This section offers some PowerPoint Slides, Word Documents, Excel Spreadsheets, and items not otherwise categorized as SAS Code or VB6 Code. As always, I have to say that these downloads are without warranty and I accept no liability for their use or installation.

This is an excellent discussion of credit scoring techniques covering topics such as logistic regression, probit regression, neural nets, CART, incorporating economic indictors in models, profitability scoring, and behavior scoring. At the end of the article, the author provides a comprehensive list of references.

This is a program written in SHAZAM that estimates a Pooled Cross Section Time Series Model and simulates the result of a change in employment. The example used does not represent real data, but is shown for illustration purposes only.

This is a summary manual for LifeCast Pro for new product forecasting. Even if you are not interested in purchase, the information provided here may prove interesting.

This is a good 2004 article about forecasting new product launches using diffusion theory.

This is an upgrade to LifeCast Pro software. It requires a previous version to be already loaded on your computer.

This is an Excel file showing you how to develop the Bass model in a spreadsheet. The Bass Model is often used for forecasting a new product or service.

This document discusses an approach to new product forecasting using the Bass Diffusion Model and a simulation procedure through Logistic Regression.

If you have a background in regression and want to learn about neural nets, this paper is for you. It gives a discussion of neural nets from an econometrics perspective through a question and answer format. Very useful.

This is an informative discussion of credit modeling that serves as a good introduction to different best practices in the industry.

This is a powerpoint presentation discussing new product forecasting techniques. It was presented at the IBF conference in 2002.

Here is a dictionary of forecasting terms that is great for including in documentation and reports.

This is a discussion of a technique often used in credit scoring to enhance model accuracy. The paper serves as a good introduction to the subject.

This paper presents some research that shows reject inferencing is not that bad a problem if you can increase your sample size as is possible in credit card modeling. This is an excellent paper and worth reading with practical implications concerning reject inferencing.

Here is an Excel spreadsheet showing the transformation process where you can more appropriately apply linear regression. Some say you can use this method where the dependent variable is clustered around upper or lower boundaries. This is seen by some as an alternative to using a Tobit Model. This procedure is used by Moody's to estimate their LOSSCALC recovery model.

Here is a great summary of the Basel II regulatory requirements that will be imposed on many of the larger banks in the U.S. and worldwide, scheduled to go into effect in 2007. It requires banks to risk rate ALL their loans in order to reduce their capital requirements.

Very good summary of History of Credit Risk - mostly consumer based.

Good link to Economic data for Modeling - FREE!

Link to SAS code for Bootstrapping

Here is a method for numerically calculating the area under the ROC curve which is often used to determine the accuarcy of credit scoring models.

Here is a paper that describes how to use SAS to dynamically write out SAS code. I have put example code of this type in my SAS download section, but this paper should give a more detailed discussion of this very important concept.

This is a powerpoint presentation on how to integrate logistic regression with C.A.R.T. modeling. Makes sense!

This is a good practical discussion of Survival Analysis in SAS.

This is another good practical discussion of Survival Analysis in SAS. Focus here is Telecommunications.

SAS has a new procedure called PROC QLIM which can be used for a number of different types of models including discrete choice and Tobit Regression.

This is a powerpoint presentation describing a modeling platform combining VB6 and SAS. This was presented in St. Pete Florida at the 2003 SEUG conference in September.

This is a paper presented in the RMA Journal on the role of Economic Data and its usefulness in Credit Modeling (by Jiong Liu & Xiaoqing Eleanor Xu).

This is the programming code, data file, and output for a double sided tobit model in SHAZAM which will match the output from the SAS double sided tobit model using PROC QLIM that can be obtained from my SAS download page.

This is a link to SHAZAM programs written to illustrate William Greene's textbook 'Econometric Analysis' published in 2000. Great resource for econometric analytics!

Good article on portfolio stress testing.

Good article on portfolio stress testing.

Using Pooling in Portfolio Stress Testing.

Vintage Analysis for Mortgage Loans.

Economic Modeling in Risk Models.

This article compares PD modeling vs. Portfolio modeling using economic data.

This article reviews different loss forecasting techniques such as roll rates, econometrics, vintages.

This article is a good nontechnical overview of modeling credit data over the entire lifecycle.

This article is a good technical discussion of modeling Basel PD, LGD, EAD and stress testing. I personally do not agree with trying to incorporate economic data directly in those models, but here is another view.

Good summary of Basel as of 2011.

You can now do logistic regression in Excel.

A good article by the conference board on the development of a leading credit index - soup to nuts.

Great article examining what bad things can happen if you add a lagged dependent variable into your regression equation. It can swamp your other variables if there are trends.

Good article looking at the impact of log transformations in forecasting accuracy. End result - no harm done.

The original article on the subject in 1974. Different schools of thought exist, but they are leaning towards differencing today.

This is a VBA spreadsheet program which allows the user to interactively produce and collapse bins for modeling purposes. When you pull it up, resave as a .xlsm file. It also contains a Correlation macro using Excel's Data Analysis Toolkit. The macro computes pairwise and bivariate correlations and warns users if multicollinearity problems might exist given certain correlation thresholds.

This is a VBA spreadsheet program which allows the user to interactively produce and collapse bins for modeling purposes. When you pull it up, resave as a .xlsm file.

Article discusses that when looking at long run trends, it is important to test for stationarity. Examples include GDP, Industrial Production, and Employment.

This is a handy spreadsheet I found that shows you the Excel calculations to test the difference between 2 proportions.