FIBO / FIB-DM 2020

FIBO / FIB-DM 2020 fireworks

Abstract

The year 2020 brings exciting new content for the Financial Industry Business Data Model, FIB-DM, and it’s original, the Financial Industry Business Ontology, the FIBO. Four thousand five hundred entities are supporting investment management and banking loans. This article examines the latest FIBO Development version and its data model transformation.

Background

The FIBO is the authoritative model of Financial Industry concepts, their definitions, and relations. FIB-DM is a complete model transformation of the FIBO into a Conceptual Data Model.

Following the EDM Council versioning and release schedule, FIB-DM updates quarterly and comes as two data models, Normative and Informative:

  1. Normative, derived from the FIBO Production release
    1,968 entities as of 2020/Q1
  2. Informative, derived from the FIBO Development release
    4,563 entities as of 2020/Q1

Breakdown of the 500 downloads

More than five hundred users downloaded the FIBO Data Model since its launch in mid-August.

FIB-DM daily downloads (column chart)
FIB-DM daily downloads

Most downloads are from Financial Institutions (FIs) on PowerDesigner and ERWin data modeling tools. The rest are global and local consulting companies, academic institutions, and individuals. The combination chart shows the daily number for download-form submissions and the running total downloads. The actual model file downloads are 939. The higher number includes download link forwards and users downloading several formats. Downloads are sites and the number of users depends on the size of the data modeling team at the FI. I.e., a data modeler downloads FIB-DM, puts it in ERWin ModelMart, and her colleagues take it from there rather than submitting the form several times.

Most institutional downloaders, 74% are Banks. Investment Managers/Servicers, Brokerage, and Capital Markets comprise 17% of the downloads.

FIB-DM Institutional downloads (pie chart)
FIB-DM Institutional downloads

I didn’t analyze and segment the more than one hundred institutions by size and country yet, but the data suggests:

  • As expected, most downloads are from midsize FIs
  • Surprisingly most downloads originate from Europe.

The Enterprise Data Modeling Council, EDMC, doesn’t publish FIBO download numbers. However, we can expect a larger Financial Institution has the tooling, databases, and expertise to consume the RDF/OWL ontology files. Participation in FIBO conferences and content groups also suggests a keen interest from North American Investments Managers and Servicers in the ontology.

FIB-DM is a solution for midsize FIs who want to leverage the industry standard, but don’t have RDF (triple) Stores, Topbraid/Protégé, and ontologists yet. Very large FIs likely already customized the FIBO and encoded simplified versions of the Configurable Ontology to Data-model Transformation (CODT).

FIBO/ FIB-DM packages

The EDM Council describes the maturity of Development version ontology as Well-Vetted and Draft. We use the documentation status as a proxy for the draft. The bar chart shows the number of entities and maturity per package.

FIBO FIB-DM Development version maturity
FIBO FIB-DM Informative version maturity

FIBO content in production supports Securities, Derivatives, and Indicators & Indices.

Chicken and the Egg
Chicken and the Egg

While this is a priority for the 14% Investment Managers/Servicers, banks would like to see more support for banking services.

However, FIBO is not a vendor product. The FIBO development team depends on EDMC members providing human resources, model- and subject matter experts in the Content Groups.

Financial Institutions participating in the content group shape the industry-standard – they prioritize FIBO content.

Likewise disappointingly only one insurance company downloaded the FIBO Data Model so far, and I haven’t seen any Insurance presentations yet. FIBO Foundation, Business Entities, and Finance Business & Commerce are a solid foundation to model policies. However, the industry remains on the sidelines “waiting for [explicit] Insurance support.”

Offer your Insurance experts to work with the FIBO development team, and there will a FIBO (FIB-DM) insurance module.

Hence the decision to publish two data models. Commercial licensees no longer have to wait for FIBO Development modules to promote into Production. They can consult the informative data model and decide for themselves.

FIBO DEV statistics

The FIBO Product portal is the download site for Production and Development versions of the ontology. The Development version is still subject to change. Some classes may be without definition, and documentation indicates where ontologies may add more to a different module. Ontologists, subject matter experts, and a collaboration with the Object Management Group (OMG) ensure a rigorous review of the final FIBO model. As a result, the Normative Model is very stable, and while the EDM Council may add additional ontology classes and properties, there are view design revisions.

The Development version is a pipeline of FIBO content to release into production. The following is a showcase of what to expect in 2020. I transformed classes and object properties into a Dev PowerDesigner model and generated the data model reports.

FIB-DM 2020 new packages and entities
The pie-in-pie chart shows the number of entities in the current packages and the breakout pie of new FIB-DM packages.

