Table 2.
Squared bias 
in the unit of 
, variance 
in the unit of 
, relative efficiency measured with respect to the variance of UDLM estimator, and distance of the vector of the distributed lag coefficient estimates obtained from seven statistical methods under the scenario that maximum lag 
is excessively specified. Distances are the average Euclidean distance between the vector of lag coefficient estimates and the vector of the true coefficients 
i.e. 
across 
simulation repetitions
| EB1 | HB2-EB1 | HP-EB1 | GRR | HB2-GRR | HP-GRR | BDLM | |
|---|---|---|---|---|---|---|---|
| Squared Bias | 2.22 | 1.62 | 1.22 | 0.83 | 2.13 | 1.43 | 0.68 |
| Variance | 66.16 | 62.39 | 61.77 | 18.26 | 9.79 | 10.91 | 35.80 |
| Efficiency | 1.83 | 1.95 | 1.98 | 6.54 | 10.47 | 10.13 | 3.42 |
| Distance | 0.25 | 0.24 | 0.24 | 0.13 | 0.10 | 0.10 | 0.18 |
Bold values corresponding to the best performer in each row.