An Analysis of the Daily Changes in us treasury Security Yields

Table 7: Robustness Tests (with 
Δ
VXN) 
Eq. 10 
Eq. 11 
Eq. 12 
Eq. 13 
Eq. 14 
Eq. 15 
Breusch-Godfrey serial correlation LM test 
Δ
UST2Y
 
2.055 3.819 3.152 3.898** 3.621 3.616 
Δ
UST3Y
 
0.265 1.154 1.005 1.211 1.006 1.031 
Δ
UST5Y
 
0.057 0.051 0.048 0.105 0.036 0.032 
Δ
UST7Y
 
0.238 0.005 0.002 0.008 0.009 0.011 
Δ
UST10Y
 
1.005 0.448 0.452 0.218 0.505 0.515 
Δ
UST30Y
 
0.034 0.117 0.004 0.117 0.015 0.010 
Harvey heteroskedasticity test 
Δ
UST2Y
 
2.709 6.886*** 5.114*** 6.739*** 6.422*** 6.398*** 
Δ
UST3Y
 
1.470 2.883** 1.523 2.856** 2.289 2.036 
Δ
UST5Y
 
2.479 1.004 0.554 1.248 1.449 1.205 
Δ
UST7Y
 
0.991 0.457 0.510 1.108 0.725 0.925 
Δ
UST10Y
 
0.241 0.755 0.775 0.550 0.610 0.276 
Δ
UST30Y
 
1.518 1.049 0.553 1.557 1.218 1.500 
Note: 
*** and ** represent statistical significance at the 1 percent and 5 percent level, respectively
. 

---

23 
The results from estimating equations [16] to [18] are presented in table 8. The full dataset is 
used for these regressions. The results are consistent with the earlier results. Both 
Δ
TB3M and 
Δ
CRB have positive and statistically significant effects on the daily changes of Treasury yields. 
The coefficients of 
Δ
GOLD and 
Δ
DOLLAR are negative when these coefficients are statistically 
significant. Robustness tests, provided in table 9, show that the estimated equations are mostly 
free from the problems of serial correlation and heteroskedasticity.