• Articles & News

      by Published on 09-21-2011 07:24 PM     Number of Views: 486 

      Regression Model For Fontana, CA - Closed Sales 9/1/2010 - 8/31/2011

      Using Salford Systems MARS 3.0 (Multivariate Adaptive Regression Splines)

      (DRAFT ONLY - Not for commercial use without usage license from Pacific Vista Net)

      Prepared by Bert Craytor, SRA on September 20, 2011

      NOTE: This analysis has been done without any first-hand knowledge of the locale!!

      Copyright (c) 2011, Pacific Vista Net


      • City: Fontana, CA
      • Property Type: Single Family Residence
      • Sale Type: All SFR closed sales from 9/1/2010 through 8/31/2011
      • Omitted: About 3 sales from Massa Way - for which data was insufficient
      • R2 : 0.91044 (Excellent)
      • R2(Adjusted): 0.90949
      • Input variables: 13
      • Predictors: 13
      • Number of sales: 2,2463
      • Min Sale Price: $34,500
      • Max Sale Price: $515,000
      • Mean Sale Price: $216,000


        Here is the Value@Grid spreadsheet (which can be used to estimate property values in Fontana, based on 3 comparables and a subject:



      Input Variables & Importance

      Variable
      QUALITY 100.00 |||||||||||||||||||||||||||||||||||||||||| Subject quality rating that encompasses those quality features not represented by the MLS data.
      GLA 93.46 ||||||||||||||||||||||||||||||||||||||| Gross Living Area
      LATITUDE 60.25 |||||||||||||||||||||||||
      AGE 49.66 ||||||||||||||||||||
      SALETYPE2$ 47.69 |||||||||||||||||||| Type of Sale: Short, Foreclosure, etc..
      LONGITUDE 24.49 ||||||||||
      LOTSIZE 18.61 |||||||
      GARAGE 15.10 |||||
      POOLYN 12.87 ||||| Pool Yes/No: "1" indicates a pool
      TLC$ 12.07 |||| Tender Loving Care Needed specifed in MLS
      BEDROOMS 11.85 |||| Bedroom count
      SALEAGE 10.60 |||| Date of sale in days before 9/20/2011
      COOLING 0.00 Whether air conditioning is provided
      VIEWYN 0.00 Whether a good view was specified
      BATHS 0.00 Bath Count
      VIEWYN_mis 0.00 View indication missing
      VOLT220$ 0.00 220 Voltage





      Model
      Subsets for SALETYPE2$
      SubSet1 = { "Buyer Concessions", "Contract Of Sale",
      "Seller Financing", "Standard Sale" }
      SubSet2 = { "Buyer Concessions", "Contract Of Sale", "Exchange/Trade",
      "Seller Concessions", "Seller Financing", "Standard" }
      SubSet3 = { "In Foreclosure", "Real Estate Owned", "Short Payoff",
      "Standard Sale" }
      Subsets for GARAGE
      SubSet1 = { "3", "4" }
      SubSet2 = { "2", "4", "5" }
      Subsets for POOLYN
      SubSet1 = { "1" }
      Subsets for TLC$ (='Y' means "TLC needed" was indicated in the MLS)
      SubSet1 = { "N" }
      BF1 = max(0, GLA - 2056);
      BF2 = max(0, 2056 - GLA);
      BF3 = max(0, QUALITY - 1);
      BF4 = max(0, 1 - QUALITY);
      BF5 = max(0, AGE - 44);
      BF6 = max(0, 44 - AGE);
      BF7 = max(0, LATITUDE - 34.0993);
      BF8 = max(0, 34.0993 - LATITUDE);
      BF9 = ( SALETYPE2$ is in SubSet1 );
      BF11 = max(0, LATITUDE - 34.16);
      BF13 = ( SALETYPE2$ is in SubSet2 );
      BF15 = max(0, LOTSIZE - 4000);
      BF16 = max(0, 4000 - LOTSIZE);
      BF17 = max(0, QUALITY - 9);
      BF19 = ( GARAGE is in SubSet1 );
      BF21 = max(0, LONGITUDE + 117.514);
      BF22 = max(0, LATITUDE - 34.0377);
      BF25 = max(0, 3 - BEDROOMS);
      BF26 = ( POOLYN in ( "1" ));
      BF28 = ( TLC$ in ( "N" ));
      BF30 = max(0, LATITUDE - 34.1116);
      BF32 = max(0, QUALITY - 9.5);
      BF35 = max(0, -193 - SALEAGE);
      BF36 = max(0, GLA - 3394);
      BF38 = max(0, LATITUDE - 34.0971);
      BF40 = max(0, AGE - 32);
      BF42 = ( GARAGE is in SubSet2 );
      BF44 = ( SALETYPE2$ is in SubSet3 );
      SalePrice = 69453.4 + 37.1355 * BF1 - 64.954 * BF2 + 5840.35 * BF3
      - 28137.2 * BF4 - 1411.55 * BF5 + 1585.38 * BF6
      + 6.58228E+006 * BF7 + 846840 * BF8 + 25423.3 * BF9
      + 5.96611E+006 * BF11 + 35208.8 * BF13
      + 0.591537 * BF15 - 10.6819 * BF16 + 27536.8 * BF17
      + 15652.4 * BF19 - 236148 * BF21 + 762708 * BF22
      - 11102.5 * BF25 + 9214.49 * BF26 + 14908.5 * BF28
      - 1.90949E+006 * BF30 + 227491 * BF32
      + 25.4779 * BF35 + 26.5947 * BF36
      - 4.96146E+006 * BF38 + 1019.18 * BF40
      + 7107.6 * BF42 + 25056 * BF44;

