

Writing your own horizon down before you sign anything

The discipline that separates an investor from a hopeful owner is one page of paper, written before the money moves and kept somewhere you will find it again. On that page: the date before which you will not sell, chosen against the phase window of the scheme your plot actually sits in rather than against a sales conversation. What you believe will have happened by then, in specific dated terms. The observable that would tell you the belief was wrong. And what you intend to do on the day you see it.

Date it and sign it. Five years from now the person reading it will be you, in an argument with a version of yourself who has already spent money and would very much like to be right. That page is the only witness you will have to what you actually believed at the start, and it is worth more than any forecast, because it makes your own thinking falsifiable. The dated items that come due soonest are collected in the 2027 watchlist, which is where I would start such a page today.

What the arithmetic can and cannot tell you

None of this predicts a price, and anyone doing this honestly has to say so plainly. The arithmetic on this page cannot tell you what land here will be worth in 2033. What it can tell you, exactly and without anybody's opinion in it, is the size of the bar the land must clear before your wait has earned anything: 5.9 percent gone at entry, an unpublished second layer on top of that, another 5.9 percent coming out of your future buyer's budget, nothing arriving in any year in between, and inflation quietly raising the bar the entire time.

Against that bar sits a plan whose own schedule runs into the 2040s and whose most testable historical forecast, the 2020 target, missed. The correct conclusion is not that the project fails. Trunk infrastructure is recorded complete, an expressway was reported inaugurated on 31 March 2026, a fab is being built and a rail line is approved. The correct conclusion is about the shape of money that belongs here. This asset suits capital that can sit for seven to fifteen years without being needed, without producing anything, without a reliable price to look at, held by somebody who checks three dated milestones a year and is honest with themselves when one of them moves. If your money cannot do all of that, the problem was never the horizon. It was that the horizon was never yours to choose.

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