Unit 02             Credit Analytics

Learning Outcomes:

What the candidate must do:

2.1.1    Review examples of analysis and discuss their merits 

2.1.2    Discuss the difference between manual underwriting and scoring

2.1.3       Discuss causality assumptions  

What the candidate must know:

2.1.4    A model for analysis

2.1.5    What makes good analysis?

2.1.6    The difference between insight and predictive models

Learning Outcomes:

What the candidate must do:

2.2.1  Read a Gains chart

2.2.2  Calculate the optimal response rate  

What the candidate must know:

2.2.3    How to read a Gains chart

2.2.4    Understand the concept of “break even analysis”

Learning Outcomes:

What the candidate must do:

2.3.1  Calculate required sample size

2.3.2  Find a probability from a Normal distribution table

2.3.3  Discuss the impact of a small sample

What the candidate must know:

2.3.4    How to use Normal distributions to estimate likelihood

2.3.5    Understand what “Nuisance factors” are

2.3.6    The formula for sample size and error

2.3.7    Know when to adjust for a small sample

Learning Outcomes:

What the candidate must do:

2.4.1  Discuss sampling requirements

2.4.2  Review examples of tests and discuss their merits

What the candidate must know:

2.4.3    The meaning of champion / challenger tests

2.4.4    The principles of testing

2.4.5    How Design of Experiments works and when to use it

Learning Outcomes:

What the candidate must do:

2.5.1  Calculate the test statistic z

2.5.2  Identify whether a result is significant

2.5.3  Calculate z for comparing two results

What the candidate must know:

2.5.4    What a type 1 and 2 error are

2.5.5    When to test for two proportions

2.5.6    The binomial approximation for standard deviation

Learning Outcomes:

What the candidate must do:

2.6.1  Discuss when we use discrete tests

2.6.2  Calculate Chi squared and evaluate significance

2.6.3  Find a probability from a Chi squared table

What the candidate must know:

2.6.4   When to apply discrete tests

2.6.5   How the Chi squared test works

2.6.6   When Information Value is better that Chi squared

Learning Outcomes:

What the candidate must do:

2.7.1  Calculate KS

2.7.2  Calculate Gini for a scorecard

2.7.3  Calculate Gini for other ranked data

What the candidate must know:

2.7.4   The difference between ranked and other continuous tests 

2.7.5   How to interpret Gini and KS results

Learning Outcomes:

What the candidate must do:

2.8.1  Discuss the merits of drilling down

2.8.2  Calculate the contribution of a component

What the candidate must know:

2.8.3   Why we drill down

2.8.4   How to perform component analysis