      GRAPHS

      GRAPHS
      by Published on 09-16-2011 08:57 AM     Number of Views: 493 

      This website serves to expedite collaboration among valuation professionals who are developing rules and models for property inspection and valuation. There are distinguishing characteristics of this website, as compared to other websites that deal with valuation:
      • The focus is on building useful models for property inspection and valuation.
      Almost equal importance placed on inspection as on valuation itself. A critical belief is that genuine property value estimation must always be preceded by actual or proxy inspection. Methods which purport to provide value estimates wihout such inspection can only provide "Rough Guesses" of value and should not be called value estimates.
      Unfortunately there is little theory behind the process of inspection: "Despite the incursion of inspection and monitoring activity in every sphere of human life, its theory as a scientific discipline does not have an adequately developed methodological basis. The fundamental concepts and definitions lack clarity, the approaches to research on the inspection (monitoring) of items of a diverse physical and mathematical nature are devoid of unity, the formulations of problems are not general enough, and so on.Until the middle of this century, the principal realm of engineering applications of inspection was piecework production in machinery construction, engine building, clockmaking, etc. As a rule, certain dimensions of articles, i.e., entities that could be described by a one-dimensional measurable quantity, were subjected to inspection. Inspection itself was regarded as the "determination of whether the true value of the measured quantity lies within or outside of its limiting values". The fundamental methodological components of scientific research in this kind of inspection, namely its object, subject, problems, methods of solution, etc., assumed a fairly obvious interpretation. The category of technical entities encompassed by inspection expanded continuously. They corresponded more and more closely to the new systems-design notions emerging in science and industry. Theoretical importance became attached to entities described by a set of vector quantities (e.g., production lots) or quantitative attributes (as in the case of dynamic systems and technological aggregates). New methods were required for the design and analysis of the inspection or monitoring process. Of course, they still did not necessitate a wholly new conceptualization of the basic methodological ideas and could be developed within the framework of the same initial treatments. But when theoreticians redirected their attention from discrete items to technological processes described by functional quantities, it became clear that such inspection could not be built on the old foundation. Discussions about the very definition of inspection (monitoring) commenced at that time [Bolychevtsev, Tsapenko, Shenbrot - The Methodological Foundations of Inspection Theory]."
      • Laws and court cases, including general standards such as USPAP, are discussed only as they translate into constraints on inspection and valuation. That is to say, such laws and standards are not regarded as an antecedent to good valuation, but rather as constraints that can vary independently of general valuation models from locale to locale.
      • International collaboration is seen as a way to bring together talented individuals to speed up development of new methodology.
      • Discussion, including unbiased reviews, of current hardware and software tools is encouraged.
      • Sharing of opinion polls and other member initiated market survey results is also encouraged as a basis for developming new methods and models.
      • Discussion and opinions are limited to tools, models, techniques, ideas and methodology. Rating, criticism or other comments, good or bad, directed at individuals is not tolerated. Note that caution must be exercised when reviewing company products, so as to avoid liability issues.
      by Published on 09-16-2011 09:33 AM

      Regression Model For Pacifica - Closed Sales 1/1/2010 - 2/16/2011

      Using Salford Systems MARS 3.0 (Multivariate
      ...
      by Published on 09-16-2011 09:46 AM

      What employment opportunities lie ahead for appraisers, going into the future? - There is a new product Value@Grid, coming ...

      AVM 

      by Published on 09-16-2011 09:37 AM

      In order to be able to efficiently generate MARS regression models for all of the areas around the San Francisco Bay Area, ...