Real estate and FIRE: you pay for the walls, you invest in the land
"My home is my best investment." It is probably the most common belief, and the most poorly calibrated, among people building their financial independence. Poorly calibrated, because it treats as a single asset something that is, economically, two. And these two assets have opposite fates.
When you buy a home, you are really buying a consumable object, the structure, bolted onto a long option on a piece of territory, the land. One depreciates relentlessly and demands capital just to stay standing. The other never wears out and captures the value of everything built around it. Understanding this split is not an economist's curiosity. For anyone pursuing FIRE through geographic arbitrage, it is the lens for the entire strategy.
Two assets, two destinies
The breakdown is conceptually simple. A home comes down to a reproducible structure (the walls, the roof, the kitchen, the pipes) sitting on a non-reproducible plot. The structure is a piece of equipment: it ages like a car, goes out of fashion, and demands regular injections of capital, not to grow richer, but to avoid falling behind the market. Replacing a roof or changing the windows does not create wealth; it stops the property from sliding in relative value while the neighbours maintain theirs.
Land follows a completely different logic. A well-placed square metre never needs a new kitchen. Its value does not depend on its condition but on its scarcity: how many people want to live here, which jobs, which transport links, which schools, which level of safety, which climate are concentrated in this spot. This is the intuition David Ricardo formalised back in 1817 with land rent, and that Henry George radicalised in 1879: the value of a plot is not produced by its owner, but by the community around it. Land is, in a sense, a toll you collect on the positive externalities created by others. When a new transit line opens, when a company moves in, when a strong school is built nearby, it is your plot that cashes in, without you lifting a finger.
Schematic illustration, not to scale.
What the data says
The idea that "real estate always goes up" blends these two components and hides the real engine. Two independent reference studies point the same way.
In the United States, Davis and Heathcote (2007) built the first long-run indices separating land from structures. The result: between 1975 and 2006, the real price of residential land rose by a factor of 3.7, a gain of nearly 270 percent, while structures rose only 33 percent. The rise in housing did not come from the walls.
Source: Davis and Heathcote (2007), real prices.
On the international stage, Knoll, Schularick and Steger (2017) reconstructed property prices across fourteen advanced economies since 1870. Their conclusion: roughly 80 percent of the global housing boom since 1945 is explained by rising land prices alone, not by construction costs.
of the global rise in housing since 1945 comes from land, not from construction costs.
Knoll, Schularick and Steger (2017)
The lesson leaves no room for doubt: over the long run, the return on real estate is essentially a return on land. The structure, at best, is stable, at worst a capex sink.
Why this changes everything for FIRE
Here is where the geographic angle becomes decisive. Geographic arbitrage is, economically, land arbitrage. When you leave an expensive area for a cheap one, you barely arbitrage the price of the walls: a built square metre costs roughly the same everywhere, as materials and labour converge. What you arbitrage is the price of location, the rent the destination community charges for being there rather than elsewhere. Geographic arbitrage does not exploit construction costs; it exploits gaps in land rent between territories and between jurisdictions.
This reading has three direct consequences for anyone optimising the path to financial independence.
1. A primary residence is, in part, an anti-FIRE asset. FIRE means accumulating assets that compound and compressing consumption. Yet a home is hybrid: the structure portion is a consumption good that depreciates and that you rent to yourself, while only the land portion has any potential to compound. Believing you are investing when you buy your home ignores that a substantial fraction of the price, often the majority outside high-demand areas, goes into an asset that does not compound and bleeds maintenance.
2. The buy versus rent debate must be recalculated against your equity opportunity cost. When you rent, you pay for the use of structure and land, but you tie up no capital and carry no capex risk: the capital you would have sunk into bricks stays invested, compounding at around 6 to 7 percent real on a global equity ETF. When you buy, you get the land appreciation option, but you also pay for the depreciating structure, you lock up illiquid capital, and you absorb transaction costs, often several percent on a round trip. The decision then depends on three variables: the share of land in the price (the higher it is in a scarce area, the more buying makes sense), your equity opportunity return, and your holding horizon.
3. You are buying other people's externalities, in both directions. For a geographic arbitrageur, the fact that land value is created by the surroundings cuts both ways. Settling in a cheap, up-and-coming area can hand you free land appreciation if the community develops. But an area is often cheap precisely because those externalities do not exist yet, and nothing guarantees they will appear. Conversely, buying in an already established premium area means paying full price for externalities already capitalised: you buy land at its peak, with limited upside left.
