Starting at the Top
Setting Up the Problem
Considering Multiple Vintages
Volatility Analysis and Economic Capital
Credit Scores and Account Management
Analysis of the US Mortgage Crisis
An Example Using [email protected] Data
Examples of Modelling Vintages
Epilogue: it’s about time
The methods, insights and examples in this book can all be summarised with one word: time. Why do we need to discuss time so much and look at long-range forecasting, stress testing, capital and scores that adapt to the economy? In the “good old days” if someone looked like a good credit risk, you gave them a loan and collected the fees. Why was life so much simpler then?
The basic problem is that the financial world is much faster paced, competitive and more complex than it was in the 1960s when credit scores were invented or than in the 1980s when roll-rate models became commonplace. An average of past performance is no longer sufficient for managing a financial institution. Every activity needs to be forward looking and that means we no longer have the luxury of ignoring time.
One side effect of failing to incorporate time in our models is the confusion between past and future. As lenders become ever more competitive, they naturally must operate with thinner margins and look for small advantages. When models fail to properly incorporate time, lenders can be fooled into seeing opportunities that no longer exist or are too volatile to be worth the investment.