We propose a Bayesian dynamic interbank network model where three mechanisms control the likelihood of a trade between two banks: (i) a time-series interbank activity index that expresses overall confidence of interbank across time, (ii) bank-specific latent variables describing banks’ tendency to be borrowers or lenders, and (iii) covariates characterizes pairwise past relationships. To validate the accuracy of the likelihood of trading, we predict pairwise future trades which we achieve an AUC of 90%. We find that the likelihood of trade is a novel pairwise characteristic that is an important determinant of banks’ ability to access interbank market liquidity with regression analysis. We conclude that 1) Banks tend to lend to banks having a higher likelihood to trade with a lower interest rate in both pre-crisis and crisis time. 2)Though the market interest rate increases in crisis time, banks offer a more favorable rate to those who have a higher likelihood to trade with.