Nothing quotes this asset. Its price comes from three negotiated sales that happened months ago, adjusted by somebody's judgement - and from an income lens that is more honest and much less popular.
11 min read · Level 2
The usual way to describe property pricing is that it is less efficient than a listed market, and I think that framing quietly flatters it. Inefficiency implies a mechanism that works imperfectly. What is actually happening is that there is no mechanism at all between transactions: a building is worth whatever the last similar one fetched, adjusted by somebody's opinion, until the next one completes. That is not a slower version of price discovery, it is a different arrangement in which price discovery happens a handful of times per decade per asset. The consequence people draw from this is usually that property is more stable, and it is the wrong conclusion drawn from a real observation. The observation is that the recorded value moves rarely. The conclusion should be that we do not know what happened in between - and the honest position is that the uncertainty is larger here than in a listed market, not smaller, because in a listed market you at least know what somebody would pay this morning. I would go further: the single most useful thing an investor can carry into this class is the habit of asking, of any property number they are shown, when the evidence underneath it was created. A valuation from eleven-month-old comparables and a live share price are not two versions of the same kind of statement, and treating them as comparable inputs to an allocation decision is how a portfolio ends up believing it is steadier than it is.
A listed share has a price because thousands of people are continuously offering to buy and sell it and the best of those offers is published. A building has none of that. There is no order book, no bid, no offer and no last traded price, because the thing has never traded except on the handful of occasions it changed hands.
So its price is constructed, by two methods that answer different questions.
Comparable sales. Find recent transactions in nearby similar buildings, express each as a price per unit of floor area, and apply the result to the subject. For Property P, at 80 square metres:
Comparable sold for area per sqm closed A 292,000 78 sqm 3,743.59 4 months ago B 315,000 84 sqm 3,750.00 7 months ago C 268,000 71 sqm 3,774.65 11 months ago ------------------------------------------------------ Mean per square metre 3,756.08 Applied to 80 sqm 300,486.34
That is where the 300,000 asking price comes from, and it is worth reading the right-hand column again. The most recent evidence is four months old. The oldest is nearly a year old. Three transactions, each one a private negotiation between two parties with their own reasons, is the entire market data set - and the answer, 300,486.34, is quoted to the nearest currency unit as though it were measured rather than estimated.
Capitalisation of income. Divide net operating income by a required rate. Property P produces 12,792, and at a 4.264% capitalisation rate that is exactly 300,000. The two methods agree here because they were built to, but they are not the same method: one asks what similar buildings fetched, the other asks what this income stream is worth.
An appraisal is a professional's reconciliation of those two, plus adjustments for everything the comparables did not match - a better kitchen, a worse aspect, a shorter lease, a car space. Each adjustment is a judgement, disclosed as a number.
And the buyer and the seller then negotiate, privately, and the agreed figure becomes tomorrow's comparable for the next building. The data set is made of its own outputs.
Because almost every intuition an investor brings from listed markets is wrong here, and three of them cost money.
The price is stale, and it is stale by an unknown amount. A share price is seconds old. A property valuation rests on evidence four to eleven months old, and nothing updates it in between. A building can be carried at 300,000 through a period in which every buyer in the market has quietly revised what they would pay, and the carried figure will not move until somebody nearby actually transacts. That is not an occasional failing; it is how the mechanism works, always.
The price is negotiated, so it contains the parties as well as the asset. A comparable that sold quickly because of a divorce, a probate or a relocation is in the data set at the price urgency produced. So is one that sold high because a neighbour wanted the adjoining plot. Nothing in the per-square-metre figure records why, and three transactions is a small enough sample that one of them being unusual moves the answer materially.
The spread is enormous and it is mostly invisible. A liquid share might cost 0.06% to cross. A property round trip costs about 6.0% of the price in transfer tax, legal fees, survey and agency - fifty times the 0.12% round trip a broad exchange-traded fund costs. That is not a transaction cost sitting beside the price. It is a gap between what a buyer pays and what a seller receives, which means "the price" is really two prices about six per cent apart, and the mid is a number at which nobody transacts.
Now the income lens, which is where the actual analysis lives. A capitalisation rate is net operating income divided by price, and it is the same quantity as the net rental yield from lesson two seen from the market's side rather than the buyer's. Rearranged, it prices the building:
Value = net operating income / capitalisation rate
At 4.00% 12,792 / 0.0400 = 319,800 At 4.264% 12,792 / 0.04264 = 300,000 At 5.00% 12,792 / 0.0500 = 255,840 At 6.00% 12,792 / 0.0600 = 213,200
The income never changed. Between the first row and the last, the building's value fell by 106,600, which is 33.33% of 319,800 - and it fell because of a number that describes what buyers require rather than what the building does. That single table is the most important thing in this lesson. A property owner who believes their asset is safe because the tenant is reliable has only secured the numerator.
Which also explains why this class can appear calm for years and then move a great deal at once. Nothing marks it in between. The capitalisation rate buyers require can drift for eighteen months with no transaction to reveal it, and the revelation arrives in a single comparable.
