Valuing a $500,000 condominium is a fundamentally different exercise from valuing a $25 million estate on Further Lane. At the mass-market level, abundant comparable sales, standardized unit types, and well-established price-per-square-foot benchmarks make valuation relatively straightforward. At the ultra-luxury level, every property is effectively unique, and the "comparable" in comparative market analysis becomes a term of art rather than precision.
Our approach combines quantitative modeling with qualitative assessment in a framework we call structured subjectivity. The quantitative layer draws on transaction data, listing histories, and market trends across both our markets, adjusted for the specific attributes of each property. The qualitative layer incorporates factors that resist quantification: architectural provenance, view quality, neighborhood trajectory, and the intangible quality of prestige.
Technology plays an increasingly important role in this process, but as a tool rather than a substitute for judgment. Machine learning models can identify patterns in pricing data that human analysis might miss — for example, the premium associated with corner positions in specific co-op buildings, or the discount trajectory of properties that have been on market beyond a certain threshold. But these insights are starting points for analysis, not conclusions.
The most valuable output of a rigorous CMA is not a single number but a range, accompanied by a narrative that explains the factors supporting the upper and lower bounds. A client who understands why a property might trade between $18 million and $22 million — and what market conditions or presentation improvements might push toward the upper bound — is better equipped to make strategic decisions than one who receives a point estimate.