When the Loudoun County Assessors office calculates the assessed value of the homes in Loudoun County their goal is to determine the Market Value of each and every home in the county as of January 1st, 2008. That Market Value on that day is your assessed value.
How do you Calculate Market Value?
There are two obstacles the assessors office faces in trying to come up with Market Value. First of all, they cannot go out to each and every house in the county on January 1st and perform an appraisal to determine the Market Value. That would be almost impossible to due and completely unrealistic from a manpower standpoint. Therefore, they have to use some type of algorithm or formula to apply to most of the properties in the county. This calculation shouldn't be too inaccurate given the fact that a lot of the homes in the county are very similar to one another, especially in the subdivisions of eastern Loudoun County. The second problem with determining Market Value is need for actual sales at or very near to January 1st to be used as the baseline for the Market Value calculations. And given that a very small number of homes will actually sell on January 1st, what time frame should be used. And also, should a sale be the day a contract is written, the day a contract is ratified or the day a house actually settles?
Here are a couple of thoughts on how the assessors office could calculate the Market Values given the limited resources and limited number of sales that occur on January 1st. This is not the actually processs the county uses, but a couple of guesses at possible algorithms they could utilize. Each has its limitations and downsides but used in combination they can be used as a double check against the actual assessed values.
2007 vs. 2008
First, the county could easily make the assumption that the 2007 Assessments were an accurate Market Value as of January 1st, 2007. This can be assumed and probably was assumed because if the values were not accurate than the homeowners would have contested the values and they would have been adjusted last year. Therefore, we can apply a multiple to the 2007 Assessments and thereby create the 2008 Assessments. This multiple will reflect the increase or decrease of the Market Value of the homes in Loudoun County over the past year.
What multiple does the county use? How many multiples does the county use? One for each zip code? One for each subdivision? One for townhouses? One for condos? The smaller the slice, the smaller the amount of data available and a potentially less accurate multiple. But the larger the slice (one multiple for the entire county) the higher the likelihood of entire neighborhoods being inaccurately assessed.
So the goal is to create small enough slices that will accurately reflect the Market Value for specific neighborhoods and/or housing types but large enough to allow enough data to be used to avoid a skewing of the numbers. I have to use the same determination when I generate Competitive Market Analyses for home sellers and it usually entails looking at the market from a broad perspective and then narrowing the focus down until the focus is small enough yet the data is broad enough.
Postal Cities or Zip Codes
The first slice the county might use would be postal cities but I know that there are tremendous disparities within certain cities (Sterling) that it would probably be better to narrow it down to zip codes. In eastern Loudoun County that will give them 9 zip codes to calculate. Now within those zip codes there will definitely be neighborhoods and subdivisions that should be looked at individually but that should be done later in the process.
Contract Date vs. Settlement Date
Secondly, the time frame to use for the baseline data needs to be determine. If the county wanted to be as accurate as possible in achieving the Market Value as of January 1st, the best data to use would be the houses that went under contract within 1 week of January 1st. This would be a time frame of December 25th to January 7th. The problem with that is there are only 95 homes in the entire county that sold then and many of those have not settled as of today so the actual final sales price is yet to be published anywhere. There should be more properties that make up the baseline, therefore the timeframe should be expanded perhaps from December 16th to January 15th.
The two problems that exist in this calculation can be rectified one of two ways. We can change the sales date from the day a contract was ratified to when it was settled. The problem with this is that a home that settles on January 1st could have gone under contract at least 30 days and in some cases more than 60 days previous. This means the sales numbers are more reflective of prices in October and November, not January 1st.
Sales Price/List Price = % of List Price
The second way to fix the problem is to keep the contract date as the sales date, figure out how much the final sales price is as a percentage of the list price for the transactions that have settled and apply that to the ones that are going to settle but haven't yet. This calculation would be applied to the current year but is unneccessary for the previous year. The problem with this calculation is that the percentage to list price is an estimate and could be inaccurate (but probably not by much). But a bigger problem is that a potential sale could fall out and not be a sale at all.
There are also several other factors to keep in mind regarding the data set that is used by the Assessors office. The dataset I use to make calculations in the MLS. This dataset does not include transactions that are not listed by a realtor. This means that there are no FSBO sales, very few new home sales and very few auction sales (foreclosures). Obviously this makes the dataset incomplete but when making comparisons of 2007 vs. 2008 as long as the criteria for inclusion in the dataset is the same, than the relationship between the numbers should be consistant. We are not looking for an absolute number, just an idea of how the numbers relate.
Another factor to keep in mind is the Assessors use of foreclosure properties and short sales. Are these distressed properties removed from the calculations? What if a majority of the transactions are these types of properties? Do you apply a factor to them or just eliminate them? I believe the houses that are included in the MLS should all be treated equally because they are all being marketed to the same universe of buyers. But I am sure that others will disagree with including them in the valuations and strong arguements can be made on either side.
What about Foreclosures?
There are many different variables to be considered in determining not only the dataset to be used in the calculations but how this dataset will be narrowed down and applied geographically as well as by property type. There is also a question as to the timeframe to be used in the calculations, the definition of sales date, etc. And all this has to done and delivered to the homeowners by Feb 1st.
And now that the assessments have been done, how accurate are they? I think that given all the variables left open to interpretation, the best way to check the accuracy of the assessments is to look at the data in at least 2 different ways and see how the actual assessments stack up. The second part of this series will look at the use of 2 of the calculations suggested above .
Contract date between December 1st and December 31st - 2007 vs 2008
Settlement date between December 16th and January 15th - 2007 vs 2008