The chart analyses the latest, FIBO Development November 23 download. The pie slices are FIBO modules and their number of entities. The pie in the pie breaks out the new 2020 modules. The EDMC FIBO DEV has 3952 named classes, double the current PROD number. Five new ontology modules, Loans, Business Processes, Collective Investment Vehicles (Funds), Corporate Actions, and Market Data, introduce 1179 new entities. FIB-DM extended; the commercial version adds FIBO modules once they are in Production. The numbers do not include 1384 associative FIB-DM entities derived from new object properties.

In a year from now, the Financial Industry Business Normative Data Model may have more than 3,000 entities.

FIB-DM DEV new entities for 2020
FIB-DM DEV new entities for 2020

The column chart shows new (blue) and existing (orange) entities per FIB-DM package. The PROD status indicates 1479 base entities that are already in the Financial Industry Business Data Model, FIB-DM, derived from the FIBO 2019 Q4 Production release. In other words, DEV indicates 2479 classes that are still in FIBO Development. We notice new FIBO modules/FIB-DM packages for Business Processes (BP), Corporate Actions and Events (CAE), Collective Investment Vehicles (CIV), LOANs, and Market Data (MD). Current packages, Foundation (FND), Securities (SEC), and Derivatives (DEV) many new classes/entities.

New Ontology modules / Data Model packages

The FIBO development version provides new modules for FIB-DM Informative.

Corporate Actions & Events

A Corporate Action is an event that brings a material change to the company or securities issued by the company. Two hundred thirty-nine new base entities support Corporate Actions for bonds and shares, as well as communications and elections.

Collective Investment Vehicles

The new package for Investment Funds has One Hundred Nine new base entities defining Fund types, strategy, administration, units, and management.

Market Data

In finance, market data is the price and trade-related data for a financial instrument reported by a trading venue such as a stock exchange. Market data allows traders and investors to know the latest price and see historical trends for instruments such as equities, fixed-income products, derivatives, and currencies. (Wikipedia)

Two hundred forty-nine base entities define the temporal concepts of Market Data, Security Status, and Analysis Curves.

Business Processes

Business Processes is not a generic package (yet). The focus for the two hundred eighty-one entities is on Securities Issuance and the Financial Context of Trade, Clearing, Settlement, Custody, and Market.

I advocate more generic BP content in supporting ontology derived BPMN models!

Banking – support for Loans

The largest new package with more than three hundred entities is the most anticipated enhancement for FIB-DM users. Sub packages Loan Core & Mortgage and Extended provide the framework. Loan temporal covers origination, status, and payments. Loan Types defines subtype entities for retail and commercial loans.

FIB-DM Loan Entities stacked column chart
FIB-DM Loan Entities stacked column chart

Loan (core)

The main package for sub packages supporting Collateral, Guarantees, Basic Terms, Parties, and Regulatory aspects.

Collateral

Entities in this package describe the concept of loan collateral and loan property collateral in the context of collateralized securitized loans.

Guaranties

This package defines various kinds of guaranty for different kinds of loans, along with the concept of a loan guarantor. These include mortgage loan related guaranty terms.

Basic Loan Terms

This is the core package for loan entities. These include loan purposes, loan payment terms, and parties.

Loans Parties

Entities in this package define the parties to a loan, including borrowers and third parties such as collateral providers and loan servicers.

Loans Regulatory

This package contains entities relating to regulatory requirements around loans, including disclosure rights and a small number of regulation-specific concepts. These include definitions of rights conferred on borrowers under consumer protection law, rights to equal treatment, and the like.

Loan Core & Mortgage

The main package for core and mortgage-related sub-packages.  Entities in the sub-packages define the loan contract and the US Home Mortgage Disclosure Act (HMDA) support.

Loan Core Ontology

This package defines the fundamental concepts for loans, with emphasis on mortgages.  The loan contract is the core concept. It includes the primary obligations to fund the loan and to pay it back according to payment schedules. The contract is also connected to various parties including borrowers, lenders, and originators.  Kinds of loans covered in this ontology include open and closed-end, secured, and mortgages.  Another major thread of the loan core ontology is credit risk assessment, which includes property appraisals, security agreements, pre-approvals, credit scores, and credit reports.

Loan HMDA Ontology

This package defines entities specific to the HMDA rule. This includes the concept of an HMDA report as well as specializations of the core classes for pre-approval requests, covered loan contracts.
The Home Mortgage Disclosure Act (HMDA) is US regulation.

