FINANCIAL SERVICES ANALYTICS
1. Introduction .................................................................................................................................................................. 3
2. From time series data to actionable insights ................................................................................................................ 7
3. Avoiding the “garbage in, garbage out” trap .............................................................................................................. 21
4. Supervised machine learning ...................................................................................................................................... 39
5. Prediction with time series ......................................................................................................................................... 49
6. Backtesting predictions ............................................................................................................................................... 62
7. Text as data ................................................................................................................................................................. 73
8. Robo advisory in portfolio management .................................................................................................................... 91
9. Digitalisation and payment methods ........................................................................................................................ 104
10. Digital transformation in loan and insurance decisions ....................................................................................... 118
11. Limits to data-driven decision making in finance ................................................................................................. 132
1
,2
,1. Introduction
Product oriented data science in action: KBC Kate in 2020 and KBC Kate today
Kate
- KBC’s Assistant To Ease your mind …
- Digital transformation
- Personal assistant
Goal of Kate
- Make better decisions based on advice given by Kate
What does Kate do?
- First interaction with the bank should be digital
- Timely
- Relevant
- Accuracy
Possibilities of KBC
- You can get a personalized solution (loan/insurance/…) within minutes
Development of Kate
- Quick response to common questions
o ‘Which transactions were completed? Give me an overview’
3
, o ‘What is the interest rate on a loan?’
- 14 cases which Kate should be able to do
o First date with Kate
o Money transfer
o Search transactions
o Proof of insurance/file a car claim
o Hospitalization
o Mortgage
o Kids are starting to drive
o Nudge about unused features current account
o Downsell car insurance
o Energy switch
o Damage after the storm
o Travel case
o Stop parking 4411
o Payments
Today
- Kate can do the work of 300 people
Key takeaways I
- A key skill in finance and risk is that you learn to extract business value from data for financial applications
o Very relevant and timely skill in the era of digitalisation
Solve business problems by turning them into
- Every question should start from the business
o How can we ensure that Kate answers the customers in a timely manner?
- Business solution = Kate
- Data infrastructure
o Ensures you get accurate answers on the question
- From the board room to the work floor: importance of understanding the steps that transform data into
decisions
- It creates values for all divisions:
4
1. Introduction .................................................................................................................................................................. 3
2. From time series data to actionable insights ................................................................................................................ 7
3. Avoiding the “garbage in, garbage out” trap .............................................................................................................. 21
4. Supervised machine learning ...................................................................................................................................... 39
5. Prediction with time series ......................................................................................................................................... 49
6. Backtesting predictions ............................................................................................................................................... 62
7. Text as data ................................................................................................................................................................. 73
8. Robo advisory in portfolio management .................................................................................................................... 91
9. Digitalisation and payment methods ........................................................................................................................ 104
10. Digital transformation in loan and insurance decisions ....................................................................................... 118
11. Limits to data-driven decision making in finance ................................................................................................. 132
1
,2
,1. Introduction
Product oriented data science in action: KBC Kate in 2020 and KBC Kate today
Kate
- KBC’s Assistant To Ease your mind …
- Digital transformation
- Personal assistant
Goal of Kate
- Make better decisions based on advice given by Kate
What does Kate do?
- First interaction with the bank should be digital
- Timely
- Relevant
- Accuracy
Possibilities of KBC
- You can get a personalized solution (loan/insurance/…) within minutes
Development of Kate
- Quick response to common questions
o ‘Which transactions were completed? Give me an overview’
3
, o ‘What is the interest rate on a loan?’
- 14 cases which Kate should be able to do
o First date with Kate
o Money transfer
o Search transactions
o Proof of insurance/file a car claim
o Hospitalization
o Mortgage
o Kids are starting to drive
o Nudge about unused features current account
o Downsell car insurance
o Energy switch
o Damage after the storm
o Travel case
o Stop parking 4411
o Payments
Today
- Kate can do the work of 300 people
Key takeaways I
- A key skill in finance and risk is that you learn to extract business value from data for financial applications
o Very relevant and timely skill in the era of digitalisation
Solve business problems by turning them into
- Every question should start from the business
o How can we ensure that Kate answers the customers in a timely manner?
- Business solution = Kate
- Data infrastructure
o Ensures you get accurate answers on the question
- From the board room to the work floor: importance of understanding the steps that transform data into
decisions
- It creates values for all divisions:
4
