The existing research on economic policy uncertainty (EPU) is primarily concentrated on the fields of financial market volatility and economic growth [31,32,33,34,35,36,37,38,39]. It should be clarified that EPU exhibits countercyclicality, and its impact effect increases with the expansion of economic output scale; the degree of positive and negative impact on actual economic output varies by country [31]. Concerning the relationship between economic policy uncertainty and US stock returns, Arouri [32] claimed that it is nonlinear, with a decrease in stock returns leading to an increase in EPU. EPU not only lowers stock returns but also exacerbates volatility in exchange rates. For instance, Nilavongse [33] utilized a structural vector autoregressive model to claim that fluctuations in the pound exchange rate are mostly caused by changes in UK economic policies, but uncertainty in US economic policies causes a drop in industrial production capacity in the UK, and the UK’s Brexit vote caused the pound to decline significantly. Chen [34] employed quantile regression to discover that the influence of China’s EPU on its exchange rate fluctuations is asymmetric and heterogeneous, and that the impact of the US EPU on China’s exchange rate fluctuations rises with the quantile, while Hong Kong EPU has no effect on China’s exchange rate fluctuations, Japan and the EU EPU have an inverted U-shaped relationship with China’s exchange rate changes. Bank lending gives the economy a boost, while high EPU discourages investors from making credit investments. Nguyen [35], who used the generalized least squares method to evaluate the data, claims that there is a negative correlation between bank lending and EPU, with the latter having a larger negative influence on industrialized nations than emerging ones. Phan [36], who studied EPU and financial stability data from 23 countries worldwide from 1996 to 2016, does not appear to share Nguyen’s [35] point of view; Phan [36] discovered that EPU has a greater detrimental effect on financial stability for nations with smaller financial systems, lax regulation, and high levels of competition. It is clear from the foregoing that significant changes in EPU have a negative impact on economic growth, investment, and consumption; high uncertainty causes investment to decline more than output or consumption does [37]. However, tools like Bitcoin can be used to hedge EPU risks. Demir [38] discovered that EPU can forecast Bitcoin returns, even though Bitcoin returns and EPU have a negative correlation; quantile regression research results reveal that the return rates of Bitcoin and EPU are significant at lower and lower quantiles, and investors can use Bitcoin to hedge EPU risks. According to Fang [39], EPU has a negative impact when compared to Bitcoin bonds, but a positive impact when compared to Bitcoin-related commodities and equities. As a result, bitcoin can be used in hedging operations to lower the risk of EPU.

Research on how EPU affects carbon emissions has increased along with the growth of the low-carbon economy. Pirgaip [40] found that there is a one-way causal relationship between Japan’s EPU on energy consumption, the US and Germany’s EPU on carbon emissions, and Canada’s EPU on energy consumption and carbon emissions. In other words, an increase in carbon emissions will result from increased energy use, EPU, and economic expansion. Moreover, he urged the G7 nations to consider the detrimental effects of EPU on energy conservation and emission reduction, as well as to reduce carbon dioxide emissions and consumption. Huang [41] demonstrated that outward direct investment, per capita GDP, and EPU all contribute to rising greenhouse gas emissions. Adams [42] examined how countries with high geopolitical risk emitted carbon; the study’s findings demonstrate not only that rising energy consumption raises carbon emissions but also that there is a strong negative correlation between rising EPU and carbon emissions, suggesting that rising EPU will make it more difficult for nations with high geopolitical risks to maintain environmental sustainability. Additionally, some academics have examined how EPU affects energy consumption by starting with renewable energy. EPU was shown to have an impact on the utilization of renewable energy in Nakhli’s [43] study, and there is only a one-way causative relationship between US electricity consumption and EPU, as opposed to a two-way causal relationship between carbon emissions and EPU; thus, when creating environmental policies, EPU should be taken into account. As research continues to advance, more academics are concentrating on the connection between EPU and carbon emissions at the micro level. Yu [44] asserts that China’s provincial EPU have a significant positive impact on the intensity of carbon emissions and that industrial firms are more willing to employ cheap fossil fuels to satisfy the EPUs’ continually increasing standards. Luo [45] discovered a strong inverse relationship between the unpredictability of economic policy and green innovation in businesses, while this relationship can be weakened by increasing the transparency of carbon information. Khan [46] found that carbon emissions in China, South Korea, Singapore, and Japan are positively correlated with trade, GDP, and EPU, and the environmental quality of East Asian economies is enhanced by FDI and the use of renewable energy sources.

