Theoretical and empirical studies on persistent high growth have largely focused on the extent to which companies are able to sustain their high growth rates over time and whether there are firm-level determinants of this behavior. These studies, however, neglect the question of whether persistent high-growth firms exhibit distinct growth patterns. To analyze this question empirically, we examine a database of approximately 16,000 persistently high-growth companies located in Brazil during the period 2013–2016. First, we develop a clustering analysis, finding three high-growth patterns. Based on this classification, and through ordered logit regression, we then examine whether regional dimensions explain the differences in growth. The results indicate that companies located in regions with more availability of technological, educational, and physical infrastructure are more likely to be in the groups with the highest growth. On the other hand, the increase in regional sophistication in economic terms tends to reduce the likelihood of very high growth. This specific result is confirmed when we perform quantile regressions.

中文翻译：

---

持续高增长的公司之间是否存在异质性？

关于持续高增长的理论和实证研究主要集中在公司能够在多大程度上维持其高增长率，以及这种行为是否存在公司层面的决定因素。然而，这些研究忽略了持续的高增长公司是否表现出不同的增长模式的问题。为了实证分析这个问题，我们检查了一个数据库，该数据库包含 2013-2016 年期间位于巴西的大约 16,000 家持续高增长的公司。首先，我们进行聚类分析，发现三种高增长模式。基于这种分类，并通过有序 logit 回归，我们然后检查区域维度是否解释了增长的差异。结果表明，位于技术、教育、和物理基础设施更有可能属于增长最快的群体。另一方面，从经济角度来看，区域复杂性的提高往往会降低非常高增长的可能性。当我们执行分位数回归时，这一特定结果得到了证实。