Evolution of methods: statistical tools for the appraiser

Woody Fincham, Research Fellow

I was looking for some demographics in regards to who makes up our industry. I ended up over at the website for the Bureau of Labor Statistics. I was looking through the site at what they have predicted for the real estate industry as of 2004, the last year for which comprehensive data is available. Some of the stats may interest you:

Another interesting narrative I found in the report (emphasis added):

"...employment will be held down to a certain extent by productivity increases brought about by the increased use of computers and other technologies, which make for faster valuations and allow appraisers to take on more customers and each assessor to assess more properties. In addition to growth openings, there should be numerous openings due to the need to replace the many appraisers and assessors who are expected to retire or decrease their working hours over the projection period."

"The cyclical nature of the real estate market will also have a large effect on the future of appraisers, especially those who appraise residential properties. In times of recession, fewer people buy or sell real estate, causing a decrease in the demand for appraisers. However, during a downturn in the residential real estate market appraisers often are able to switch specialties and appraise other types of properties."

"Appraisers may find the best opportunities as independent fee appraisers because the banks and other financial institutions that, in the past, employed a significant number of appraisers are increasingly contracting out to independent fee appraisers to make loan appraisals on a case–by–case basis, decreasing their need to have appraisers on staff. The increased use of automated valuation models to conduct appraisals for loan and mortgage purposes has also shifted work out of the financial sector."

I found it intriguing to look through this data as it indicates a couple of things to me. We are moving towards a shift in the workforce, in that, many long time appraisers are approaching retirement. That will mean a huge amount of experienced people will be gone over a relatively short period of time. That can be both good and bad for the industry and the people that depend on us to be here for them. This generation of appraisers going out was the first to see licensing, USPAP, and other regulations come into play. As time has moved along, required qualifications have increased, and many of the younger appraisers will have had more stringent education, although there is a current lack of quality education for pre–licensing. We will have to see how it all levels out, and hopefully it will mean positive trends for us as a group.

I know several guys that are in the bubble of possibly retiring soon (1–5 years), and many of them are very qualified appraisers. Their expertise will be missed by their clients, and by folks like me who call them from time to time to get their opinion on weird situations. I also know several that have basically been grandfathered through, and have never really gotten a well founded education in theory. True statistical analysis, regression and other types of supportable methods, are never used by them. Many only pull a minimum amount of research to evaluate. That doesn't mean they are doing poor work, far from it. These guys have more experience than I will have for several years, and they form well supported values. The can almost "feel" their way into good opinions, where younger guys need all the statistical stuff to be sure.

As we move into this transition I would think that a more commanding use of traditional methods, and more complex methods would prove much more useful to develop reports. The ability to move information from MLS format and the Aurora comp database to excel is not too complicated right now, but it would be pretty awesome if Wintotal could do this now. Importing and exporting data from program to program is an extra step that can be eliminated. In the event, as an example, you run regression analysis on a body of comps from a neighborhood to pull out support for an adjustment, running it in Wintotal would allow for the program to help sort and place the adjustment. Graphs, charts and other visual means could be produced and included for reference, or just to keep in the work file.

The other reason I think that this type of data analysis could be important is due to my work doing review appraisals. When folks do sub–prime type assignments for lenders, the reports are very closely reviewed by an underwriter. No offense to them, but they are not appraisers and many times they are interpreting their guidelines and will not understand the report they are reading. I will get a review assignment with an AVM that was pulled on the property and lo and behold, the AVM does not line up with the report. I reference that AVM very loosely – as I said before I don't really like them – only to find out in my data collection that the adjustments made are at best very loosely market based, or have been magically pulled from the air. The running joke in my office is to call it a SWAG (scientific wild a** guess).

Common sense tells us all that narrative comments should reflect why we are using certain types of adjustments, but being able to show "why" is a great way to support what we are doing. Many times I will do a review and formulate a differing opinion than the appraiser due to data being interpreted the wrong way. Sometimes adjustments jump out at us, and other times they are hard to isolate. Running models and forecasting on data can really help make sure we aren't jumping out on a limb. Sometimes we do have to jump on a limb when we have market fluctuations, but knowing we have exhausted most or all means can help us feel better about those odd ball assignments we all end up getting from time to time.

Having this set up in our software may even allow us to perform some desktop style market indicators, similar to AVMs, but maybe something that could directly compete with them. What could be a better forecasting tool for low risk assignments which could result in more money in our pockets? Lenders can't value property, never could and never will. OCC changes and new lending regulations will push the importance and independence of appraisers more towards our advantage. There are simply way too many commission based originators directly involved with the appraisal. I love to get the phone call from the guy that is basically saying "I know you worked on this for 8 hours, and you are so sure about your opinion that you even signed your name to it, certifying that you did your job right. I need $2,000 more or my deal is shattered...do you think you could squeeze it out?"

Don't you think that if lenders had had less influence on some appraisers during this current sub–prime fall out, that some consumers wouldn't have bought more than they could really afford? The lenders were still using their alternative models, but the problem is an AVM doesn't know the difference between 4 adjacent neighborhoods, where a local licensed appraiser can probably explain the differences in about a paragraph combined with a statistical analysis report that an underwriter is going to read and agree with.

I guess my point/question is where is the better "value" in service and reliability going to come from? An AVM from an office 5 states away, or an appraiser armed with a complete knowledge of the area and professional tools and reports? Lined up like this an AVM looks like what it really is: a heartless pile of silicone that is only as valuable as the person using it. And unless that person is appraising his or her own backyard...they aren't really doing their job right, are they?

I think by incorporating some of the tools they pass off as valuations will only lend to improving our position amongst the lending industry. We certainly don't need to rely on an AVM to determine value, but some of the tools that go into the reports they make are effective. Also, by having the ability to create additional reporting examples and possible visual guides (charts, graphs, etc.), we will only prove what most of us know, that we are doing our job better than a computer ever will.

These types of services will never replace the men and women that have laid the foundations for all of us as appraisers. They will act as a step towards the appraisal industry moving in a positive symbiotic relationship with out largest client bases. They will also help the large number of newer appraisers understand what they are doing better. There are many of these guys that have been taught incorrect methods, and procedures to formulate valuations, not even looking at statistical methods.

So what do my peers out there think? Could we see some advantages from our software being capable of doing these types of things for us? How many of you really think an appraiser armed with this kind of data could really benefit from it? I look forward to reading your replies, so keep them coming.