The mobility paradox
This is where the most uncomfortable tension sits for a FIRE strategy built on geographic arbitrage, one no classic real estate analysis raises, because it only arises for someone who plans to move.
The asset that appreciates the most, land, is also the one that most sabotages the mobility that geographic arbitrage lives on. Owning a plot means anchoring yourself: illiquidity, high transaction costs, and above all lasting exposure to one jurisdiction, its tax regime and its political climate. Yet the whole power of geographic arbitrage rests on the ability to move to a cheaper area, or a lower-tax jurisdiction, when the time comes.
There is a trade-off here that deserves to be named: capturing long-run land rent means holding a fixed position, while geographic arbitrage means staying mobile. You cannot maximise both at once. For a path that plans several stages, for example a long residency phase in one country, then a later switch, this argues for anchoring in property only where the holding horizon is genuinely long and the area genuinely scarce, and for renting everywhere else, letting capital compound in liquid assets.
A tax nuance belongs in the calculation. In most systems the only component you can depreciate is the structure, the part that loses value; the land is not depreciable at all. Yet it is the land that creates the real wealth and that drives the taxable capital gain on resale. Depreciation and wealth creation therefore fall on the two opposite halves of the asset.
A decision grid before you sign
Rather than falling for a kitchen, take the price in hand and ask four questions.
Illustrative split, varies by local market.
- What share of the price is structure, and what share is land? In a slack area (where plots remain available), the structure makes up most of the price and value will track construction cost: little appreciation potential, the point is use. In a tight area it is the reverse, you mostly pay for scarcity.
- What bet are you really making? A studio in a city centre is a pure land bet. A well-placed parking space is land plus rental yield, with zero maintenance and zero emotional pull. A large isolated house is a bet on expensive structure dressed up as wealth. Filing all of these under "real estate" hides that they are radically different assets.
- Does buying beat my equity opportunity cost, net of maintenance and fees? If the rent you save (the use yield) does not exceed what the same capital would earn invested, after capex and the liquidity discount, then renting plus investing remains the FIRE-optimal default.
- How long will I really stay, and what does this anchoring cost me in mobility? The more your strategy relies on future geographic or tax switches, the more any land you own must be justified by a long horizon and strong scarcity.
The other side of the ledger: four forces the land breakdown underrates
Taken to its limit, the land argument pushes toward "rent and invest the rest." But that conclusion ignores four mechanisms that argue strongly for buying, above all your primary residence. None contradicts the land/structure split; all of them complete it.
1. Leverage: the only long, fixed, non-callable credit a private individual can get
An equity ETF returns 6 to 7 percent real, but with no accessible leverage. You cannot borrow against a stock portfolio on the same terms as a home: a securities-backed loan is callable, floating-rate and low loan-to-value. Property is the only asset where a private individual borrows 80 percent of the value over twenty or twenty-five years, at a fixed rate, secured by the asset, with no margin call. That is what turns a modest unleveraged return into a high return on equity.
The illustration is striking. On a €500,000 property funded with €100,000 of equity, a 3 percent rise on the total value is €15,000, which, against your €100,000 of equity, is +15 percent gross on capital deployed, before interest. Leverage creates value as long as the property's total return (appreciation plus net imputed rent) beats the cost of the loan. But the symmetry is merciless: a 3 percent fall in the property is minus 15 percent on your equity. Leverage amplifies both ways, and land can fall.
The specifically FIRE point is often forgotten: leverage requires income to obtain it. A bank lends against a salary, not against a portfolio withdrawal rate. The window to borrow is therefore the salaried accumulation phase. Once in early retirement, credit is no longer available on the same terms. If leverage is part of the plan, it has to be set up before you press the FIRE button; the fixed-rate loan then runs into early retirement, and beyond.
2. Tax treatment: the primary-residence exemption changes the calculation
Land generates a taxable capital gain on resale, with one major exception in most tax systems: the home you actually live in. Many jurisdictions exempt the primary residence from capital-gains tax, fully or partially, while a rental or second property is taxed, sometimes only tapering to exemption after decades of ownership. On top of that, the imputed rent of an owner-occupied home is generally not taxed.