Educational, not personalised financial advice. Nothing in the Investor Academy is a recommendation to buy, sell or hold anything — what is right for you depends on your own circumstances and the rules where you live.
Two appraisals of Property P, done in the same week, by two competent people.
Appraiser one, leaning on comparables. Takes A, B and C, computes 3,756.08 per square metre, applies it to 80 square metres for 300,486.34, then adjusts: minus 6,000 because the subject has no parking where B did, plus 4,000 for a recent bathroom. Concludes 298,486.34, and calls it 298,500.
Appraiser two, leaning on income. Takes net operating income of 12,792 and asks what capitalisation rate the market requires for this kind of unit in this location. Judges 4.50%, which gives 12,792 / 0.045 = 284,266.67. Concludes 284,000.
A gap of 14,500 between two defensible answers on one building in one week, and neither of them is wrong. The first is an estimate of what similar things fetched; the second is an estimate of what this income is worth. Both rest on a judgement - the adjustments in the first case, the required rate in the second - and both are presented as a single figure.
Now watch the second appraiser's judgement do the work. At 4.264% the same income supports 300,000; at 4.50% it supports 284,266.67; at 5.00% it supports 255,840. The required rate moved by 0.736 of a percentage point across that range and the value moved by 44,160, which is 14.72% of 300,000. The income stayed at 12,792 throughout.
This is why an appraisal is a range presented as a point, and why why-valuation-is-a-range applies here more forcefully than anywhere else in the curriculum. In a listed market the range is collapsed by transactions every second. Here it is collapsed once, by two people signing a contract, and then reopened immediately.
The practical consequence is unglamorous and worth stating plainly. If you are buying, the comparables are the seller's argument and the capitalisation rate is yours, because the comparables describe a market and the rate describes what you need from this building. Arriving with only the first means negotiating inside somebody else's frame.
Treating an appraised value as a price. It is an estimate of what a transaction might achieve, produced from evidence months old, with adjustments that are judgements. A share price is a fact about a transaction that occurred; an appraisal is an opinion about one that has not.
Capitalising gross rent instead of net operating income. Property P's gross rent of 24,000 at a 4.264% rate gives 562,851.78 - an overstatement of 87.6% - and the arithmetic is perfectly correct throughout. This is the most expensive single slip available in property valuation.
Reading a small comparable set as a market. Three private negotiations, the most recent four months old, are not a price discovery mechanism in any sense a listed investor would recognise. One unusual transaction in a set of three moves the answer by several per cent.
Ignoring the capitalisation rate because the tenant is secure. The rate moved Property P's value from 319,800 to 213,200 without touching the income. A reliable tenant secures the numerator and says nothing at all about the denominator.
Mistaking the absence of marks for the absence of movement. Nothing revalues a building between transactions, so a period of no news looks like a period of no change. Lesson six is about what that does to every risk statistic quoted about this class.
Recorded sale prices are public in most jurisdictions. Pull the actual transactions rather than the asking prices, convert each to a price per unit of area, and write the completion date beside it. The dates are half the information: a set whose most recent entry is eleven months old is telling you the market is thin, which is itself a finding about how easily you will sell.
Comparables give you what similar buildings fetched; capitalised income gives you what this income stream is worth to a buyer requiring a given return. When the two disagree by more than a few per cent, that disagreement is the most useful thing you have found - it usually means either the comparables are not comparable or the income is unusual for the building type.
It is your number and it is the one you control. Anchoring it to the asking price is the whole of how a negotiation is lost. Decide what income return this building must produce given its location, its condition and what else you could do with the money, and only then divide the net operating income by it.
For Property P, 4.00% gives 319,800 and 6.00% gives 213,200 on unchanged income - a 33.33% difference. Doing this before you buy converts an abstract risk into a number you can hold in mind, and it is the clearest available demonstration that securing the tenant is only half of securing the asset.
When appraising any specific property, when a valuation you have been given needs testing, when negotiating a price, and whenever a carried value has not moved in a long time and you need to know whether that means stability or silence.
Do not use comparables where there are fewer than about three genuinely similar recent sales - the method needs a sample and three months of nothing is not one - and do not use a capitalisation rate borrowed from a different property type or a different location, because the rate is precisely where the differences between them live.
Recorded sale prices carry no reason. A transaction that completed below the local range because the seller had to be out by a date, and one that completed above it because the buyer owned the adjoining property and wanted both, sit in the data set as identical evidence. With only three or four genuinely similar recent sales, one such transaction moves the derived per-unit figure by a noticeable margin, and it moves it in a direction nobody can detect from the numbers alone. Agents and long-standing local valuers often know the stories; the public record never does. This is one of the few places where a conversation genuinely outperforms a spreadsheet, and it is worth having before relying on a small comparable set.
Ask what was happening in each comparable transaction, not just what it sold for. A small sample cannot average away a motivated seller.
Create a free account to read the rest and apply it to your own investing.