In this analysis I'm going to test the term structure using the Dickey and Fuller if the series m1, m12 are integrated I(1). Check using the Engle-Granger methodology the two series are cointegrated. Estimate a VAR for the two series ordering the series as m1, m12 in the Choleski decomposition. Estimate a VAR for the two series ordering the series as m12, m1 in the Choleski decomposition. Compare the result, discussing and explaining any difference in the paths with the two different groups of IRF. Which group would you choose, and why? Simulate the Forecast Error Variance Decomposition for your model of choice and comment on the result.