ARDL bug in EViews 14 ?

[Bob]

Dear EViews team,
Using the ARDL procedure in EViews 14, the estimation runs normally and the cointegration equation is correctly estimated, but all ARDL coefficients are reported as zero (t = 0, p = 1) !
Best regards

for exemple :
Dependent Variable: RATE_10Y
Method: ARDL
Date: 02/09/26 Time: 00:04
Sample: 2003M04 2025M09
Included observations: 270
Lag selection: Automatic (deplags=12, reglags=0)
Selected model: ARDL(3,0,0) using AIC (12 models evaluated)
HAC standard errors & covariance (Bartlett kernel, Newey-West fixed
bandwidth = 5.0000)

Variable Coefficient Std. Error t-Statistic Prob.*

Distributed-lag Regressors

Dependent
RATE_10Y(-1) 0.000000 0.086579 0.000000 1.0000
RATE_10Y(-2) 0.000000 0.150383 0.000000 1.0000
RATE_10Y(-3) 0.000000 0.078682 0.000000 1.0000

Independent
INF 0.000000 0.003984 0.000000 1.0000
CONV_PM 0.000000 0.005289 0.000000 1.0000

Deterministic Regressors

C 0.000000 0.013360 0.000000 1.0000

R-squared 0.998085 Mean dependent var 2.990556
Adjusted R-squared 0.998049 S.D. dependent var 1.180612
S.E. of regression 0.052149 Akaike info criterion -3.047465
Sum squared resid 0.717943 Schwarz criterion -2.967500
Log likelihood 417.4078 Hannan-Quinn criter. -3.015355
F-statistic 27521.89 Durbin-Watson stat 1.980610
Prob(F-statistic) 0.000000

*Note: p-values and any subsequent test results do not account for model
selection.




Deterministics: Rest. constant (Case 2)

Variable Coefficient Std. Error t-Statistic Prob.

INF 1.136709 0.957254 1.187468 0.2361
CONV_PM 0.494911 0.475037 1.041836 0.2984
C 0.438690 1.450053 0.302534 0.7625

[EViews Gareth]

Build date of EViews? (Help->About EViews)

[Bob]

Jan 5 2026 and now Feb 19 2026 with the same problem !

[EViews Gareth]

Could you provide the workfile?

[Bob]

for exemple

[igor]

Funny case: Estimation - Option - Coefficient name (return C instead of ardl). All will be fine

[EViews Gareth]

Ah, yeah, should be an easy fix for us - next patch.

[Bob]

ok ok ;)