统计与数据科学学院知识库

在全球经济增长放缓和内外部冲击叠加的背景下，全球经济面临增长下行与通胀波动的双重挑战。这不仅加剧了通胀不确定性和经济下行风险，还削弱了宏观政策的传导效率，制约了经济的可持续发展。当前，中国经济正经历增长下行与通胀波动的非线性互动，对经济高质量发展构成挑战。为实现“稳增长”与“稳物价”的经济目标，亟需精准评估宏观经济尾部风险的程度及其关键影响因素，并系统分析货币政策的调控效应，以提升政策决策的前瞻性和有效性。

鉴于此，本文构建了“在险测度—影响因素—动态关联—政策调控”的系统研究框架，以评估中国在险通货膨胀（Inflation at Risk，IaR）和经济在险增长（Growth at Risk, GaR），并揭示其关键影响因素及货币政策的非线性调控效应。主要研究内容如下：第一，构建基于高维数据的在险测度体系。采用分位数因子模型（Quantile Factor Model, QFM）整合高维经济金融数据的横截面与分位数信息，构建涵盖尾部预测值、在险概率、预测尾部均值及分布高阶矩的综合在险评估体系，以弥补传统因子模型对经济变量厚尾分布特征刻画不足的问题。第二，“分位数—时频域”双维度风险影响因素分析。结合分位数回归分析与非参数尾部依赖方法，揭示影响因素与IaR和GaR的尾部依赖模式及持久性。同时，利用时频域小波多元分析方法，在短、中和长期不同周期上识别关键风险源。第三，IaR与GaR的动态关联分析。突破已有仅关注均值研究的局限，采用带有随机波动率的时变参数向量自回归（TVP-SV-VAR）模型探讨极端情景IaR与GaR的时变交互机制，并利用分位数向量自回归方差分解网络（QVAR-DY）刻画不同经济情景下的风险溢出路径与网络结构。第四，货币政策的非线性调控效应分析。基于局部投影（Local Projections，LP）方法，研究货币政策在不同经济状况、高低通胀区制及不同规模政策冲击下的非线性效应。系统评估其对IaR与GaR的影响机制，并揭示不同风险组合下的政策协同与非对称效应。

主要研究结论显示：第一，高维数据下的在险评估。（1）QFM捕捉了数据的异质性结构，使在险测度框架更精准地刻画经济增长和通胀的极端波动情景；（2）中国GaR展现较强韧性，中长期平稳，短期波动较大，2012年后经济“换挡”特征显著，IaR则呈现阶段性波动；（3）IaR与GaR均具有显著的“事件驱动”特征，金融危机、COVID-19疫情及俄乌冲突等重大事件均对其产生冲击；（4）通胀与经济增长的在险概率与均值之间存在“阈值效应”。第二，在险因素探源研究。（1）金融状况指数（Financial Conditions Index，FCI）负向影响经济增长，投资和消费起到经济稳定器作用，而金融紧缩及外贸依赖加剧经济下行风险；（2）COVID-19重塑IaR和GaR与其影响因素之间的尾部依赖关系；（3）IaR短期受通胀预期和外部市场波动驱动，中期受FCI、国际大宗商品价格和全球产出缺口影响，长期受FCI和成本输入型推动；（4）GaR短期受投资和房地产驱动，中期受消费、房地产、进出口和投资的多重影响，长期受财政、能源及消费等结构性因素影响。第三，IaR和GaR的时变关联分析。（1）IaR对其他经济变量的影响呈现“短期>中期>长期”的衰减趋势；（2）GaR冲击对IaR有促进作用，但疫情后，经济增长对通胀的推动效应减弱；（3）极端风险情景下，系统风险传导效应最强，网络关联密度最高；（4）系统风险的传递具有时变性和周期性，重大事件会强化风险溢出效应。第四，货币政策的非线性调控效应。（1）价格型政策对通胀调控更有效，数量型政策短期内促进经济增长，温和的数量型政策效果更显著，且政策冲击在高、低通胀区制具有非对称性，通胀在高区制更敏感，经济增长在繁荣期的政策响应强于衰退期；（2）单一货币政策难以有效应对经济尾部风险，货币与财政政策配合可显著提升调控精准性和有效性；（3）适度宽松的财政政策提高货币政策传导效率，降低经济衰退风险；（4）一致性目标的政策组合能有效缓解通胀波动与经济下行风险。

基于上述结论，本文提出“数据监测、风险防控、协同联动”三层次政策预警与防控体系。首先，在数据监测方面，构建全面的经济风险监测框架，强化金融市场、房地产市场、国际大宗商品价格波动及外部冲击的监测，提升极端风险情景下的动态预警能力，并建立金融市场与宏观经济的联动监测体系。其次，在风险防控方面，动态调整经济风险评估策略，优化人民币汇率及大宗商品价格应对措施，并强化跨市场联动监管，防范系统性风险。最后，在协同联动方面，提高货币与财政政策目标一致性，增强政策的非线性调控能力，强化国际经济政策协调，深化“一带一路”倡议合作，提高经济抗风险能力。

Admist the backdrop of slowing global economic growth and compounded internal and external shocks, the global economy faces dual challenges of economic downturn risks and inflation volatility. These challenges not only heighten uncertainty in inflation and amplify downside risks to growth but also weaken the transmission efficiency of macroeconomic policies, thereby constraining sustainable economic development. At present, China’s economy is experiencing a nonlinear interaction between economic slowdown and inflation fluctuations, posing significant challenges to high-quality development. To achieve the dual objectives of “stabilizing growth” and “maintaining price stability,” it is imperative to accurately assess the extent of macroeconomic tail risks and their key driving factors while systematically analyzing the regulatory effects of monetary policy. This will enhance the forward-looking nature and effectiveness of policy decisions.

