Missing Forecasts: What Went Wrong?*



Case Code : CLMISC012
Publication date : 2010
Subject : Miscellaneous
Teaching Note : Available
Industry : Information Technology
Length : 02 Pages
Price : Rs. 100

To download this case click on the button below, and select the case from the list of available cases:

Miscellaneous Short Case Studies
Miscellaneous Case Studies **
ICMR Case Collection
ICMR Courseware
View Detailed Pricing Info

Key words:

Regression analysis, Model specification, strategy, Null Hypotheses, ROCE, F stat, ANOVA, t stat, multi-collinearity, autocorrelation


* This caselet is intended for use only in class discussions.
** More comprehensive case studies are priced at Rs.200 to Rs.700 (US $5 to US $16) per copy.











ICMR India ICMR India ICMR India ICMR India RSS Feed

This caselet is about the decision of a small software company ITsol to change its strategy and to decide which head it should invest in so to improve its financial performance. It used a regression analysis for this purpose and based on the results, made certain changes to its strategy. However, the company failed to meet its forecasts. The caselet provides an opportunity to students to analyze the results and discuss what went wrong.


Understand the issues related to regression analysis
Analyze and interpret the regression results correctly
Appreciate the importance of missing variables (specification of model)
Analyze the pros and cons of Regression analysis
Understand why Regression analysis should not be used when the model is not correctly specified
Discuss and debate the issues related to the ITsol CEO's decision to change its strategy
Explore ways in which ITsol could have used the results properly


Sandeep Singh (Singh), the CEO of ITsol India Ltd. (ITsol), was wondering what had gone wrong. His company had missed its forecasts for the fiscal year 2009-10, the third consecutive year that it was doing so...

Questions for Discussion:

1. What are the Null Hypotheses for β's? If there is a 5 unit change in D/E ratio what kind of effect will it have on ROCE?
2. The value of R-square is quite low yet the model is significant. What could the reason be for that?
3. If some more variables are added in the model, what would the change be in the value of R- square?
4. Why is the F stat value used in the ANOVA table rather than the t stat value?
5. What are the possible reasons for ITsol's strategy failure?
6. What is Model specification and why it is so important?
7. What would your decision have been had you been the CEO of ITsol and why?
8. What is multi-collinearity and what are its effects on results?
9. What is Autocorrelation and why it is useful for the analysis of results?

Cases on Related Topic:
1. Avant Garden: Exploring Possibilities of Improving Revenue Management