How we compare performance

Each vehicle states a target return when it opens. We compare that with what the vehicle has actually paid over the trailing twelve months, divided by the capital subscribed to it: its register-level trailing yield.

Capital subscribed is the sum of settled subscriptions in the vehicle, before the platform fee. Income is cash actually distributed and settled on the register, not accrued, declared or expected. Nobody’s individual holding is read to produce it, which is what makes every figure on this page publishable to a signed-out visitor.

Signed in, your portfolio runs the same comparison against your position instead: income received on it over the capital you subscribed or paid for units, on the same before-fee basis. Both come from the same double-entry ledger that produces your statements, so the comparison and the statement can never disagree.

The median trailing yield of vehicles in the same category on this register: office against office, industrial against industrial. It is computed from each vehicle’s own distributions over the capital subscribed to it, never from anybody’s individual position, so no investor’s holdings are visible to another.

A median, not an average: on a book this size one exceptional quarter would move an average and tell you nothing about the typical vehicle. And no median is shown at all below three comparable vehicles, because two numbers do not have a middle.

Four cases produce no figure rather than a misleading one.

• Held under a year. A trailing-twelve-month yield needs twelve months. Annualising a partial year would flatter or punish a position for nothing but the date it started.
• Vehicles that have wound up. An exited or closed vehicle stopped paying because it repaid, and any payments left inside the trailing window are a fraction of a year read against a full-year target. Its outcome belongs to a different measure, the realised multiple at exit, not to a running yield it no longer has.
• Vehicles that repay on completion. A development note pays nothing until it completes, by design. Showing 0% against a 16.5% target would read as catastrophic underperformance on an instrument behaving exactly as written.
• Fewer than three comparable vehicles. Stated as “too few to compare” rather than dressed up as a median.

Every vehicle on the public book, its stated target, and what it has actually paid over the trailing twelve months, computed exactly as described above, from the same ledger that produces investor statements. Where a figure is refused, the reason is stated in its place.

Measured over the twelve months to 12 September 2026. The window opens 12 September 2025.

We do not plot these vehicles against a listed property index, and this is a deliberate omission rather than an unbuilt feature.

A listed REIT reprices every second on what buyers and sellers will pay. These vehicles are re-marked two to six times a year by an independent valuer. Draw those two on one chart and the register looks calm and the index looks violent, but that difference is an artefact of how often each is measured, not of the underlying risk. Property valued twice a year is not less volatile than property valued daily; the movement simply is not observed. Reading a smooth line as safety is the single most expensive mistake available in this asset class, and we will not draw the chart that invites it.

The same reasoning rules out candlesticks, intraday ranges and a live ticker on these pages. There is no continuous price to draw, so drawing one would be an illustration, not a measurement.

Past distributions are not a forecast, a target is not a promise, and nothing on this page is advice about whether to invest. RWA oa is not licensed to give it. See the risk warning for what can go wrong with these investments, which is the other half of any honest comparison.