Collateral Risk Network. The Language of Data. April Elizabeth Green

Size: px

Start display at page:

Download "Collateral Risk Network. The Language of Data. April Elizabeth Green"

Linette McDonald
5 years ago
Views:

1 Collateral Risk Network April The Language of Data Elizabeth Green 1

2 2 CRN April 2012 Appraisal Prose? I came to explore the wreck. The words are purposes. The words are maps. I came to see the damage that was done and the treasures that prevail. - Adrienne Rich, 1929

3 3 CRN April 2012 Rubrics & Methodology Its not only what you say, its how you say it. MISMO provides a common vocabulary and definitions. Mathematical pictures; statistical evidence requires standardized and recognized methods, data and definitions.

4 4 April 2012 Data Chaos

5 5 April 2012 Origins Appraisal forms were originally honed around the assumption that the appraiser and his intended reader had a common understanding of the subject s market. Lender collateral guidelines are based on underwriting practices with delegated authority. The GSEs never took the appraisal report before UMDP

6 6 April 2012 Data Chaos Theory

7 7 April 2012 A Step Toward Standardization UAD defines how to fill out the form to meet GSE business requirements. Certain topics enumerated but still uses commentary in key areas like market details, neighborhood descriptions, supporting details about condition & quality are textual descriptions.

8 8 April 2012 Technicalities of MISMO v2 Series MISMO version 2 is transaction centric, based on older technology framework(dtd), & difficult to extend. 2.6 is published as an XSD but not a true schema-based format, no strong data typing that can be recognized by commercial parsers.

9 9 April 2012 XML Benefits Using Schema-based validation, it is possible to express an XML document's structure and content in terms of the data model that was implicit during its creation. The XML Schema data model includes: Vocabulary (element and attribute names) Content model (relationships and structure) Data types

10 10 April 2012 Technicalities MISMO v3 Series True schema, strong data typing, Namespace aware, SMART Doc securityready and highly extensible. All GSE UAD extensions were added as named fields in MISMO v3.1 in Loan file centric.

11 11 April 2012 Friction Underwriters try to understand the appraisal by applying their own judgment using data sources not necessarily used by the appraiser and without the benefit of local market knowledge.

12 12 April 2012 Re-engineering Its not about another appraisal form; its about relevant content. Back to basics, the appraisal process: Identify the problem Access & verify the data Analyze the data Conclusions & opinions

13 13 April 2012 The New Foundation - Data Using a common vocabulary, data can exchanged with more integrity and confidence. Information in the appraisal process can be dynamically suited to the scope of work and unique market/property characteristics.

14 14 April 2012 Up & Coming Trends Statistically-based tools for both appraisers, underwriters and investors. Big data and open standards in government and regulating agencies. Data ownership, copyrighting and authenticity.

15 15 April 2012 Reference Dictionary MISMO v 3 is not a single file format but rather a centralized dictionary for all terms needed to support the mortgage origination and servicing process. Data can be more easily be exchanged and applied throughout the loan lifecycle.

16 16 April 2012 Data Sourcing Credible sources and verifiability. Data ownership and terms of use. Limitations, reuse and freshness. Accountability.

17 17 April 2012 MISMO v3 Series

18 18 April 2012 MISMO v3 Series

19 19 April 2012 Better Data for All To improve the appraiser s analysis and the underwriter s capabilities, a consistent flow of clean, credible, ready to use data must be made more readily available. Apples to apples - we must all be able to recognize, define and spell apple the same way.

20 20 April 2012

21 21 CRN April 2012 Semantic Preciseness MISMO 3.2 provides a common vocabulary that encompasses the entire loan file (credit, capacity & collateral). Common dictionary eliminates confusion and misunderstandings and allows tasks, such as appraisal reporting to be succinct and meaningful.

22 22 April 2012 Compliance Strength Testing Reliable, accurate and clear data interpretation are keys to successful securitization AND to building compliant solutions to Dodd Frank & IAG mandates: New Valuation Products USPAP Compliance C&R

23 23 April 2012 MISMO v3 Series

24 24 April 2012 Better Lending Decisions The lender needs more data than the submission to an investor to ensure that the guidelines are met and repurchase risk is fully mitigated. Much of the supporting data in traditional forms are trapped in freeform text which can only be reconciled by a human reading the comments.

25 25 April 2012 UAD Support in v3 Series All UAD data points have been reconciled in v3.1 & v3.2 adds new architectural features. GSE UAD specific codes for Condition and Quality have been added as a data type that provides for a link to the GSE definitions.

26 26 April 2012 Who s On First? Should your investor know more about the loans you sold them than you do? The mechanics to validate and ensure the accuracy of appraisal results need more details in the valuation process and collateral adjudication then the summary results that are submitted to the investor.

27 27 April 2012 Granular Property Details

28 28 April 2012 MISMO v3 Series

29 29 April 2012 MISMO v3 Series Detail at the item level.

30 30 April 2012 Think Outside the Grid MISMO v 3.2 provides an attribute at the property level that identifies its context as SUBJECT or COMPARABLE. Adjustments at the item level. Any detail that can be described on the subject property can also be identified on the comparable property.

31 31 April 2012 Separating the Form Mentality The appraiser can be empowered with automation to interact with data resources and stakeholders appropriately and with greater precision and expediency without undue influence or further deterioration of fees for services rendered.

32 32 April 2012 One Dataset Multiple Views V3.2 empowers a more complete data set than beyond the traditional appraisal forms. Because the data elements are commonly defined, reports that are specific to the scope of work and intended use can be dynamic to meet business requirements.

33 April 2012 Data Box Imagine the full supporting data file being delivered in a way that allows intended user to easily drill down and perform

33 33 April 2012 Data Box Imagine the full supporting data file being delivered in a way that allows intended user to easily drill down and perform what if scenarios. Solutions can be developed that assist both the reader and the writer in the appraisal process when we use a common vernacular.

34 34 April 2012 Visualizing Data Visualization of the ACTUAL DATA, rather than statistical analysis alone is the perfect hybrid. New visualization tools can show the data itself and business users can draw conclusions using pattern matching and trend recognition.

35 35 April 2012

36 36 April 2012 Correlation is not Causation We like to construct simple cause-and-effect adjust comparable properties to explain the market as we experience it. Thinking statistically is tricky. We need to train both underwriters and appraisers. If you can t write, you can t think. The same is now true in math. Statistics is the new grammar.

37 37 April 2012 Data as Currency If knowledge is power then is data is gold? Data as a currency is viable provided others can perceive its usefulness and value. Understanding, verifiability and source are key.

38 38 April 2012 Future Appraisers Gen Y/MIllennials are Google-centric. Access to data and solutions are expected to easy, intuitive and instant. Technological training tools can in many cases do a better job in ensuring the competency level of an apprentice.

39 39 April 2012 Common Ground Reducing data translation and cycles removes friction from the overall process. Increased accuracy and precision in understanding appraisal results benefit the client and the appraiser, and all downstream systems and processes. Open data standards encourage technology innovation. Advances in technology and data access can be rapidly made available to appraisers and appraisal clients.

40 FOLLOW ME ON Thank You!

86 years in the making Caspar G Haas 1922 Sales Prices as a Basis for Estimating Farmland Value

2 Our Journey Begins 86 years in the making Caspar G Haas 1922 Sales Prices as a Basis for Estimating Farmland Value Starting at the beginning. Mass Appraisal and Single Property Appraisal Appraisal