GET SOCIAL WITH US. #vision2016. Tweet, follow, share throughout the session.

Transcription

1 GET SOCIAL WITH US Tweet, follow, share throughout the session Experian Information Solutions, Inc. All rights reserved. 1

2 Growth through dimensional decisioning Experian and the marks used herein are service marks or registered trademarks of Experian Information Solutions, Inc. Other product and company names mentioned herein are the trademarks of their respective owners. No part of this copyrighted work may be reproduced, modified, or distributed in any form or manner without the prior written permission of Experian.

3 Introducing: Herman Jopia American Savings Bank, FSB Armando Ramos Experian

4 Outline Introduction Business context Scoring modeling and the binning code The pricing problem Closing remarks Questions and answers 4

5 Introduction What is dimensional decisioning? 5

6 In the news 6

7 What is dimensional decisioning? Dimensional decisioning is the use of a series of scores and attributes to make better decisions and grow the portfolio Layering of scores and attributes Create optimal prospect lists Increase conversion rates Build customer loyalty Increase campaign effectiveness to align with business objectives Increase customer profitability 7

8 Business context Objective and goals Banking industry / economic environment Analytics: a core asset at ASB Prescreen campaigns 8

9 Business value Objective Drive volume and profits through data and analytics Portfolio: Personal unsecured loan Channel: Direct mail (pre-approved) Optimization (2015/2016) What should we do? Predictive analytics (2013/2014) Why did it happen? What will happen? Goal +20% Goal +20% Descriptive analytics (2012) What has happened? +1.6% +9.2% Intelligence / complexity 9

10 Challenges and opportunities Economy, population, competition, skills Central Pacific Bank Others (5) First Hawaiian Bank Bank of Hawaii Analytics Skills Gap Map of University Programs in Data Analytics 10

11 Challenges and opportunities What should we do? Targeting: the right offer for the right person Others (5) Central Pacific Business as usual Bank was not going to work First Hawaiian Bank Bank of Hawaii Analytics Skills Gap Map of University Programs in Data Analytics Explore new segments (beyond traditional lending) Develop talent and build the analytics infrastructure 11

12 Leadership team Fully committed with analytics 12

13 Leadership team Example: Analytics internship / intern data analyst 13

14 Goals vs. actuals Did we make it? How? +65% +11% +20% +20% +1.6% +9.2% Understand the market Understand your business Understand your capabilities Evaluate results and models Find high value opportunities and compete wisely What are the drivers of your profits? Are you able to implement and execute your ideas? Yes, metrics but are you getting the expected results? 14

15 The market Matured and standardized What : Installment loan When: By season Who : Prime consumers How : Pre-approved offers Elvis A. Presley 1234 Ala Moana Blvd Apt 123 Honolulu, HI APR 17.99% Better targeting Scoring modeling Better offers Price optimization The right offer for the right customer 15

16 Direct mail The process Data Raw data [Experian] Rules + Scoring Model (Response) Returned mail rate Selected population Printing postage Response rate Returned mail Response No response Cost of funds maintenance Credit risk Loan Interest income Fees (-) Profitability (+) 16

17 Scoring modeling What is it? Scorecard example and definitions The development process Optimal binning R package smbinning 17

18 What is a scoring model? An algorithm that, based on known characteristics, generates a number or Score that represents the probability of a certain event Characteristics Algorithm Score 18

19 Scorecard Example: Response model Characteristics Customer Relationship Credit Score 1 Attributes or Bins No Relationship 70 No Score / Missing Months Months Months Months Net Worth Credit Score 2 No Assets 100 No Score / Missing 150 < $1, < $ 10, < $ 100, $100, Points Min Score = 300 Max Score =

20 Skills The development Decisions / stages Business case Sampling Correlation Documentation Implementation monitoring Data Binning Modeling Reporting recalibration Generation of predictive characteristics Decisions External Development Internal Credit Score 1 (CS1) Binning Model Custom Generic Budget Time Accuracy Budget Time Accuracy Budget Time Accuracy Not Likely No Score / Missing Performance chart 20

21 The development Sample dataset / binning CS1: Numeric; FResponse Binary (1: Response, 0: No Response); N = 100,000 21

22 1 Meaningful groups 2 Assumption about the relationship 3 Measure of association IV > 0.3: Strong IV < 0.1: Weak 22

23 The optimal relationship between credit score and response rate 23

24 R Package: smbinning Optimal binning for scoring modeling is Free! # Once the data is loaded in R... > result = smbinning(df=dfpultrain, y= FResponse, x= CS1, p=0.05) # Plot Response Rate > smbinning.plot(result,option= goodrate,sub= Credit Bureau... ) # Information Value > result$iv [1]

25 R Package: smbinning Optimal binning for scoring modeling Documentation References Video R Code examples Documentation 25

26 Price optimization The pricing problem Personal unsecured term loan Real life optimization Results 26

27 Pricing problem Key drivers Skills Credit model(s) Response model Expenses Income Key drivers External factors Data Printing Postage Credit risk Maintenance Cost of funds Profitability Interest income Fees Term APR Amount Principal reduction Early prepayment Price sensitivity Low interest rate environment 27

28 The product Personal unsecured loan Loan amount $10,000 Annual rate 9.99% Term 48 Months Monthly payment $ x 48 Total payment $12, Interest income $2, What if customer pays the loan on month 30? $1, What if customer pays additional 50 $/month? $1, What if customer does not pay at all? ($10,000) 28

29 Pricing problem Real life optimization RESPONSE RATE PROFITABILITY TARGET CONSTRAINT Set profitability targets (%, $) Make money! Quantify driver s contribution Set the right constraints Understand price sensitivity Test (play) and learn 29

30 Closing remarks 30

31 Summing up Build your own path to success Understand your market, business, and capabilities. Then apply PA Demonstrate the benefits of PA and earn trust It s all about results Embrace complexity: Coding saves time, a lot of time (example: smbinning ) Scoring models help to make better decisions (example: response) Apply real life optimization (moving targets, dynamic constraints) Data (lack of) is never the issue make assumptions, test, repeat Be the first in the market Business as usual will never get you where you want / need to be 31

32 For additional information, please contact: Follow us on 32

33 Share your thoughts about Vision 2016! Please take the time now to give us your feedback about this session. You can complete the survey in the mobile app or request a paper survey. Select the Survey 1 button and complete 2 Select the breakout session you attended 33

34 34