Master of Science in Enterprise Risk Management

Coding for Risk Management

Course Overview

Risk Management requires programming.  The tasks that we might think are specific to business analysts are becoming common throughout companies.  And, the approach of using shared Excel files across a network is being replaced with sharing database access firmwide and processing the data with SQL and code.  Even at the most senior levels, decision makers must be able to casually grab and view datasets and run scripts.  As tools emerge to  automate tasks and make analysis more friendly, facility with programming is required to interface and take advantage of the tools.  Ironically, automation requires more facility with programming.  The reason for all of this is that the benefit is so high.  Companies can find information, communicate it, and make decisions faster using automation.

Coding for Risk Management provides the knowledge that students need to thrive in today’s businesses.  The course offers a hands-on approach to studying the common tools of SQL for data gathering, Python for data analysis, R for analytics and data visualization, and Amazon Web Services for the use of Cloud infrastructure for secure and scalable infrastructure.  These tools are explored by coding up risk management concepts that appear in Market    Risk, Credit Risk, and Insurance Risk.  Students have the opportunity to learn the landscape of different syntaxes and be ready to adopt the local programming language and technical conventions of whatever firm they work at.

Learning Objectives

At the end of the course, students will be able to:

•    L1 Code up essential risk management concepts in the Python and R programming languages

•    L2 Adhere to and opine on best coding practices

•    L3 Decide on which languages are best for different tasks

•    L4 Rapidly adapt to and learn new syntaxes

•    L5 Query internal corporate databases and external web resources to gather and organize data

•    L6 Visualize data and create interactive dashboards for decision making

Readings

Forta, Ben. (2019). SQL in 10 Minutes a Day. Sams Publishing; 5 edition (December 20, 2019).

Lander, Jared P. (2017). R for Everyone. Addison-Wesley Professional; 2nd edition (June 18, 2017).

Yan, Yuxing. (2017). Python for Finance 2nd Edition. Packt Publishing 2 edition (June 30, 2017).

Yan, Yuxing. (2018). Financial Modeling using R 2nd Edition. Legaia Books USA; 2nd edition (January 18, 2018).

Assignments and Assessments

Weekly programming assignments will enable students to immerse themselves in different programming languages and styles of coding.  The assignments will be graded without partial credit; students will need to meet the challenge of producing successful code that accomplishes the assigned tasks.  The final exam is a repeat of these coding challenges.  Small case studies will allow students to work on more complete programming projects.  Case studies may very across different semesters.

Here are examples:

Case Study: FRTB Capital Estimation:

Estimate the capital requirement for a small bank based on the bank’s trading data held in a SQL database.

Case Study: Merton Model Default Risk Estimation:

Estimate the Default Risk of a company using Merton’s model

Case Study: Altman’s Z Score model of credit risk based on a balance sheet.

Refit the classic Altman’s Z Score model for data relevant to a specific industry.

Case Study: Agent Based Modeling.

Use an agent-based modeling framework to model the behavior. of economic agents.

Case Study: CDO Default Risk Estimation:

Use copula modeling functionality in R to estimate and backtest correlated defaults.

Case Study: Counter-party Credit Risk Modeling:

Estimated the counter-party credit risk across a firm by gathering its trade data from a databased and building appropriate model for each asset type.

Case Study: Lending Data:

Grab Lending Clubs default data and fit a model to it.

Grading

The final grade will be calculated as described below:

FINAL GRADING SCALE