The research focus of global economic policy uncertainty (GEPU) is comparable to that of EPU, and it examines how GEPU has affected global financial and commodity future markets. Yu [47] contended that as China’s integration into the global economy progresses, the unpredictability of global economic policy will become one of the main causes of, and a substantial contributor to, the volatility of the Chinese stock market. Miao [48] was of the opinion that the addition of a regime switching model could increase the precision with which GEPU forecasts change in the US stock market. Qin [49] thought that gold could be utilized as a hedge against GEPU risks during economic downturns; gold prices fluctuate as a result of GEPU, but GEPU is also impacted by gold prices both favorably and unfavorably. Shao [50] claims that there is an imbalanced relationship between GEPU and the transmission of global grain prices, with the effects of GEPU being more beneficial for soybean prices than maize and wheat prices.

While GEPU research methods are relatively rich and unique, the present research methods on PMI are pretty straightforward. Yu [51] used the DCC-MVGARCH method to study the dynamic correlation between the PMI and GDP growth rate of manufacturing in the US. Yanik [52] examined the causal relationship between manufacturing PMI and Türkiye’s stock market through the Granger causality test. Zhang [8] used the state space equation and KALMAN filter algorithm to test the fluctuation of PMI potential signals of manufacturing industries in the US, the EU, Japan and China. The nested regression and VAR models were used to study the relation between economic activity and uncertainty by Shaikh [53]. In the latest research method on GEPU, Zhang [54] used the TVP-VAR model to study the time-varying relationship between EPU, geopolitical risk (GPR), and inbound tourism in China. Gu [55] used the TVP-VAR model to compare the dynamic impact of GPR and EPU indices on the oil market.

Although the above research clearly proves that EPU and GEPU affect the stability of stock returns [32], exacerbate exchange rate fluctuations [32,33,34], inhibit investment and output [35], and affect carbon emission intensity [41,42,43,44], the above-mentioned research methods on EPU and GEPU are mostly based on linear models such as the VAR model and its evolutions [31,32,33,40,41,42,43], quantile regression [34], the least square method [35], etc. These methods are relatively simple. Moreover, it is impossible to capture the dynamic time-varying characteristics of EPU or GEPU shocks, and it is impossible to further analyze the time-point shocks caused by crucial events in EPU and GEPU. Among the above research methods on PMI, Granger causality can only test the overall causality among variables and cannot reflect its dynamic correlation [52]; the state space equation and KALMAN filter algorithm are limited to the correlation study of two variables; the DDC MVGRACH model has relatively low accuracy in the long-term prediction of variables [51]; the VAR model can better predict multiple variables, but the conclusions drawn lack completeness and accuracy due to its failure to consider the changes in model coefficients with different periods. Although the above methods can, to some extent, reflect the dynamic interaction between variables, they cannot reflect the dynamic correlation between variables in different periods and the marginal impact effects at different time points. With the deepening of the research, the dynamic and time-varying impact of emergencies in EPU and GEPU on the economy has gradually attracted the attention of scholars. Zhang [54] used the TVP-VAR model to empirically analyze the impact of GPR emergencies such as the 911 incident and the US financial crisis on tourism but did not involve the impact of COVID-19 or the Russia–Ukraine conflict on tourism. Gu [55] also uses the TVP-VAR model, but the selected research scope was up to September 2020; thus, it cannot deeply analyze the impact of COVID-19 on international oil prices, and it does not involve the Russia–Ukraine conflict which has had an important impact on the global economy. Therefore, this study will draw on the gaps and deficiencies of the above research methods and for the first time, try to put the impact of global economic policy uncertainty on manufacturing PMI within a unified analysis framework, pay more attention to the dynamic time-varying characteristics of GEPU, and use the TVP-VAR model to analyze the time-varying impact of GEPU on manufacturing PMI in China, the US, and the EU. In this paper, we perform a comparative analysis of the impact of the US financial crisis, the European debt crisis, and especially the China–US trade war, COVID-19, the Russia–Ukraine conflict and other major global events on manufacturing. TIt also helps to compare and analyze the risk resilience of manufacturing in these three economic systems, providing relevant empirical references for industrial investors and policymakers.