Incompetence Corrupts...

and just a little incompetence can corrupt absolutely.

Lord Acton may not have considered the above twist, but my new analysis of market areas in the City of Akron is raising some significant questions in my mind about the whole process of determining a "market value" for residential property in some markets. Conventional practice relies upon the Sales Comparison Approach to provide a means of forming an opinion of value. The data I am finding, however, seems to imply that sales comparison may be entirely inappropriate in some markets, and the insistence by lenders on the use of the technique may be an important contributor to the current difficulties in the mortgage lending arena.

Most appraisers are familiar with the "big three" approaches to deriving a value opinion : cost, income, and sales comparison. In residential mortgage lending, underwriting doctrine insists that sales comparison is most appropriate. Lenders (as well as most appraisers) have believed that by comparing a subject property to the market experience of other nearby homes of the same age, size, type, and condition, it is possible to predict the reaction of "typical" buyers to an attempt to market the subject. What has been ignored are the basic assumptions that must be present in developing an opinion regarding "what if we tried to sell it?".

The market area I was researching was the Summit Lake Neighborhood in the City of Akron. The city's official neighborhood data sheet for that neighborhood gives a year 2000 owner-occupancy statistic for "South Akron" as 57.3%, but also has a note that "Slightly over one-third of housing units are owner-occupied, significantly less than 55% in the City." My research from the Census 2000 data shows an average tenant-occupancy rate in the Summit Lake Neighborhood of about 70%. Consequently, the most probable buyer for a single-family home in that neighborhood, at any given time, would be an investor.

In examining the statistics from brokered MLS sales and the public records data for all sales in that area, a serious discrepancy was evident. The mean sales price for brokered sales declined from $27,167 in 2001 to $13,741 in 2007, with the median declining from $24,000 to $10,000. In complete contrast, the mean for all sales increased from $44,347 in 2001 to $63, 509 in 2006, and the median for all sales increased from $42,700 to $70,000.

What is not evident from the raw stats is the fact that the MLS sales are mostly REO sales -- the banks have been hiring agents to dispose of their foreclosed properties -- whereas the public records stats only include the final transaction price for a given home at the end of the calendar year. The private sales by the investors, who are dealing both among themselves and also with owner-occupant buyers, involve numerous "flip" sales where the REO homes have been purchased, "rehabbed", and resold. There is no ready information with regard to marketing times or seller concessions and incentives. Further, if one looks closely at the major players in the high-end sales, one finds that at the end of 2007, a number of them were either under indictment or already in jail for mortgage fraud.

Even if all the transactions taking place were honest deals, there is still a problem with trying to determine what the typical buyer and seller would be willing to agree on in this market. An owner-occupant is usually looking for a house that is ready to move in to. An investor who wishes to be a landlord will be looking for a home that will be able to generate an income stream with minimal repairs and maintenance. The investor who is rehabbing to resell will be picking up the sales at the bottom of the pile. The "market value" of the home should not differ much between each category of buyer, since open market competition for the property should drive the final sales price. However, the appropriate method of arriving at the "market value" may be something other than simple sales comparison as anticipated by the standard appraisal forms.

Out-of-town appraisers, who may be relying on public records sales data without understanding the owner/tenant composition of the neighborhood, may be providing value opinions that do not really meet the needs of the intended users or their intended use. Users of AVMs in such a neighborhood run the risk of developing value estimates completely out of line with the reality of the marketplace. This situation warrants further exploration.