Loan Extended

The main package for LN extensions. The current development packages LN, LNX, and Loan are likely to be reorganized in Production

Loans General

This package contains a small number of additional entities for general loan terms. These are mainly based on lists of values such as income verification methods, risk levels, and documentation types, as referred to in the loan application process.

Loan Generic Ontology

This ontology defines a number of concepts that are not specific to a loan and belong in more general FIBO ontologies. They are in the process of being moved over. This ontology will eventually go away.

Loan Temporal

The main package for temporal aspects such as loan status, phases, and events. Sub packages are for  Loan Applications, Appraisal, Borrowers, Construction Loans, Collateral, Events, LTV, Phases, Status, and temporal.

Loans Temporal core

Loan concepts with a date or time component. For example, amounts outstanding, loan balances, and interest rates at a given point in time.

Loan Applications Temporal

The package defines entities for the loan application process, including the temporally sensitive concept of loan phases, along with kinds of the loan application and the parties thereto.    These concepts include mortgage loan applications, which will need to be integrated with the HMDA-based work in Loan Contracts.

Loans Appraisal

Entities relating to the appraisal of a loan, including mortgage borrower assessment information and repossession history.

Loan Borrowers Temporal

The package provides temporally sensitive entities elating to loan borrowers themselves, including such concepts for different kinds of the borrower (corporate and personal). These include things like credit statuses, payment history, and facts about the borrower at assessment and at other times.

Construction Loans Temporal

Temporally sensitive concepts relating to construction loans.    At present The package simply defines the basic concept of a snapshot in time for a construction loan. Additional details will need to be added.

Loan Collateral Temporal

The package covers date-sensitive facts about the collateral provided for a collateralized loan, including security collateral.

Loans Events

The package defines a wide range of events relating to loans. These include legal proceedings, prepayments, events affecting the collateral such as valuation, purchase, and sale and details of many of these concepts.    Note that most of this material was derived from an earlier EDMC activity with IBM Research and is substantively based on a logical data model from that exercise.

Loans LTV

The package defines the concept of a loan to value (LTV) ratio, including both the original LTV at application and the same concept as a variable over time.

Loan Phases

Entities in this package define the phases in the life of a loan, including delinquent loans, up to the point where the loan is repaid.    Future integration of this work should be based on the lifecycle concept, in the same way as product lifecycles.

Loans Status

The package defines the kinds of status that a loan may be in at various stages in its lifecycle, based on the more general temporal concept of statuses. These include the statuses of legal proceedings, of loan performance, of prepayment, and of the loan itself, e.g., active or in arrears.

Loan types

The main package for loan types. Loan types are Construction Loans, Commercial Loans, Credit Products, General Loans, Mortgage Loans, Personal Loans, Loan Products, Student Loans, Secured and Unsecured Loans. The package is a folder for sub-packages and doesn’t contain entities.

Construction Loans

Construction loans are loans in name only, in that the concept referred to as a construction loan is effectively a credit facility, with separate draw-downs (loans as defined in these ontologies) upon evidence of completion of agreed stages of the construction project. Note that for completion., Some basic project management terms such as milestones are already included but will need framing within more foundational concepts for project management.

Commercial Loans

Commercial loans are loans where the loan purpose is some commercial purpose. Note that these are distinguished by the loan purpose, not by the borrower type – borrowers may be corporate or personal, though in the majority of cases they would also be corporate loans that are loans with a corporate borrower.

Credit Products

Terms specific to the existence and extension of credit. Includes terms like home equity lines of credit which may be referenced in asset-backed securities. It also includes credit card facilities.

General Loans

This ontology is for different kinds of general loans such as motor loans and marine finance.   Other types of general loans such as those for motorhomes or other specified purposes would also go in this ontology, which is not complete.

Mortgage Loans

Loans that have collateral posted as security and where that collateral is real-estate and the real estate which makes up the collateral is purchased with the funds loaned. This ontology covers a range of mortgage concepts and parties, along with categories of mortgage loan purpose (remortgage, second home, etc.)   Note that much of the material in this ontology will also have been covered in the separate Loan Contracts module, and is to be integrated. The terms here are not intended not to be specific to US mortgage loans or HMDA reporting but are more general.

Personal Loans

Loans where a borrower is a natural person.

