Modernize comps? Silly do. Everybody knows what’s a good comp!
Or do they?
This simple claim forms the major dividing line between the old way, and the new way – the legacy “trust me” appraisal paradigm – and all other forms of valuation and risk assessment.
This Modernizing Appraisal series is about the analysis process, not FannieMae and FreddieMac reporting requirements. The process is identical: 1) The problem; 2) The data; 3) Predict or adjust; 4) Communicate. We continue here on #1 – defining the problem to be solved. Read the earlier parts of this series here.
Recall that historically, appraisal was created for sparse, difficult to verify, subjective, and singular sales information. Who you knew was as important as what you knew (as an appraiser).
Collecting comps was a personal activity. The goal was to collect a handful. Five or six was “adequate.” For residential, three was the magic number. For experienced professionals, personal credibility was/is the goal. The historical focus of USPAP is to be credible — “worthy of belief.”
Today, the goal is better than being worthy. The goal is reliable results. Measurable results.
This is the ninth of the series on the analysis process, (not reporting process). Here we start the second of the four components of valuation. These components are exactly the same for appraisals, valuations, evaluations, hybrids, AVMs, BPOs, waivers, or “inspection plus” alternatives..
- Problem delineation, hypothesis, scope of work;
- Data reduction to relevant information;
- Adjustments and predictives;
- Delivery interface.
This four-part delineation is true for all types of valuations.
The idea of modernizing data delineation is simple. But it is the key to evolving from “worthy opinion” to reliably-measurable results. This is the essence of data science. It is critical to the proper prompting using artificial intelligence. We call it EBV, Evidence Based Valuation©. A result, not opinion.
In common legacy practice it is important that we “support” that opinion.
In data science, it is universally known that 80% of the analysis means get the right data! This is more so true for the application of Ai through data science — for valuation and risk questions.
Here, in this blog series, we proceed from #1: problem ID and scope of work, to #2 covering the theory/models of data selection. This summarizes the analysis and mechanics which are the focus of the Valuemetrics.info classes, starting with the Stats, Graphs, and Data Science 1 class.
We consider such issues as:
- What is a comp?
- How do we identify “similarity”?
- How many sales do I need for these methods?
- What is the role of expert judgment in data selection?
- Where does expert intelligence work with artificial intelligence?
- What new competencies are needed for the professional asset analyst?
It appears that the need for traditional “form-filler” appraisers is declining rapidly, and will soon disappear. Step by step. The time to train and start is now.