The SARS-CoV-2 outbreak made clear the urgent need to depart from traditional statistics, typically released with a lag and available at a relatively low frequency. This led to unparalleled efforts to put forward high-frequency indicators to track economic developments timely. By resorting to non-traditional data sources, we propose a novel daily economic indicator to track economic activity in Portugal. It corresponds to the latent variable of a set of daily series within a factor model framework. We find a sudden and sharp drop in economic activity in mid-March 2020, when the lockdown of several activities was declared due to the COVID-19 pandemic. Since in this approach we address the complexities of high-frequency data without further smoothing, we are able to identify sudden changes of economic activity in a timely and daily manner in contrast with other approaches.

SARS-CoV-2 的爆发明确表明迫切需要摆脱传统统计数据，传统统计数据通常发布滞后且提供频率相对较低。这导致了前所未有的努力提出高频指标来及时跟踪经济发展。通过利用非传统数据源，我们提出了一种新颖的每日经济指标来跟踪葡萄牙的经济活动。它对应于因子模型框架内一组每日序列的潜在变量。我们发现 2020 年 3 月中旬经济活动突然急剧下降，当时由于 COVID-19 大流行而宣布封锁多项活动。由于在这种方法中，我们无需进一步平滑即可解决高频数据的复杂性，因此与其他方法相比，我们能够及时、每日地识别经济活动的突然变化。