Annex 1
The Relationship between CPI-Combined and WPI
The relationship between WPI and CPI was examined based on monthly data from January 2000 to December
2013, for Food and Core (non-food non-fuel) parts separately.
For this, the variables were initially seasonally adjusted using X-12-ARIMA technique. The inter-relationship
was tested using Granger Causality and SVAR framework. First difference of the log of seasonally adjusted
data was used.
A. Investigating the relationship between CPI-Combined Food and WPI-Food |
Dependent variable |
Independent variable |
Test statistic |
Remark |
Dlog(WPI_food) |
Dlog(CPI_food) |
9.62 |
Bidirectional causality |
(0.02) |
Dlog(CPI_food) |
Dlog(WPI_food) |
8.58 |
(0.04) |
Figures in parentheses are p-values. The optimum lag length is 2.
Conclusion: Impact of an increase in WPI-food inflation on CPI-Combined food inflation is significant and
leads to increase in CPI-Combined food till two months, while an increase in CPI-Combined food inflation
leads to a corresponding increase in WPI-food inflation.
B. Investigating the relationship between CPI-Combined Core and WPI-Core
Here core is defined as headline CPI/WPI excluding Food & Fuel.
Dependent variable |
Independent variable |
Test statistic |
Remark |
Dlog(WPI_Core) |
Dlog(CPI_Core) |
3.29 |
No evidence of existence of causality between WPI-Core and CPI-Core. |
(0.35) |
Dlog(CPI_Core) |
Dlog(WPI_Core) |
3.23 |
(0.36) |
Figures in parentheses are p-values. The optimum lag length is 3.
Conclusion: No significant impact of WPI-Core on CPI-Combined Core and vice versa.
|