Bracketing

Some differences resist measurement. A view, the quality of a renovation, a floor plan buyers prefer, a location factor the market has not produced a clean pair for: each matters, and each may have no pair of sales from which a figure can be read. Bracketing is the method for that case, and it is often underused.

Rather than assigning a figure to the difference, comparables are chosen on both sides of the subject for that feature. One is superior to the subject in that respect and one is inferior. The subject falls between them, and the two adjusted prices bound the subject’s position for that feature without a figure being assigned to it.

The subject has an average kitchen. One comparable has a full renovation and sold for more; another has the original kitchen and sold for less. Without measuring what the renovation is worth in that market, the analysis can state that the subject sits between those two sales, bounded above by the renovated house and below by the original one. The same construction works for a partial water view against a full view and no view, for a corner lot against a cul-de-sac and a through street, and for the house with the plan buyers like against the one with the plan they tolerate.

Reading a bracket with figures

The subject is a 1,950-square-foot colonial whose kitchen was last updated about fifteen years ago: serviceable, dated in finish, and not original. Three sales in the segment have been adjusted for time, living area and garage at measured rates. Sale A, with a full kitchen renovation completed the year before it sold, indicates $512,000 after those adjustments. Sale B, with its original kitchen, indicates $478,000. Sale C, with a kitchen of about the same age and finish as the subject’s, indicates $496,000.

No figure has been assigned to the kitchen, and the bracket still says a good deal. The subject’s kitchen is inferior to Sale A’s and superior to Sale B’s, so on the evidence of those two sales the subject sits below $512,000 and above $478,000. Sale C, which matches the subject on the kitchen, falls inside that bracket, and its position confirms the bracket rather than contradicting it. The $34,000 gap between A and B is the market’s response to the whole distance between an original kitchen and a full renovation in that segment, together with whatever smaller unmeasured differences the two houses carry.

The bracket also tests any figure proposed for the feature. If a figure of $45,000 had been assigned to the difference between Sale A’s renovated kitchen and the subject’s, Sale A would indicate $467,000, which is below Sale B’s $478,000 for a house whose kitchen is inferior to the subject’s. A figure that turns the bracket inside out is larger than the market in that segment supports, and the bracket shows it without any further data.

Bracketing and measured adjustments together

Bracketing and adjustment are used together. Bracketing sets the boundaries of the range for the features that cannot be measured; measured adjustments narrow the position within those boundaries for the features that can. An analysis that brackets condition and measures size, garage and time is stronger than one that assigns an unsupported figure to condition and calls it measured.

Bracketing as a selection discipline

Bracketing is also a selection discipline. When the comparables in an analysis all sit on one side of the subject for a feature, the range is open-ended on the other side, and the analysis cannot say where the subject stops. The search for a comparable on the missing side is worth more than a fourth comparable on the crowded one.

The check is made feature by feature across the whole set. For the colonial above:

Living area and the kitchen are bracketed. The garage is measured from pairs and needs no bracket. The finished lower level is not bracketed: every sale is inferior to the subject on that feature, so for that feature the three indications are floors, and nothing in the set marks where the subject stops. If the finished lower level cannot be measured from pairs in that segment, one sale with a finished lower level, even an older one carrying a market-conditions adjustment, closes the bracket and does more for the analysis than any other addition to the set.

When the missing side cannot be found

In a thin market, a sale on the missing side may not exist within the period or the area the search began with. The search is then widened in whichever direction yields sales the subject’s buyer would have considered: further back in time, with a market-conditions adjustment, or further out, into an area that competes for the same buyer. Where neither direction produces the sale, the analysis states that the bracket is open on that side and treats the indications as bounding the subject from one side only. Active and pending listings with the feature show where sellers of superior property are asking, which is useful guidance for positioning the price, but a list price is not a sale and does not close a bracket.

Condominiums, new construction and rural property

Bracketing is often the only workable method in three settings. Within a condominium complex, units of the same model differ by floor level, end or interior position, and outlook, and sales on a higher and a lower floor, or of an end unit and an interior unit, bound the subject’s position where no clean pair exists. In new construction, builders sell the same plan with different upgrade packages; a sale of the base package and a sale of the full package from the same builder and phase bracket a subject with a middle package. In rural markets, sites rarely match in acreage, frontage, topography and usable land, and a sale with more usable land and one with less bound the subject’s site more reliably than a per-acre figure drawn from dissimilar parcels.

Bracketing is also the clearest way to show a seller where the house sits: the renovated sale above it, the original-kitchen sale below it, and the subject’s own kitchen between them.

How it applies. A grid in which every comparable is superior to the subject in condition, or every comparable is larger, generally has a range with no floor or no ceiling for that feature, unless the feature was measured from pairs, as the garage was in the table above. The question for the grid is whether the subject is bracketed on the features that matter most in that market, and where it is not, whether a sale on the missing side exists.