Good Morning Good Morning Dr by fanzhongqing


									           Good Morning

          Dr. Michael Furick

–Faculty member at Georgia Gwinnett
College, School of Business

–Teach Management Information Systems
       and Marketing
             Today’s Topic

Using neural networks to develop
 decision support systems to chose
 tenants for apartment rental.

  Results of a pilot study
If we asked about rental property…

  Owning rental property is the best
  financial decision you will ever make


  Owning rental property is the worst
  financial decision you will ever make
Three levels of apartment tenants

Good tenants…….heaven on earth

Bad tenants……….hell on earth
         worry                       s this

No tenant………….empty unit
         worry and            Fear
         lose money            of
   Picking “good” tenants is vital to
    rental business success

   and sanity
Many decisions about tenants get
made every year

   34 million households live in rental
    housing (held steady due to
   20% renters above $60k income
   20% renters below $10k income
   56% of rental units owned by
How do other industries pick

   Most rely on credit reports and
    credit scoring to predict consumer
    financial behavior
       Banks
       Car dealers
       Mortgage brokers
       etc
What is a credit report and score?

   Credit report- a multi page report
    that profiles a consumer’s financial

   Credit score- mathematical means
    of summarizing the credit report
    into a three digit number
          Credit score is widely used because it
          is predictive and easy
                              Delinquency Rates by FICO Mortgage Risk Score

Rate of Delinquencies

                        20%                                       15%
                                                                            5%       2%      1%

                              below   500 to   550 to    600 to   650 to   700 to   750 to over
                              499      549      599       649      699      749      799 800
                                                        FICO Score Range
Success has caused credit score use
to spread to other industries

Auto industry uses credit score to
 Determine who gets auto insurance

 What price to charge for an auto policy

Two studies found that a lower credit score
 Up to 50% more accidents

 Bigger claims ($918 vs. $558)
Credit data and credit scores should
work in apartment rentals

Two Part Study
 First part of study looked at tenant
  performance vs. six commercially
  available credit scores
  (statistical analysis)
 Second Part: If credit is not
  predictive then what is predictive?
  (neural network analysis)
Credit data and credit scores should
work in apartment rentals

   First part of study looked at tenant
    performance vs. six commercially
    available credit scores

   22 different credit scores are
    available from Experian
Six scores tested from Experian
   FICO Mortgage Risk Score
   FICO Advanced Risk Score
       Derogatory credit in 24 months
   FICO Installment Loan Score
       Repay short term loans auto etc.
   FICO Finance Score
       Loans from non-traditional sources
   National Risk Score
   Sureview Non-Prime score
       (non-prime bankcard applicants)
Correlation examined:
   credit score vs. tenant performance

   Data collected
       One apartment complex
       200 tenants that moved in during 2002
       6 scores collected on each tenant
       Tenant performance followed in
        satisfying lease over 12 months

   Traditional statistical methods used
    to examine correlation
Results Part 1: credit data not
      predictive of tenant performance

   No correlation between credit data
    and tenant performance in
    satisfying the terms of their lease

   R square approaching zero

“We have as much trouble with people with
 good credit as we do with people with bad
       credit”          property manager quote
Why are commercial scoring models
not predictive in selecting tenants?

   ??????
   Many “good working” models filter
    out consumers with
       Less job tenure
       High ratios of debt to income
       Older vehicles
What would be predictive in Part 2?

   Hints from the decision process
    used in the apartment rental
Picking tenants more complex than
picking customers

   Financial consideration
   Non-financial considerations

   Non financial consideration affected
    by Fair Housing Laws
Decision process mostly manual with a
range of data and big dose of “gut feel”
96 units Baltimore        Reject if landlord problems or criminal
                          Reject if bankruptcy
395 units Chicago         Credit score in top 15%
                          Reject if landlord problems or criminal
264 units Chattanooga     Reject is landlord problem or criminal
                          Reject if bankruptcy
210 units Athens, Ga.     Income 3 times monthly rent
                          80% satisfactory accounts
                          Reject if landlord problems or criminal
68 units Washington D.C   Reject if landlord problem or criminal
                          Reject if bankruptcy
Nationally, property managers make
rental decisions on a range of items

   33.8% ran criminal backgrounds
   62.6% ran credit reports
   65.5% called references
                          Rental Property Reporter

   50.6% ran credit reports
   52% verified income
   75.5% relied on personal interviews
                                 U.S. Census Bureau
Opportunity to standardize decision
making with a Decision Support Model

   Data to be a mix of financial and
    non-financial items (matching
    current decision process)

   Apartment managers suggested 76
    possible variables
Sample of data elements used in
neural network model
   7   From out-of-state (Application)
   11 Size of employer     (Chamber of Commerce)
   12 Number of years with employer
   15 Income       (Verification)
   20 Number of people to occupy apartment
   34 Type of vehicle one      (Application)
   35 Age of vehicle one        (Application)
   48 Estimated monthly installment loan payments
    (Credit Report)
   68 Number of driving infractions     (DMV report)
   73 Information found on county criminal search
Data collection process

   One apartment complex
   Data elements collected on 60
    tenants as they moved in during
   Tenants lease performance tracked
    over 12 months
Why use neural networks to create this

   Neural network – an artificial
    intelligence system that is good at
    finding and differentiating patterns

   modeled after the brain’s mesh-like
    network of interconnected
    processing elements (neurons)
Why use neural networks to create this

   NNs good with unstructured data
       how do data elements interact with
        each other or with the output
   Analyze nonlinear relationships
   Learn and adjust to new
Layers of a Neural Network

Input Layer   Hidden Layer   Output Layer
Why use Palisade’s NeuralTools®

 Over 50 NN software packages
 Evaluated about a dozen

 Feature, function, benefit
Actual model creation details not
covered here

   Data divided into test and training
   Model run several hundred times
    using various combinations of
   Prediction accuracy recorded and
    analysis completed
What did the model find?

   Model accurately predicted 69.1%
    of tenants (good and bad )
   Three data elements became most
    important in choosing tenants
    1.   Percent satisfactory accounts on credit
    2.   Total applicant income
    3.   Driving record of applicant
 Comments on driving record as
 predictor in apartment rentals

    Auto Industry
                     Predicts   Record

    This Study

  Credit          Predicts
Performanc                      Driving
     e                          Record
Limitations with this pilot NN study

   Small data set
   Single geographic region (one
    apartment complex)
   Data set of those who moved in
    (sample selection)
Next Step for researchers

   Proposal submitted to National
    Science Foundation to fund
    expansion of the study to the
    Southeastern U.S.
Thanks to Palisade Corporation for
hosting the conference
Thank you for attending

Questions now and later

 Dr. Michael Furick
 Copies of the detailed result and model are available
  for purchase from

 Document UMI number: 3215298

 Citation:

Using neural networks to develop a new model to
  screen applicants for apartment rentals. Furick,
  Michael T., PhD. 2006.

To top