In this lesson · 8 sections
The phrase "real asset" is used so loosely that it has almost stopped meaning anything. It gets applied to a flat in Wiedikon, a bar of gold in a Zurich vault, an oil futures contract, a toll road in France, a warehouse near Basel and, increasingly, to things that are neither real nor assets. If the category is going to be useful — and it is — it needs a definition that does work rather than one that just gestures at physicality.
Here is the one that holds up. An asset is real when its value is anchored to the price of a physical thing that has to be produced, replaced or extracted, rather than to a promise denominated in a currency. A bond is a claim on a fixed number of francs. A share is a claim on a business's residual profits, which are themselves currency amounts, though the business may have real assets behind them. A building is a claim on the use of a specific piece of ground and the structure on it — and if the general price level doubles, the cost of building that structure again doubles too. That anchoring to replacement cost, not the fact that you can kick it, is what makes an asset real.
The definition that actually does work
Test the definition against edge cases and you can see where the boundary sits.
A residential building is real: rents are set in a competitive market against tenants' nominal incomes, and the cost of constructing a substitute rises with wages and materials. A thirty-year fixed-rate mortgage on that building is the opposite of real — it is a nominal liability whose burden shrinks when prices rise. This is why leveraged property is such a potent inflation position: you are long a real asset and short a nominal one at the same time.
Gold is real in a narrower sense. It produces nothing, its industrial demand is small relative to the above-ground stock, and its price is not tethered to any production cost in the way copper's is. What anchors gold is the willingness of people to hold it as money-that-nobody-issues. That makes it a monetary asset wearing a physical costume, which is why the sibling lesson on gold treats it separately rather than lumping it in with commodities.
A commodity futures contract is not the commodity. Buying a barrel of oil and buying an oil future are economically different positions, and the difference — the shape of the futures curve — is the single largest reason retail commodity products underperform the headline spot price they advertise. That gets its own lesson too.
Timberland, farmland, infrastructure and shipping sit clearly inside the category: physical, productive, cash-generating, replacement-cost-anchored. Cryptoassets sit clearly outside it, whatever the marketing says: their scarcity is a rule in software, not a production constraint, and no cost curve pins the price.
Three sources of return, and only three
Every real asset pays you through some combination of three channels. Being explicit about which channel you are relying on is most of the analytical work.
Income. Net cash the asset throws off after the costs of keeping it operating — rent minus maintenance, tolls minus upkeep, crops minus inputs. This is the only source that arrives whether or not anyone else wants to buy the asset from you.
Real growth in that income. Rents that rise with wages, tariffs indexed to consumer prices, yields per hectare that improve with technique. This channel is what separates a productive real asset from a static one.
Repricing. The multiple the market applies to that income — the capitalisation rate for property, the multiple for infrastructure. Repricing is dominated by real interest rates, and it is the channel that produces almost all of the short-run pain and pleasure.
Gold and industrial commodities have no first or second channel at all. Held physically, they generate no income; they cost money to store and insure. All of their return is channel three, plus, in the case of futures-based exposure, a collateral return on the cash you post and a roll return that can be positive or negative. If someone tells you gold has an expected real return, ask them where the cash flow is. There isn't one, and that is not a fatal objection — but it changes entirely what the position is for.
Replacement cost is the actual inflation mechanism
"Real assets hedge inflation" is a slogan. The mechanism underneath it is replacement cost, and it explains both when the hedge works and when it fails.
Consider a warehouse that would cost CHF 20 million to build today: CHF 5 million of land, CHF 15 million of structure. If wages and materials rise 30% over a decade, rebuilding the structure now costs CHF 19.5 million. No developer will build a competing warehouse and let it out at a rent that fails to service CHF 24.5 million of total cost. So new supply stops until market rents rise enough to justify the new, higher construction cost — and the existing building's rent gets dragged up with them. The hedge is delivered by the supply response, not by magic.
That framing tells you immediately where the hedge is strong and where it is weak. It is strong where supply is genuinely constrained and construction costs are a large share of value: Swiss residential property in the lake basin, city-centre logistics, regulated infrastructure with indexed tariffs. It is weak where supply can arrive fast (secondary office space, commodity production with idle capacity), where the income is contractually fixed for a long period (a twenty-year lease with no indexation is a bond with a building attached), and where the asset's value is mostly land in a place that is not getting more desirable.
It also tells you the hedge is slow. Leases reset on renewal, construction responds over years, tariffs index annually with a lag. Over one or two years, a real asset's price is driven by real interest rates, which usually rise when inflation surprises upward. The hedge shows up over five and ten years, not over one — and the sibling lesson on inflation hedging quantifies which assets have actually delivered it over which horizons.
Expected versus unexpected inflation
The distinction that resolves most confusion is between inflation you already knew about and inflation you did not.
