Real Estate's Low Volatility Is Partly an Illusion of Measurement

Real estate appears in asset comparisons as the steady line next to equity’s jagged one, and the gap is read as a difference in the assets. Part of it is a difference in how the two are measured.

What follows concerns appraisal-based return indices - the basis for most institutional property series - and why the volatility they report is a blend of the asset’s behaviour and the valuation process’s own inertia. The research base is US and UK commercial property; the mechanism is not jurisdiction-specific.

An Appraisal Is Not a Transaction

Most property return series are built from a valuer’s periodic estimate rather than a realised sale price. Appraisers anchor partly to previous valuations and partly to comparable sales that themselves lag the current market. Both anchors point backward, which mechanically smooths the series.

The effect has a precise statistical signature. Geltner, MacGregor and Schwann characterise appraisal smoothing as a temporal lag bias, and the resulting autocorrelation in reported returns is what makes long-horizon risk look small relative to short-horizon risk.

Unsmoothing Is a Standard Correction, Not a Fringe Objection

This is settled enough to have generated its own toolkit. The filtering procedures in common use derive from Geltner (1991, 1993) and Fisher, Geltner and Webb (1994), which recover an estimate of underlying market values from a smoothed valuation-based index.

The correction is material rather than cosmetic - large enough that unsmoothed series support different asset-allocation conclusions from the raw ones, which is the reason the literature exists at all. The procedures are themselves contested on the margins, with documented bias in the standard technique; what is not contested is the direction. Reported appraisal-based volatility understates the asset’s true variability.

The Indian Case Is the Mechanism Without the Measurement

India has no transaction-based residential or commercial property index built to the standard these studies rely on. That is worth stating plainly rather than borrowing a foreign statistic and presenting it as a local one.

What transfers is the mechanism, not a number. Wherever a reported price comes from periodic appraisal rather than continuous trading, the reported volatility measures the appraisal process alongside the asset - and Indian property valuation is, if anything, less frequent and less standardised than the markets the research was conducted in.

Illiquidity Is a Separate and Real Point

None of this argues real estate is secretly as volatile as equity. Illiquidity is a genuine reason property should carry a different risk profile from a stock that trades every second, and an asset that cannot be sold in a panic is not exposed to the same reflexive price dynamics.

The claim is narrower: some of the observed steadiness is a property of the asset, and some of it is a property of the measurement, and a single reported volatility figure does not separate them.

What This Means for Allocators

Do not put an appraisal-based volatility next to a transaction-based one in the same optimiser. The two numbers are not measuring comparable things, and the mean-variance output will systematically over-allocate to the smoothed series.

Check the autocorrelation of any property return series before using it. Positive serial correlation is the signature of smoothing, and it is visible directly in the data without needing the unsmoothing machinery.

And treat the low number as a floor on risk rather than an estimate of it. The asset is genuinely less volatile than equity. It is also less volatile on paper than it is in fact.