In this paper, we propose a network-based analytical framework that exploits cointegration and the error correction model to systematically investigate the directions and intensities in terms of the short-run disequilibrium adjustment towards long-run equilibrium affecting the international stock markets during the period of 5 January 2007 to 30 June 2017. Under this setting, we investigate whether and how the cross-border directional interconnectedness within the world’s 23 developed and 23 emerging stock markets altered during the entire period of 2007–2017, and two specific periods of 2007–2009 Global Financial Crisis and 2010–2012 European Sovereign Debt Crisis. The main results indicate that the magnitude of the short-run disequilibrium adjustment towards long-run equilibrium for individual stock markets is not homogeneous over different time scales. We report that the changes in directional interconnectedness within stock markets worldwide did occur under the impact of the recent financial crises. The derived networks of stock markets interconnectedness allow us to visually characterize how specific stock markets from different regions form interconnected groups when exhibiting similar behaviours, which none the less provides significant information for strategic portfolio and risk management.