Developing countries typically integrate into the world economy by first opening up to trade and then later, if at all, by integrating their capital markets. I study the effects of postponing the opening of capital markets in a standard trade model with financial frictions and firm dynamics. As trade barriers fall, the model predicts that capital misallocation declines in the aggregate, but increases among exporters. Allowing capital inflows helps all firms but it also magnifies the losses from misallocation. In the quantitative experiment calibrated to the Hungarian integration episode of the 90s, the benefit of cheaper capital dominates the adverse effect of growing capital misallocation on productivity, leading to higher output, consumption, and welfare than under closed capital markets. Moreover, Hungary could have gained an extra 1 % in welfare, on top of the overall gain of 7 %, by immediately allowing capital inflows after the reduction in trade barriers.