Existing empirical studies of cross-border acquisition completion by emerging market multinational enterprises remain highly contextual, yielding inconsistent evidence regarding the determinants of deal success or failure. We apply machine learning to expose underlying complexities. The learning results of LightGBM, from data on 24,693 cross-border acquisition deals involving 29 emerging countries, unveil a comprehensive picture of the relative importance and impact patterns of 59 predictors that were fragmentally, inconsistently, or not at all presented in the extant literature. Our findings offer fresh insights into the deal completion of cross-border acquisitions by emerging market multinational enterprises, suggesting novel future research priorities.

现有的关于新兴市场跨国企业跨境收购完成的实证研究仍然具有高度的背景性，对于交易成功或失败的决定因素产生的证据不一致。我们应用机器学习来揭示潜在的复杂性。LightGBM 的学习结果来自涉及 29 个新兴国家的 24,693 笔跨境收购交易的数据，全面揭示了 59 个预测因素的相对重要性和影响模式，而这些预测因素在现有文献中是支离破碎、不一致或根本没有呈现的。我们的研究结果为新兴市场跨国企业跨境收购的交易完成提供了新的见解，提出了未来新的研究重点。