Does a regression passed on time series data with one independent variable and one dependent variable which yields parameters with very high standard errors (t-values) and also a high r-squared imply that the regression results are spurious and that there is no real correlation between the variables?
The model is: $ln(y_t)=a+bln(x_t)$, $R^2>0.95$
edit: just to clarify, these standard errors arent just high but are also high in comparison to their equivalent parameter estimates.
Thank you for any help.