For anyone steering their taxable income to preserve means-tested benefits or stay under a threshold, this exemption is doubly valuable: the gain simply does not enter taxable income, so selling your home does not blow up your tax picture in the year you sell, whereas a taxable rental gain can. Note the tension with the mobility paradox: the tax code tends to subsidise precisely the asset that anchors you (your home), partly offsetting its mobility cost, but only if the property genuinely is your primary residence when you sell. Exact rules vary by country, so check your local regime.
3. Inflation hedge: a fixed-rate loan freezes your housing cost
A long fixed-rate loan freezes the nominal cost of housing. Inflation then erodes the real value of the debt and of the monthly payment. The renter, by contrast, faces rent re-indexed every year (annual rent indexation); over twenty-five or thirty years, cumulative rent inflation is considerable. A fixed-rate loan is, in effect, a short position on inflation, handed cheaply to the individual: inflation becomes the borrower's friend.
The stake is central in FIRE. In early retirement you may live forty years on a fixed income or one driven by a withdrawal rate; housing inflation over several decades is one of the biggest threats to the plan. Owning, a paid-off home or a fixed-rate loan, takes the most inflation-exposed expense line out of the equation.
A note on framing: this freeze covers the financing and the location rent, not maintenance (capex stays exposed to construction-cost inflation) nor property tax (often rising sharply). So it is the cost of credit that is frozen, not the total cost of occupancy. It still neutralises the heaviest and most systematically re-indexed expense.
4. Behavioural robustness: a paid-off roof lowers the withdrawal rate
A purely mathematical analysis forgets that a FIRE plan only succeeds if you stick to it. A paid-off home lowers fixed costs and therefore the withdrawal rate you need, which directly improves the portfolio's survival probability, and lets you ride out a crash without panic-selling, because the home is locked in. The "security" of a paid-off roof is not just comfort: it has a measurable value in withdrawal-rate terms, through the drop in mandatory cash needs.
The flip side to keep in mind is concentration: a paid-off home locks a huge share of net worth into a single, illiquid, undiversified asset tied to one jurisdiction, which brings back the mobility paradox. The same security can, in theory, be reached with a larger liquid cushion; but many investors do not behave as if liquid wealth were safe, and behaviour is what decides whether the plan survives. Worth counting as a real parameter, then, but without hiding its cost in diversification.
The synthesis
These four forces do not refute the land breakdown: they sharpen the distinction the article already drew. The breakdown is the right weapon against treating your home as an investment, and against rental or secondary property, where you get neither the primary-residence exemption nor any escape from the taxable land gain. But for the primary residence specifically, accessible leverage, the tax exemption, the inflation hedge and behavioural robustness combine into a radically different, and often winning, calculation. A primary residence and a rental investment are not the same decision, which is exactly why filing everything under "real estate" misleads.
In short
Real estate is not an asset: it is a depreciating consumable, welded to an option on a territory that captures the value created by others. The long-run data, American and international, all show that the return comes from the ground, not the walls. This split remains the decisive weapon against the idea that a primary residence is an "investment", and against the illusory returns of many rental projects.
But it does not settle the buy versus rent question on its own. Four forces, long fixed-rate bank leverage, the primary-residence capital-gains exemption, the hedge against rent inflation, and the behavioural robustness of a paid-off roof, strongly rehabilitate buying your primary residence, above all for a FIRE path exposed to several decades of inflation and sequence-of-returns risk. The discipline that follows is therefore not "never buy", but: separate the investment from the use, pay for property only where the land share is scarce and the horizon truly long, borrow while you still have the income to do so, and everywhere else keep your capital liquid and your mobility intact, because mobility remains the engine that buying jams.
Sources
- Morris A. Davis, Jonathan Heathcote, "The Price and Quantity of Residential Land in the United States", Journal of Monetary Economics, vol. 54, no. 8, 2007.
- Katharina Knoll, Moritz Schularick, Thomas Steger, "No Price Like Home: Global House Prices, 1870-2012", American Economic Review, vol. 107, no. 2, 2017.
- David Ricardo, On the Principles of Political Economy and Taxation, 1817 (land rent).
- Henry George, Progress and Poverty, 1879 (collective capture of land value).