Skip to main content

Table 7 Weak instrumental variables tests

From: Price distortions and municipal bonds premiums: evidence from Switzerland

 

MGIV2SLS

M1IV2SLS

M2IV2SLS

Weak identification test (Cragg-Donald Wald F statistic)

5.80E + 004

7328.989

1351.68

Stock-Yogo weak ID test critical values:

10% maximal IV size

 

16.38

 

15% maximal IV size

 

8.96

 

20% maximal IV size

 

6.66

 

25% maximal IV size

 

5.53

  1. This table presents large C-D Wald F statistics on all models, and lower S-Y weak ID test critical values, rejecting the null hypothesis that the instruments in the models are weak. If any of the critical values were larger than the C-D Wald F values, then there would be at least one weak instrument in the models. Source: Own estimations