dc.description.abstract
The Covid-19 pandemic accompanied by social distancing, lockdowns, and partial business closures led to a global recession in the year 2020 that extent was exceeded only by the two World Wars and the Great Depression, considering the current and the last century. Although the global automotive industry was even before Covid-19 in a highly disruptive and challenging environment, the impact of the pandemic and related after-effects changed the prevailing market conditions completely and heavily affected new passenger car registrations (NPCR). In this context, this thesis is concerned with the evaluation of the quantitative impact of Covid-19 and the resulting after-effects on NPCR in Western Europe. To establish a clear baseline for the assessment Seasonal Autoregressive Integrated Moving Average (SARIMA) models have been fitted in R to datasets of the ACEA for NPCR by countries and by manufacturers (OEMs) in a specified pre-Covid time frame. In the next step, the model with the lowest mean absolute percentage error (MAPE), in a specified pre-Covid-19 verification time frame, was selected for forecasting European NPCR of a specific country or OEM in the observed post-Covid-19 period (2020 and 2021). Hence the forecasted events can be considered as realizations of NPCR, which neglect the disruptive Covid-19 effects and consider the long-term pre-Covid-19 systematic pattern (trend, seasonality) only. As a result, this approach allows for an adequate evaluation of the quantitative Covid-19 impact by comparing the observed NPCR and the forecasted realization in the specified post-Covid time frame.In the presented numerical studies, Covid-19's impact and resulting after-effects on NPCR have been observed for 21 countries of the EU27, the EFTA, and the UK and for 21 OEMs that are selling cars in the EU14, the EFTA, and the UK. The proposed studies state results for each of the 21 selected countries and 21 OEMs and a comparison of the results concerning the months with the deepest Covid-19 impact, the speed of recovery, and the overall performance in 2020 and 2021. In summary, it can be stated that the year 2020 was characterized by a Covid-19 induced deep impact on NPCR in March and April 2020 and a subsequent strong recovery phase. In contrast, the effect of the supply chain disruptions and shortages on NPCR in 2021 was less severe in a single month but persisted the whole year. Based on the proposed evaluation approach, the overall impact on NPCR in the 21 countries (-23,7%, appr. -3,44 Mio. NPCR) and on the 21 OEMs (-25,8%, appr. -3,13 Mio. NPCR) in 2021 was even more severe than in 2020 (21 countries: -21,5%, -3,08 Mio. NPCR; 21 OEMs: -21,5%, -2,57 Mio. NPCR). It is suggested that this is a result of the semiconductor shortages, which drastically intensified in the second half of the year 2021.While some speculations on potential causes for differences in the results for the observed countries and OEMs were stated, scientific evidence of such differences concerning specific car types (fuel-based, electric, hybrid, etc.), supply chain resilience, innovative retail strategies, governmental automotive policies, and governmental responses to the pandemic, etc., could be part of future research. Another future research project could be concerned with a separation of the long-term Covid-19 effects on NPCR, and the effects attributed to the prevailing highly disruptive environment in the automotive sector (vehicle electrification, connectivity, autonomous driving, and shared mobility). Additionally, factors like governmental CO2- and NOX-emission targets, a change in mobility behavior, and increasing political tensions and trade restrictions could be considered in this respect. The capability of the proposed time series (SARIMA) forecasts to reasonably establish the pre-Covid-19 systematic patterns (trend, seasonality) of NPCR, as evidenced by their fairly good fit during the verification time frame, and their ability to highlight the Covid-19 impact and the resulting after-effects on NPCR make these forecasts suitable for more disaggregated analyses.
en
