Do Real-World Assets Create Real-World Markets? A real(T) case analysis
Jean-Baptiste Pleynet
Nathaniel Pleynet
Illiquid, non-fungible assets like real estate rely on periodic valuations to anchor prices. Between updates, the market must set prices based on expectations rather than fresh appraisals.
In this article we test whether valuation age—the time since a property’s last valuation—shows up in YAM trades. Do price premiums drift upward as the last valuation gets older, and do net yields compress as a result?
This follows last week’s note on how pricing relates to time-to-execution; here we focus on how time since valuation relates to the prices investors actually pay.
RealT does not publish a fixed calendar for property revaluations. Updates are made as needed (e.g., market moves, tax reassessments) and communicated when effective via email, Community Calls, and platform updates. A revaluation does not change scheduled rents; it changes the reference property value, which mechanically shifts the catalog net yield (rent ÷ value).
Our goal in this note is to see how the YAM market prices periods between official updates—i.e., whether traders anticipate revaluations and incorporate latent appreciation into executed prices.
Valuation age (days): time between the last official valuation and the trade’s block timestamp (UTC).
Token NAV at last valuation: property valuation at the last update ÷ total tokens outstanding.
Price premium (%): (trade price per token ÷ token NAV at last valuation) − 1.
Net yield at trade: annualized net rent (after management fees, taxes, vacancies, RealT fee) ÷ property valuation at the trade block. As defined in articles 2.
Catalog net yield: the platform’s reference net yield using the most recent valuation and rent.
Yield premium (percentage points): net yield at trade − catalog net yield.
Scope. Executed equity trades only (no loan tokens). “Cruise-speed” window: 1 Jan 2024 – 30 Jun 2025 (As defined in articles 1). “Stress” window: May–Aug 2025 (As defined in article 3).
Weighting & hygiene. Charts and regressions are size-weighted by traded notional (USD-stablecoin equivalent). Extreme outliers are winsorized.
Estimation (baseline). We fit a simple linear model: Premium = alpha + beta × ValuationAge + error (no entity or time fixed effects in the baseline; robust standard errors clustered by property).
The fitting takes into account trade weights.
This is a first-pass estimate; a fuller model with city/month fixed effects and other controls will follow later in the series.
Plotting price premium (trade price vs. token NAV at the last valuation) against valuation age shows a clear upward drift: the longer a property goes without a fresh valuation, the higher the price investors are willing to pay. In other words, the market appears to anticipate the next valuation update.
A size-weighted linear fit on the cruise-speed sample points to a slope of ~4.31% per year (≈ 0.36% per month). Practically, that’s about +2.1% after six months without a valuation and +4.3% after twelve months.
This is a first-pass relationship, not proof of causation. City mix, property characteristics, and timing effects can all matter; we’ll explore those controls in later notes.
Detroit matters here because it accounts for the largest share of YAM trading, as we saw in a previous article. When we run the same “price premium vs. valuation age” fit on Detroit-only trades, we get a slope of ≈ 4.06% per year—very close to the all-market estimate (≈ 4.31%/yr). In practical terms, that’s roughly +2.0% after six months without a valuation.
This is consistent with the idea that traders anticipate revaluations and price in some capital appreciation between official updates. It also makes sense mechanically: Detroit’s large weight in the dataset pulls the aggregate slope toward the Detroit number.
A note on context: public indices reported strong Midwest house-price gains in 2024 while the national market slowed under high rates. Detroit outperformed, with Cleveland solidly positive and Chicago more moderate.
Caveat. RealT’s Detroit properties are not the entire Detroit market. Differences in property type, renovation status, rent policy, token float, and whitelist density can all affect how quickly prices drift between valuations.
We’ll dig into city-by-city slopes in the next article and test whether these patterns hold once we control for city and time effects.
A natural counter-hypothesis is that rising rents between two valuation updates could explain higher token prices with no real change in investor return. If rent goes up, a steady yield would mechanically justify a higher price.
That is not what we see on YAM. When we plot the yield premium (net yield at trade minus the catalog net yield) against valuation age, the relationship slopes downward: as the last valuation gets older, executed trades clear at lower net yields than the catalog figure. In other words, prices appear to rise faster than rents in the interim, so investors accept a yield give-up—consistent with an expectation of capital appreciation at the next update.
Practically, the pattern extends below zero: beyond a certain valuation age, the median yield premium turns negative, meaning the market’s expected net yield dips below the catalog yield. That can happen either because (a) token prices get bid up relative to the last valuation while rents are largely unchanged, or (b) traders are explicitly willing to trade some near-term yield for a shot at realizing a capital gain when the property is revalued. Notes. Results are size-weighted and conditional on execution. Rent updates can occur between valuations (lease turnovers, adjustments).
In the July–August 2025 stress window, volumes spiked and headline yields moved up. The question here is whether the valuation-age effect survived—do traders still price how long it’s been since the last valuation?
Short answer: yes, but the gradient softens.
On prices, the size-weighted fit of price premium vs valuation age drops from ~4.31% per year in cruise speed to ~3.33% per year in the stress window. That’s roughly +1.7% at six months without a fresh valuation (vs +2.1% in cruise) and +3.3% at twelve months (vs +4.3%). Fundamentals remain in play; the market still anticipates revaluations—just less aggressively.
On yields, the yield premium (net yield at trade minus catalog net yield) stays negative with valuation age—the curve slopes downward in stress as well. The relationship is flatter, and the level is higher (overall yields were elevated during stress), but the pattern holds: as valuations get older, trades clear at lower net yields relative to catalog, consistent with investors pricing in some capital appreciation ahead of the update.
Interpretation. Even under pressure, participants didn’t forget fundamentals. They bought faster and competed for bargains, yet they still priced valuation recency—only with a milder drift. Put differently, stress changed the level (higher yields overall) more than the logic of how valuation age feeds into price and yield.
Valuation age matters. As time since the last valuation grows, prices run ahead of rent. At cruise speed the priced-in drift is about +4.3%/year (~0.36%/month); in the July–Aug 2025 stress window it softened to ~3.3%/year.
Yield compresses. The yield premium (trade net yield − catalog net yield) falls with valuation age, often turning negative—evidence that investors trade some near-term income for anticipated capital gains at the next update.
How to use this. Treat ~0.36%/month as a yardstick. If you believe a property’s next revaluation will exceed that drift, late-cycle buys can make sense; if you’re income-first, prefer younger valuation age or neutral/positive yield premium.
Caveats. Executed equity trades only; city/property mix, rent timing, whitelist frictions, and selection bias mean RealT’s pool isn’t the whole market.
Next up: a city-by-city cut. We will see what a focused analysis of cities like Detroit, Cleveland, Chicago, Toledo can tell us, and how local fundamentals line up with what YAM is already pricing.
We strive to ensure that our work remains objective and grounded in data. In the interest of full transparency, we disclose that we hold investments in RealT assets and, more broadly, recognize the company’s approach as a noteworthy and innovative contribution to the real world asset sector.
Preliminary results are shared with the RealT team one week prior to publication, solely for the purpose of identifying potential bugs or inaccuracies in the data presented. The RealT team does not participate in the analysis, nor do they influence the methodology or conclusions in any way.
This study is conducted independently and is entirely self-funded.
This research draws partly on the free, open-source analytics suite maintained by the RealToken Community.
