Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd Logo
Upload

Welcome to Scribd!

  • 99 views

    The Amex Campus Super Bowl Strategy Case Study Vfinal

    Uploaded by

    Deep Agrawal
    The document discusses creating an initial credit line assignment strategy for newly approved credit card customers to maximize long-term profitability. It states that the strategy is critical for driving profitable revenue growth while controlling risk. The strategy needs to balance assigning lines that are too low, reducing spending, or too high, increasing credit losses. Performance will be evaluated based on total long-term profitability over two years from the line assignments. Randomly assigned lines in the provided data will be used to select best performing lines for each customer segment.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    The Amex Campus Super Bowl Strategy Case Study Vfinal

    Uploaded by

    Deep Agrawal
    99 views10 pages
    The document discusses creating an initial credit line assignment strategy for newly approved credit card customers to maximize long-term profitability. It states that the strategy is critical for driving profitable revenue growth while controlling risk. The strategy needs to balance assigning lines that are too low, reducing spending, or too high, increasing credit losses. Performance will be evaluated based on total long-term profitability over two years from the line assignments. Randomly assigned lines in the provided data will be used to select best performing lines for each customer segment.

    Original Title

    The_amex_campus_super_bowl_strategy_case_study_vfinal

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The document discusses creating an initial credit line assignment strategy for newly approved credit card customers to maximize long-term profitability. It states that the strategy is critical for driving profitable revenue growth while controlling risk. The strategy needs to balance assigning lines that are too low, reducing spending, or too high, increasing credit losses. Performance will be evaluated based on total long-term profitability over two years from the line assignments. Randomly assigned lines in the provided data will be used to select best performing lines for each customer segment.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    99 views10 pages

    The Amex Campus Super Bowl Strategy Case Study Vfinal

    Uploaded by

    Deep Agrawal
    The document discusses creating an initial credit line assignment strategy for newly approved credit card customers to maximize long-term profitability. It states that the strategy is critical for driving profitable revenue growth while controlling risk. The strategy needs to balance assigning lines that are too low, reducing spending, or too high, increasing credit losses. Performance will be evaluated based on total long-term profitability over two years from the line assignments. Randomly assigned lines in the provided data will be used to select best performing lines for each customer segment.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 10

    • Once we approve a credit card application for a customer, the

    next step is to assign initial line (credit limit) to the newly


    approved lending card.

    • Line on a credit card account is the guaranteed amount till


    which a customer can have balance with us in good standing.

    • Initial line assignment strategy is extremely critical decision


    which not only drives profitable revenue growth for Amex but
    also customer engagement while controlling for our risk and
    exposure.

    • It’s a game of balance – assigning too low a line decreases


    customer capacity to spend thus reducing profitability, assign
    too high a line and risk increases exposing us to credit losses.
    Getting this balance right is the one of the most critical
    problems to solve.

    • Think about how a credit line would affect you and your
    spending and borrowing behavior.
    Create Initial line assignment strategy for customers who got approved for credit card to maximize long-
    term profitability

    • You’ll need to consider model predicted line and cover for change in profitability

    • You need to decide what line each customer should get

    Total long-term profitability (2 years) generated from line assignment will be the leading success metric for
    evaluation
    • Data provided (both development and leaderboard data) for the challenge is random assigned line data where for any
    approved applications line was assigned on random basis

    • Purpose of random assignment of line is to gauge segment level performance at different random line and select lines for
    each segment where performance (actual 2 years performance) is the best

    • Data provided has following types of attributes

    Internal Data External Data Performance Data

    Includes the internal data like product Includes the bureau data like bureau score,
    Actual 2-year performance data
    details, model predicted outputs customer attributes

    Product Details – Proprietary or cobrand,


    Bureau Score
    product fee details
    +
    + Average spend, balance, revenue, loss &
    Delinquency details on external trades
    Model predicted risk score profit
    +
    +
    Balance and Paydown on external trades
    Model predicted optimal line
    • Leverage attributes provided to define different segments

    • For each segment, evaluate performance across different random line assigned within a segment

    • While evaluating performance apart from profitability, look out for loss coverage ratio and efficiency as well
    • Profitability = Revenue – Loss

    • Loss Coverage Ratio (LCR) = Revenue/Loss :


    • Provides us information on how much leverage we have in long term over losses covering for recessionary scenario
    (where losses could double with revenue being stagnant)
    • Efficiency = Profitability/Line
    • Provides us information about how efficient is our line assignment process. Lower efficiency means we are giving higher
    line but not getting enough profit out of it.

    • For each segment, define line assignment strategy based on above analysis leveraging already provided model output line (your final
    line should be using model output line as one of the inputs)

    • Key decision points to keep in mind are customer risk, product type, model line, external line, external behavior and customer
    capacity
    This competition will run in 3 rounds :

    Round-1:
    Round-2: Round-3:
    Leaderboard
    Deck Submission Virtual Interaction
    Submissions
    • Participants will build their line assignment strategy using development data provided
    • Score the leaderboard dataset using their chosen strategy logic
    • Submit your output in form of CSV with two columns without header Unique ID and Final Line
    • The Final Line value should be from the below unique values:
    • 1000, 2000, 4000, 6000, 8000, 10000, 15000, 20000, 25000, 30000
    • Final line should match random line for at least 10% of the leaderboard data
    • Frequency of any Final line value (as given in point 1) shouldn’t be greater than 30%

    • Evaluation Metric Score: = 0.4 * Delta Profitability + 0.4 * Delta LCR + 0.2 * Delta Efficiency
    • Delta Profitability (in BPS)= (Profitability at Candidate Line - Profitability at Benchmark Line)/ (Profitability
    at Candidate Line)
    • Delta LCR (in BPS) = (LCR at Candidate Line - LCR at Benchmark Line)/ (LCR at Benchmark Line)
    • Delta Efficiency (in BPS) = (Efficiency at Candidate Line - Efficiency at Benchmark Line) / (Efficiency at
    Benchmark Line)

    • Teams will see their score post each submission.


    • Top 20 teams with highest score will be displayed on a Leaderboard
    Top 10 teams in leaderboard based on Round 1 results are eligible for this
    round.

    Submit your code with brief about your approach, considerations and any
    analysis used to finalize the campaign

    Teams will also create a presentation detailing their approach


    and insights derived.

    Evaluation Criteria – Qualitative


    ▪ Methodology & Thought process: Ideas & Approaches Tried
    ▪ Assumptions made
    • Shortlisted teams from Round 2 will qualify for Round 3 based on
    their score from Round 1 and deck submission from Round 2
    • They will be asked to present their respective deck to a panel.
    • Presentations will be evaluated based on novelty in approach, ease
    of execution of the solution and clarity in articulation.

    Top 3 teams will be selected as winners based on leaderboard


    Data score as well as qualitative and presentation evaluation

    You might also like