Z score

Question 1

What are A,B,C,D and E in Altman’s z score?

Answer 1

A= WC/TA
B= RE/TA
C= EBIT/TA
D= MV Equity/ BV of Liabilities
E= Sales/TA

Question 2

What are X1, X2, X3 and X4 in Tafflers z score?

Answer 2

X1= PBT/CL
X2= CA/TL
X3= CL/TA
X4= CA-CL/ (Sales-PBT-Depn)/365

Question 3

What do Agarwal and Taffler (2007) study?

Answer 3

• Test the z score model from Taffler (1979) over 25 years from 79-2003
• Tested on 27243 observations of which 232 were Fs

Question 4

What tests do Agarwal and Taffler do to look at relationship between failure and z score?

Answer 4

• They compare the z score model against a simple model where if PBT < 0 then failure
• a naive model which says that no firms fail

Question 5

What do Agarwal and Taffler (2007) find from their relationship tests?

Answer 5

• Stronger relationship between z score<0 and fail than PBT<0 and fail
• Average time to failure was 13 mths and 2.4 years before failure z score was <0
• Magnitude of the negative z score has a strong relationship with the likelihood of failure

Question 6

What do Agarwal and Taffler (2007) find out about predictive ability of z-scores?

Answer 6

• Overall success rate PBT=85% and z=74%

- Success for failed companies is z=96% and PBT=68%, these are the most costly mistakes

Question 7

What do Hunter and Isachenkova (2006) examine?

Answer 7

The ability of macroeconomic variables to predict distress over and above the forecasting power of financial accounting ratios.

Question 8

What was the empirical evidence before Hunter and Isachenkova (2006)?

Answer 8

• Goudie and Meeks (1991) if exchange rates rise quickly, failure may be linked to firms exposure to exports
• Bhattacharjee et al (2002) sharp decline in £1$ rate may disadvantage domestic firms which import
• Robson(1996) unexpected changes in inflation and interest rates may lead to increase in failures as firms struggle to make interest repayments on debt.
• Bunn + Redwood (2003) as GDP decreases demand falls and failures may rise

Question 9

What are the two models Hunter and Isachenkova (2006) test?

Answer 9

1. Predicted failure- financial ratios from B1X1 up to B11X11
2. Predicted failure from the financial ratios and exchange and interest rates too

Question 10

What are the results from Hunter and Isachenkova (2006)?

Answer 10

• 1 year before failure 7 significant ratios, 2 years before 4 significant
• With Macroeconomic model both are significant 1 year before failure and 1 is 2 years before failure
• Higher R^2 for model with ACC+macro
• Acc+macro model better at classifying sample, also better at predicting failure
• Hold out sample just the Acc model is better
• 2 years before failure the Acc model is best

Question 11

What are the criticisms of Hunter and Isachenkova (2006)?

Answer 11

• 7 ratios but why those 7?
• Multiple different cut offs to choose from
• Changes in % correct
• Level of insolvency varies
• Interest rates are low
• 13 years old so not of use today
• Pre IFRS regulations
• Fair values nowadays may make it easier to predict failure and more accounting rules