Travel restrictions and social distancing imposed to curb the spread of the new coronavirus have been strongly hitting tourism since March 2020. Tourism forecasting literature addressed the effects of shocks in contexts characterized by a predictable route to recovery. COVID-19 is without precedents. In this article, monthly overnight stays for the period January 2010 to December 2020 are used to estimate the impact of the pandemic in Lombardy, Italy’s most affected region. A model-based approach is implemented, and the number of overnight stays up to December 2023 is forecasted. Four models are compared. Estimation results from an augmented SARIMA model suggest that, provided a new lockdown is averted, the domestic segment will recover in a relatively short period of time, whereas international tourism might need an external intervention to speed up its recovery process.

自 2020 年 3 月以来，为遏制新型冠状病毒的传播而实施的旅行限制和社会疏远一直在强烈打击旅游业。旅游业预测文献在以可预测的复苏路线为特征的背景下解决了冲击的影响。COVID-19 是没有先例的。在本文中，使用 2010 年 1 月至 2020 年 12 月期间的每月过夜住宿数来估计疫情对意大利受影响最严重的伦巴第大区的影响。实施基于模型的方法，并预测到 2023 年 12 月的过夜人数。比较了四种模型。增强型 SARIMA 模型的估计结果表明，如果避免新的封锁，国内市场将在相对较短的时间内恢复，