Abstract: Preventing and defusing risks in the real estate sector is essential for maintaining national economic stability. The key lies in clarifying the intrinsic interrelationships among real estate risk factors and uncovering the underlying risk formation mechanisms. This paper proposes a node importance evaluation method that integrates node degree and constraint coefficients to address the limited discriminative power of traditional network analysis methods in identifying critical factors. By applying a TOP1 network to filter key interaction paths among risk factors, it reveals how risks emerge and compares the risk structures across different cities. The results indicate that consumer purchasing power and the availability of development funds are the core determinants of real estate market risk nationwide. However, city-level heterogeneity exists. In first-tier cities, risk is predominantly driven by land prices, whereas in second-tier cities, urbanization levels play a more significant role. Specifically, the key risk pathways in first-tier cities exhibit a “chain-like” structure (land price→consumer purchasing power→economic development level→land supply), while in other cities, the structure is more “star-shaped”, with consumer purchasing power as the central hub directly linked to multiple influencing factors.

Key words: real estate market risks, formation mechanism of risks, complex networks, TOP1 ranking network

摘要： 防范化解房地产风险对维护国民经济稳定至关重要，其关键在于厘清房地产风险因素的内在关联性，并揭示风险的形成机制。提出一种融合节点度与约束系数的节点重要性评价方法，以克服传统网络分析法中关键因素区分度不足的问题，进一步利用TOP1网络筛选因素间的关键作用路径，揭示风险形成机制，并对比分析不同城市间的差异。结果表明，消费者购买力和开发资金到位情况是全国房地产市场风险的核心影响因素，但存在城市异质性：一线城市风险显著受土地价格驱动，而二线城市中城镇化水平的作用更为突出。具体而言，一线城市的关键作用路径呈现“链状”特征（土地价格→消费者购买力→经济发展水平→土地供给量），而其他城市则以消费者购买力为核心呈现“星状”特征（多因素直接关联核心）。

关键词: 房地产市场风险, 风险形成机制, 复杂网络, TOP1等级网络