The issuance of special bonds by local government plays a crucial role in supporting the construction of key national projects and promoting high-quality regional economic development. To study the risks of special bond, starting from the core-periphery theory in regional economics, this paper analyzes the unbalanced distribution of special bond risks between core and peripheral cities as well as the characteristics of risk evolution. It constructs a risk assessment system, taking the Pearl River Delta urban agglomeration as an example. The CRITIC-TOPSIS model was used to measure the risk index of each city in the region from 2017 to 2023, and the spatio-temporal pattern evolution was analyzed in combination with GIS technology. The results show that the special bond risk index of this region presents a fluctuating-contracting-expanding trend, with significant differences among cities: in the initial stage, the risk distribution was scattered; after regularization, distinct patterns emerged, with low risks concentrated in economically developed central cities and high risks in economically underdeveloped peripheral cities. The risk center shifted from the west bank of the Pearl River in 2017 to Zhaoqing, Foshan and other places in the northwest in 2023, with the overall distribution scope expanding. Markov chain analysis reveals regular transfer among the five risk levels (low, low-medium, medium, medium-high, high): low risks tend to transfer to medium risks and above, medium risks face upward pressure, and the long-term stability of high risks decreases. Based on these findings, suggestions are proposed, including improving the monitoring system and optimizing special debt management, to enhance risk management and facilitate the sustainable development of the regional economy.