Loan Products

The ontology describes loan and mortgage loan products. A product as defined here is something marketed and offered for sale, in this case, a loan, such that when a customer takes on the product, they become signatories to a corresponding loan contract. A loan product in this sense is a contractual, off-the-shelf product with terms corresponding to those in the actual contract when it is signed.   Note that considerable work has been done on products to support a separate loan product concept in the HMDA-derived Loan Contracts module, so this work now needs integrating with that or deprecating

Secured Loans

This package provides entities relating to secured loans. All loans are either secured or unsecured, and secured loans are secured either by collateral or by the guaranty. This ontology covers these core concepts, so that mortgage loans, where the collateral is more specific, can be built on to of this. These concepts may not be needed for stand-along application-specific ontologies for mortgage lending as the concepts here provide the abstractions for the definition of mortgage loan concepts and differentiates them from other categories of loan. These intermediate abstractions may or may not be presented in the separate Loan Contracts module for mortgage loans.

Student Loans

A loan or series of loans made for the purposes of study at some institution of learning.   This ontology and much of the common supporting information on loan applications are based on extensive review and input from Sallie Mae in the US, and there may be other variants of student loans that are not covered here. For example, in principle, a student loan may be framed as a credit facility in some arrangements and as a single loan with separate payment phases in others.

Unsecured Loans

This package is a companion to the Secured Loans package and defines loans for which there is no security, either as a guaranty or as collateral.    Note that for most applications this ontology would not be needed. It is provided in order to provide a conceptually complete account of the kinds of loans there might be. While most people would not lend on this basis, the concept still exists.

Loan types / hierarchies

The diagram gallery on the Bank-ontology website shows the loan types sub-package. There are four major hierarchies, Commitments, Loan Agreements, Borrowers & Facilities, and Mortgage Purpose.

Loan Agreements

The agreements (contracts) hierachy is subtype to the Fundamental Concept defined in the Froundation package.

Loan Agreements (PNG)
Loan Agreements

The different types of contracts, in particular, Loan Contracts and Credit Facilities.

Loan Commitments

Debit and Credit Agreement Terms. Both roll up to the Financial Industry Business Concept, Commitment.

Loan Commitments

Borrower, Facility and Mortgage hierarchies

Both Borrower and Facility hierarchies have the common ultimate supertype Thing in Role. The Thing in Role is most familiar as an Autonomous Agent, but it can be any Thing that plays a role in a given context, here Loans.

The mortgage purpose hierarchy comprises of Purchase, Construction (loans), Remortgage, Renovation, Right to Buy, Investment, Debt Consolidation, Equity Release, Government Sponsored Loans, and Combination Mortgages.

Release plan, versions, and licenses

There is no official release plan for the Financial Industry Business Ontology. Again, the FIBO is an industry initiative – not a vendor product. Finalizing the Development modules, rigorous review, and QA depend on the availability of EDMC members providing subject matter experts.

FIB-DM has a major annual release. There are quarterly interim model upgrades, following the EDM Council FIBO Production schedule. Licensees with maintenance contracts get the FIB-DM version six weeks after the FIBO quarterly release.

FIB-DM core is the Open Source (GPL-3.0 copyleft) version for smaller financial institutions, academic research.

FIB-DM Commercial is the full 1968-entity Normative version including current FIBO Production modules for Securities, Derivatives, and Indices & Indicators, plus the 4563-entity Informative model. Licensees with a maintenance contract receive upgrades on a quarterly basis.

Both models are released as SAP PowerDesigner Conceptual and Logical Data Models. Most data modeling tools can import native PowerDesigner files. For other tools and uses FIB-UM, the Universal Model is the FIBO in UML-XMI.

Licenses are priced for institution size, using the EDMC membership tiers as a segment. Prices will rise in 2020 as the data model gets bigger. The maintenance contract locks in the current introductory price and entitles for 12 months of free upgrades.

Conclusion

The FIBO Development team with support from EDMC members has a firework of new Financial Services content lined up. The Financial Industry Business Ontology becomes the unchallenged standard for quality and comprehensiveness.

The Financial Industry Business Data Model makes the standard available for midsize financial institutions, without the need for specialized tools, databases and ontologists.

FIB-DM is the reference for Enterprise Data Models in Finance.

References

FIBO on the EDMC website: https://edmcouncil.org/page/aboutfiboreview

FIB-DM website: https://fib-dm.com/
Downloads of the data model, Education PowerPoints, Articles, browsable full model report, and the complete MS-Excel of PROD/DEV Packages, Entities, and Definition.

Bank Ontology website: https://bankontology.com/banking-data-model/
Focus on FIB-DM for Retail and Commercial Banks with scalable (SVG) diagrams of the Loans hierarchies.

YouTube channel with education videos: https://www.youtube.com/c/fib-dm

LinkedIn pages: https://www.linkedin.com/showcase/fib-dm/ and https://www.linkedin.com/showcase/bank-regulation-ontology/
For updates and discussion.