Connecticut Avenue Securities: Sharing Data

10 Pages Posted: 19 Feb 2019

See all articles by Kenneth C. Jr. Lichtendahl

Kenneth C. Jr. Lichtendahl

University of Virginia - Darden School of Business

Elena Loutskina

University of Virginia - Darden School of Business

Yael Grushka-Cockayne

University of Virginia - Darden School of Business; Harvard University - Business School (HBS)

Gerry Yemen

University of Virginia - Darden School of Business

Abstract

This case uses one of Fannie Mae's credit-risk transfer instruments (CRT) to explore its data set platform and predict loan defaults through machine learning algorithms. The CRT, called Connecticut Avenue Securities (CAS), issued bonds valued on the performance of preselected pools of mortgages. The material works well to unfold natural language processing using Python. Through a three-class series, students will learn to wrangle data, experience Python, scale up to a full data set in a cloud computing environment, and use Tableau to report findings. In addition, the material allows for an analysis of the drivers of mortgage loan defaults.

Excerpt

UVA-QA-0903

Feb. 13, 2019

Connecticut Avenue Securities: Sharing Data

It is one of the most exciting innovations in the mortgage market in my career.

—Chris Hentemann, chief investment officer at 400 Capital Management

As a chartered financial analyst, Jayla Lewis worked at a large full-service investment firm in the structured securities group. The fun part of her work was creating models to perform securities analysis. The not-so-fun parts were the long, late-night hours it required. In the fall of 2017, Lewis was working with loan data to predict future default rates on Fannie Mae's single-family guaranty book of business. Fannie Mae was one of the largest mortgage buyers on the secondary mortgage market. As such, Fannie Mae had four credit-risk-transfer (CRT) instruments that offered investors an opportunity to invest in single-family mortgage-backed securities.

. . .

Keywords: code, visualize data, test hypotheses, validate model results, Brier score, variable importance plot, AI

Suggested Citation

Lichtendahl, Kenneth C. Jr. and Loutskina, Elena and Grushka-Cockayne, Yael and Yemen, Gerry, Connecticut Avenue Securities: Sharing Data. Darden Case No. UVA-QA-0903. Available at SSRN: https://ssrn.com/abstract=3335541

Kenneth C. Jr. Lichtendahl (Contact Author)

University of Virginia - Darden School of Business

P.O. Box 6550
Charlottesville, VA 22906-6550
United States

Elena Loutskina

University of Virginia - Darden School of Business ( email )

P.O. Box 6550
Charlottesville, VA 22906-6550
United States
434-243-4031 (Phone)

Yael Grushka-Cockayne

University of Virginia - Darden School of Business ( email )

P.O. Box 6550
Charlottesville, VA 22906-6550
United States

Harvard University - Business School (HBS) ( email )

Soldiers Field Road
Morgan 270C
Boston, MA 02163
United States

HOME PAGE: http://https://www.hbs.edu/faculty/Pages/profile.aspx?facId=263650

Gerry Yemen

University of Virginia - Darden School of Business ( email )

P.O. Box 6550
Charlottesville, VA 22906-6550
United States

Register to save articles to
your library

Register

Paper statistics

Downloads
2
Abstract Views
129
PlumX Metrics