Home Part II: What You Should Know About the Direct Comparison Approach and...

    Part II: What You Should Know About the Direct Comparison Approach and Were Afraid to Ask

    1664
    1
    Open folder with visible diagram lying on desk and hands pointing at it.

    (This post is Part II to the previous article “Standardizing the Direct Comparison Approach)

    The DCA is described as an approach. The dictionary describes an approach as “to come near or nearer to something, someone in space, time, quality and amount”. Somehow that definition does not fit the function of the DCA. It is not trying to get near to something, it is supposed to be a representation of the behavior of the actions of buyers and sellers interacting in a specific arena (market place). Therefore, I’d like to suggest that the word “approach” should be removed and replaced with “model”- the Direct Comparison Model (DCM). A model is defined as “a representation of a system that allows for investigation of the properties of the system and, in some cases, prediction of the future” as taken from Investorwords.com.

    Appraisers should consider themselves as part of the social science network of people that study the actions of humans. So whether a social scientist is studying the changes in the hem lines of women’s skirts from 1865 to present, the change in the business suit over the same period or why people are violent, it is all based on one common feature – – data of Human Interaction. Data is the key to understanding real estate markets. Logically then, data analysis and having a standardized DCM model would assist the appraisal profession significantly.

    Always remember the famous quote from George Box, a noted statistician.

    “All models are wrong some are useful”…

    In my opinion, the profession can benefit from building a DCM that is standardized that follows the rigors of proper data analysis and the validation of adjustment through testing.  Currently, what appraisers have is a framework that needs to be brought forward to the 21st Century.

     THE BASIC FUNDAMENTALS OF THE DCM

    What is the Function of the Direct Comparison Model?

    The Direct Comparison Model (DCM) has only one job. No, it is not to determine the value of a property. That is the result. The real function of the DCA is to do only one thing:

    Explain and Reduce the Variations in the Selling Prices of the Comparable Sales Based Upon Systematic Testing of the Data”

    Where does the role of Qualitative and Quantitative variables and the Unit of Comparison play in the DCM?

    Unit of Comparison

    Why do we ever need one? Well, the simple truth is the unit of comparison such as the selling price per square foot of building is the cornerstone of reducing the variation in the selling prices of the comparables.  It is the easiest to apply.

    For example, here are some sale prices.

     

    SALE DETAILS

     

    Index #1 Index #2 Index #3 Index #4 Index #5 Index #6 Index #7
    Address of Indices

     

    ADDRESS 1

     

    ADDRESS 2 ADDRESS 3 ADDRESS 4 ADDRESS 5 ADDRESS 6 ADDRESS 7
    Date of Sale December-12

     

    February-10 May-12 October-15 October-15 April-13 January-13
    Number of Square Feet

     

    5708.00 4439.00 3659.00 5290.00 6881.00 9239.00 7700.00
    Sale Price

     

    $525,000 $545,000 $392,000 $500,000 $650,000 $850,000 $950,000
    Selling Price Per Square Feet $91.98 $122.78 $107.13 $94.52 $94.46 $92.00 $123.38

     

    The overall selling prices have a difference of 142%. By using a unit of comparison of selling price per square foot of building, the sales have been reduced down to 34%. Just by using a unit of comparison we gained 108% in explaining the differences in the selling prices of the comparables. Now all we have to work on is explaining the 34%!

    When it comes to the unit of comparison there are no rules. It is all about reducing the variation in the selling prices of the data. The easiest way to determine the unit of comparison is based upon the type of real estate under appraisal.  We can determine the best unit of comparison by trying several of them and calculating the unit of comparison that has the lowest dispersion as a percentage.

    Qualitative and Quantitative Adjustments

    A qualitative adjustment reflects something in appearance and value. It is a measure of something by its quality not its quantity. A quantitative adjustment pertains to the quantity of something that can be measured in numbers.

    These attributes are going to be found in various parts of the DCM. That is fine as long as you know which variables belong to which category and more importantly how to make the adjustment.   We are going to group adjustments into two groups.  The first group are the base ones in order to remove more variability within the Unit of Comparison. The second group are those adjustments that pertain to the specific details of the comparable properties.

    The first set of variables found in the DCM is in the top section. It is shown as follows:

     

    FIRST SET OF VARIABLES IN THE DCM
    VARIABLE
    Property Rights(qualitative)
    Mortgage Financing(quantitative)
    Motivation(quantitative)
    Time(quantitative)
    Total Adjustments “Before” Other Adjustments

     

    Here is what that section should look like.

    Part 2 - table1
    Remember earlier we reduced the sale prices down to 34%? Now we have to lower this figure even more. However, we need to have a discussion about variables.The bottom section of the DCM deals with the variables that pertain to reducing the unit of comparison even further.

    Variables

    It is a common discussion about what variables to use for which property type. There is no set rule as variables can change over time. Variables are therefore not absolutes but rather something like a moving target. The real estate market changes so why can’t variables? We can categorize variables with certain types of properties. For example:

    Vacant Land for commercial development – variables are location, site size, site shape, servicing, zoning, official plan.

    Vacant land for residential development – variables are location, site size, site shape, servicing, zoning, official plan and estimated time of development.

    Vacant land for agriculture – variables are shape, lot size, topography, tiling, soil type, amount of bush or workable acres.

    Vacant land for recreational use – variables are lot size, lot shape, physical characteristics in terms of ponds, interior lakes, rocky outgroups, vistas, access and zoning.