Expected inflation is already in the price of everything. If the market expects 2% inflation, nominal bond yields, property cap rates and equity discount rates all embed it. Owning a real asset gives you no advantage against inflation that everyone has already priced.
Unexpected inflation is what a hedge is for, and it does two opposing things to a real asset at once. It raises the future nominal cash flows — helpful. It usually also raises real interest rates, because central banks respond by tightening — harmful, because the discount rate applied to those cash flows rises. Which effect dominates depends on the asset's duration and how quickly its income actually resets.
That is why an inflation shock can hurt a long-lease office block and help a self-storage business with monthly pricing, even though both are "real estate". Reset frequency is the variable that matters, and it is knowable in advance from the lease terms.
The costs, and the measurement illusion
Real assets carry frictions that financial assets do not, and honest accounting for them changes the answer more often than people expect.
Transaction costs on direct property run to several percent of value on each side once notary, land-registry, agency and, in most cantons, transfer taxes are counted. Holding costs are continuous: maintenance and capital renewal typically consume around 1% of a building's value per year, storage and insurance eat into a physical metal position, and an oil futures position pays a roll cost when the curve is upward-sloping. None of these appear in a headline price index.
Then there is the illusion. Direct property and unlisted funds are valued by appraisal, usually quarterly or annually, and appraisers anchor on the last valuation and on stale comparable transactions. The result is a reported return series that is smoothed: measured volatility looks low and correlation with equities looks low, and neither is a property of the asset — both are properties of the measurement. Listed real estate holding nearly identical buildings shows several times the reported volatility because it is priced continuously by people who can sell. The buildings are the same. Only the honesty of the mark differs. Illiquidity is a genuine cost being reported as a risk reduction.
The Swiss angle
Three things make the Swiss case distinctive, and they run through the whole course.
First, Swiss inflation has been structurally lower than in most peer economies for decades, so the domestic case for an inflation hedge is weaker than the imported literature — mostly American — assumes. A hedge against 2% is worth less than a hedge against 6%.
Second, the franc has tended to appreciate against the dollar and the euro over long periods. Commodities and gold are quoted in dollars, so an unhedged Swiss investor holding them is also running a short-franc position. Over multi-decade windows that currency leg has subtracted meaningfully from franc-denominated returns on dollar-priced commodities. It does not make the exposure pointless, but you must decide whether you want the currency risk that comes bundled with it.
Third, the Swiss tax code treats real assets differently from securities. Capital gains on privately held movable assets — shares, funds, gold — are generally free of income tax for private investors, while gains on real property are taxed by the cantons through the property gains tax, and property is subject to annual cantonal wealth tax on its tax value. Those asymmetries are large enough to change the ranking of otherwise similar exposures, and they are taken apart in the Swiss property lessons that follow.
What to take away
- An asset is real when its value is anchored to the replacement cost of a physical thing, not to a currency-denominated promise; this is why leveraged property, long a real asset and short a nominal debt, is the sharpest inflation position most households will ever hold.
- Return on a real asset comes from income, real income growth and repricing — gold and physical commodities have only the third, so their entire expected return depends on someone paying more later.
- The inflation hedge is delivered by the supply response to rising construction costs, which means it works where supply is constrained and income resets often, and fails where leases are long and fixed.
- Unexpected inflation raises both future cash flows and real discount rates, so real assets frequently fall in the first year of an inflation shock and only earn their keep over five to ten years.
- Appraisal-based valuation makes unlisted property look far less volatile than it is; low reported volatility from infrequent marks is a measurement artefact, not risk reduction.
Sources
- Òscar Jordà, Katharina Knoll, Dmitry Kuvshinov, Moritz Schularick and Alan M. Taylor, "The Rate of Return on Everything, 1870–2015", Quarterly Journal of Economics, 2019.
- Katharina Knoll, Moritz Schularick and Thomas Steger, "No Price Like Home: Global House Prices, 1870–2012", American Economic Review, 2017.
- Piet M. A. Eichholtz, "A Long Run House Price Index: The Herengracht Index, 1628–1973", Real Estate Economics, 1997.
- Claude B. Erb and Campbell R. Harvey, "The Golden Dilemma", Financial Analysts Journal, 2013.
- Gary Gorton and K. Geert Rouwenhorst, "Facts and Fantasies about Commodity Futures", Financial Analysts Journal, 2006.
- David M. Geltner, Norman G. Miller, Jim Clayton and Piet Eichholtz, Commercial Real Estate Analysis and Investments — on appraisal smoothing and unsmoothing.
- Bank for International Settlements, residential property price statistics and accompanying methodology notes.
- Swiss National Bank data portal and the Federal Statistical Office (BFS/OFS) for Swiss price, rent and construction cost series.