If confidence intervals overlap in a crossed-effects model, aren't they significant?

I tried Hierarchical age-period-cohort analysis using Stata. Below is the command I used and got the random effects estimate.

• mixed dep indep || _all: R.YEAR || _all: R.COHORT, reml enter image description here

Then, I tried to plot the YEAR and COHORT results with 95% CI on the graphs. The images show that both YEAR and COHORT tend to grow, but the confidence intervals overlap.

enter image description here

Because I can see some differences between the first bars and the end bars on both graphs, I doubt whether I cannot say that there are some differences between years and cohorts.

Please help me how can interpret the data in my research.

Thanks.