    Improved house – variables are site size, lot shape, lot frontage, house age, house size, garage type, basement finishing, condition, interior amenities outside amenities and location.

    Improved converted office building – variables are age, architecture, building size, parking, lot size, location, building size and condition.

    Improved Custom office building – variables are parking, building size, lot size, age, income, condition and location.

    Retail Plazas – variables are use lot size, age, parking, income, tenant type, location, building size and building size.

    Campground use – variables are location, number of sites, income per site, potential for future development, condition, house type (single family residential, above store as in apt or trailer) and lot size.

    Rules of the Use of Variables

    Overview

    The assumption is that the appraiser will become attuned to proper data analytical techniques when embracing the DCM discussed in Part III. It is important that the appraiser not overplay the role of variables in explaining price. Given a data set of sale prices of houses, variables will not tell you the contributory value of a vegetable garden relative to no garden, or a garden shed against no shed or the difference between a rear yard with a chain-link fence as opposed to a wooden fence. Variables are not microscopic in terms of their role in explaining every detailed improvement of housing, for example. They represent the majority of why prices are what they are and would only explain up to 99% of selling price variation. What about the other 1%? That is known as statistical noise. Adding more variables into the mix may not necessarily improve the function of the model.

    There are only two rules to follow regarding variables.

    1. Do not have many variables (*e.g. seven) when using four sales. This goes against the grain of basic data analysis and statistics. It is very difficult to determine the market effect based upon a given data set by trying to put four sales through the proverbial analytical ringer to squeeze out the impact of seven variables. It is called Degrees of Freedom. The variables need to roam through the sales in order to be able to extract any meaningful conclusion as to the impact the variables have on the data set. It would be like being asked to record the behavior of a dog while it is in a kennel. The behavior or variables will be limited to eating and sleeping even though you really want to measure more detailed behavior. However, if you let the dog out of the kennel your behavior variables are expanded to include such items as determining intelligence, how it eats, how it plays, how it listens, how it obeys, etc. Well, with data it is the same thing. Don’t ask four little sales to reveal all of its behavioral secrets when you place them in a kennel and start demanding variables regarding site size, building size, architectural features, number of parking spaces, basement finishing, condition, etc. The variables will breakdown because of the lack of “degrees” or mobility to freely association with the number of sales.
    1. Just because the sale price is expressed in dollars that does not mean that the adjustments have to be in dollars. The sale price of any property is nothing more than the end decision of the buyer and the seller reaching a conclusion about a piece of real estate. No buyer buys real estate by saying: “ I will give you $25,000 for the lot, $100,000 for the base house, $14,500 for the garage, $13,000 for the basement or $15.00 per square foot for the recreation room and $123 per square foot for the house, etc. What they say are verbal statements such as:
    • “I really like this house. Will you take $225,000 for it?”
    • “Your property has a great location. Perfect for commuting to work.”
    • “You have great tenants and the four-plex is in good shape. I would like to offer you “X” amount of dollars.”
    • “Nice house but it needs a lot of work.”

    Therefore, buyers and sellers interact with one another using words not individual prices. They cannot tell you the absolute reasons why they paid or sold at a specific price. So if that is the case, then how can we use the language of the buyers and sellers to make adjustments? To even take it a bit further, how can we take the language of buyers and sellers and convert it all to dollars and cents that matches the unit of comparison?

    See Part III to be published next week, August 26.

    AIC Disclaimer:

    This post is part of the AIC’s innovative program to explore new and creative concepts for valuing real property within the broader context of advancing the profession to meet and complex marketplace and evolving profession. To achieve this end the author(s) of these blogs/articles have the freedom to raise, express and discuss ideas and opinions that are not necessarily endorsed by  the Appraisal Institute of Canada’s (AIC) or comply with its professional guidelines and standards. While the AIC edits all blogs/articles for literary correctness it does not judge or edit the merits of the blog’s/article’s ideas or concepts. Readers are encouraged to discuss the ideas and contents of these blogs/articles on-line, and to share their own thoughts and ideas through the comment section below.

    George Canning, AACI, P.App

    George Canning is the principal of Canning Consultants Inc., a real estate appraisal and consulting firm. He has over 30 years of practical and diversified experience with several of the largest real estate firms in Southwestern Ontario. George now specializes in providing specialty consulting services to meet the needs of the clients that were not being met by traditional valuation methodologies. In particular, he provides solutions to complex real estate problems using modern techniques that in the past could not be reliably solved. He is one of a very few real estate appraisers/consultants that employs modern statistical methods and modeling tools with a common sense approach based upon many years of analyzing real estate.

    SHARE
    Previous articlePart I: Standardizing the Direct Comparison Approach
    Next articleBREXIT and the Price/Value Equation

    1 COMMENT

    1. I agree with the suggested name change from DCA to DCM – this term more accurately captures the generally accepted appraisal practice of building a economic model to validate and support the prediction of the sale price of a property at a certain date.

      A very useful part of this blog is the discussion of degrees of freedom. Attempting to reconcile all physical and economic differences between a small number of comparison sales and a subject property isn’t possible or statistically valid. George was right in recommending appraisers avoid adjustment for variables that likely have a minor contribution to the sale price of a comp are not useful – I would go further and suggest the influence of these variable is almost impossible to substantiate and is likely not a factor in purchaser motivation. We need to rely on a combination of data exploration (statistical analysis) and our experience to understand whether a variable is a material influence on the sale price in a certain market.

      Thanks George.

    LEAVE A REPLY

    Please enter your comment!
    Please enter your name here