In response to these challenges, this study constructs a systematic research framework encompassing “tail risk measurement—driving factors—dynamic interactions—policy regulation” to assess China's Inflation at Risk (IaR) and Growth at Risk (GaR). It aims to identify key influencing factors and analyze the nonlinear regulatory effects of monetary policy. First, a tail risk measurement system based on high-dimensional data is developed. The quantile factor model (QFM) is employed to integrate cross-sectional and quantile information from high-dimensional economic and financial data. This approach establishes a comprehensive tail risk assessment framework, incorporating tail forecasts, risk probabilities, predicted tail means, and higher-order distribution moments. It addresses the limitations of traditional factor models in capturing the heavy-tailed distribution characteristics of economic variables. Second, a dual-dimensional risk factor analysis is conducted by combining quantile regression analysis with nonparametric tail dependence methods to uncover the dependence structure and persistence of key factors influencing IaR and GaR. Additionally, a wavelet-based multivariate analysis is employed to identify critical risk sources across short-, medium-, and long-term economic cycles. Third, the dynamic interactions between IaR and GaR are explored. Moving beyond the traditional focus on mean-based analyses, this study employs a time-varying parameter vector autoregressive model with stochastic volatility (TVP-SV-VAR) to examine the evolving interaction mechanisms between IaR and GaR under extreme economic conditions. Furthermore, a quantile vector autoregression with Diebold-Yilmaz spillover index (QVAR-DY) is employed to map risk spillover pathways and network structures across different economic scenarios. Fourth, the nonlinear effects of monetary policy regulation are analyzed. Using the local projections (LP) method, this study examines the nonlinear effects of monetary policy under varying economic conditions, inflation regimes, and policy shock intensities. It systematically evaluates the transmission mechanisms of monetary policy on IaR and GaR while identifying the coordination and asymmetry of policy responses under different risk scenarios.

The main findings of this study are as follows: First, tail risk assessment based on high-dimensional data. (1) QFM effectively captures the heterogeneous structure of data, enabling a more precise depiction of extreme fluctuations in economic growth and inflation. (2) China’s GaR exhibits strong resilience, remaining stable in the medium to long term but experiencing greater short-term volatility. After 2012, the economy demonstrated a distinct structural transition, while IaR exhibited phase-specific fluctuations. (3) Both IaR and GaR are significantly influenced by major events, including the global financial crisis, the COVID-19 pandemic, and the Russia-Ukraine conflict, which have all triggered substantial economic impacts. (4) A threshold effect exists between the tail risk probability and the mean of inflation and economic growth, suggesting nonlinear dynamics in risk transmission.

Second, identification of key risk factors. (1) Financial condition index (FCI) negatively impact economic growth, while investment and consumption serve as economic stabilizers. However, financial tightening and increased reliance on foreign trade elevate downside economic risks. (2) The COVID-19 pandemic reshaped the tail dependence between IaR, GaR, and their influencing factors, altering their interdependencies. (3) Short-term IaR is primarily driven by inflation expectations and external market fluctuations, while medium-term IaR is influenced by FCI, global commodity prices, and the global output gap, and long-term IaR is largely shaped by FCI and imported cost pressures. (4) Short-term GaR is driven by investment and the real estate sector, while medium-term GaR is affected by consumption, real estate, trade, and investment, and long-term GaR is shaped by fiscal policies, energy dynamics, and structural consumption patterns.

Third, time-varying interactions between IaR and GaR. (1) The influence of IaR on other economic variables follows a decaying pattern: short-term > medium-term > long-term. (2) GaR shocks enhance IaR, yet post-pandemic, the impact of economic growth on inflation has weakened. (3) Under extreme risk scenarios, systemic risk transmission intensifies, leading to the highest network connectivity density. (4) Systemic risk transmission exhibits time-varying and cyclical characteristics, with major events amplifying risk spillover effects.

Fourth, the nonlinear regulatory effects of monetary policy. (1) Price-based policies are more effective in controlling inflation, while quantity-based policies stimulate economic growth in the short term, with moderate interventions yielding stronger effects. Moreover, policy shocks exhibit asymmetry across high- and low-inflation regimes, where inflation is more sensitive in high-inflation periods, and economic growth responds more strongly during economic booms than in recessions. (2) A single monetary policy is insufficient to mitigate tail risks, whereas coordinated monetary and fiscal policies significantly enhance regulatory precision and effectiveness. (3) Moderate fiscal easing improves the transmission efficiency of monetary policy, reducing the risk of economic downturns. (4) Policy coordination with aligned objectives effectively stabilizes inflation volatility and mitigates downside economic risks.

Based on the above findings, this study proposes a three-tier policy framework for data monitoring, risk prevention, and coordinated policy response to enhance economic resilience and stability. First, in terms of data monitoring, a comprehensive economic risk monitoring system should be established to strengthen surveillance of financial markets, the real estate sector, international commodity price fluctuations, and external shocks. This would improve dynamic early warning capabilities under extreme risk scenarios and facilitate the development of an integrated monitoring mechanism linking financial markets with the broader macroeconomy. Second, in risk prevention, economic risk assessment strategies should be dynamically adjusted to optimize responses to fluctuations in the renminbi exchange rate and commodity prices. Additionally, cross-market regulatory coordination should be reinforced to mitigate systemic financial risks and enhance economic stability. Finally, in coordinated policy response, greater synergy between monetary and fiscal policies should be pursued to enhance the nonlinear regulatory effectiveness of macroeconomic policies. Furthermore, international economic policy coordination should be strengthened, particularly through deeper engagement in the Belt and Road Initiative, to enhance global economic cooperation and bolster China’s resilience against external risks.