International trade is crucial for meeting national demand, enhancing efficiency, and stimulating innovation. Selecting appropriate trading partners is paramount, impacting economic benefits and long-term development. Traditional methods like comparative advantage and trade costs are often insufficient. This study proposes using economic complexity, which measures a country's economic development and product capacity more precisely, enabling future-oriented partner selection. The increasing tensions between China and the US create uncertainty in the global trade environment, making reassessment of trade strategies necessary. Traditional methods such as gravity models focus on bilateral trade and lack macro-level data-driven approaches. Economic complexity, based on endogenous growth theory, offers a more effective way of predicting economic growth than classical indicators, and has been successfully applied in various fields. This study utilizes economic complexity indexes (ECI and PCI) and product proximity algorithms to analyze Chinese and US trade networks, aiming to provide a more comprehensive and accurate framework for selecting trading partners, considering both the current and future state of economic development. The study contrasts the world trade network with those of China and the US markets to identify strategic variations in partner selection and product diversification.

Existing research on economic complexity primarily focuses on bipartite trade networks, often neglecting partner selection. Studies using exporter-product bipartite networks reveal the current economic status and potential from a global perspective. Other research focuses on national or regional trade, providing insights into domestic economic development. Studies focusing on single countries provide rankings of exporting countries based on that country's market, but lack comparative analysis of product similarities and differences. Furthermore, data limitations in single-country studies may introduce errors when selecting trade partners. The current study addresses these limitations by using comprehensive international trade data and employing a comparative analysis of the Chinese and US markets alongside the global market as a benchmark.

This study uses international trade data from the Harvard University Growth Lab, covering 128 countries from 2001 to 2015. The data were cleaned and processed to ensure reliability. Three sub-datasets were created for the world, China, and the US markets. The methodology includes the following steps: 1. **Data Collection:** Data on exporting countries, importing countries, product codes, and values were extracted from the dataset. 2. **Data Processing:** Sub-datasets were filtered for the world, Chinese, and US markets. Data were screened to ensure consistent results across the three perspectives, including data quality, GDP export information, population size, and annual export value. 3. **Product Proximity Algorithm:** This algorithm, based on Revealed Comparative Advantage (RCA), calculates the proximity between products based on the conditional probability of their co-exportation. A proximity matrix was created and visualized as a heatmap, showing the proximity between different products across the three markets. 4. **Economic Complexity Index (ECI) and Product Complexity Index (PCI):** These indexes were calculated using the Method of Reflections (MR), providing measures of the economic development level of exporting countries and the complexity of exported products, respectively. These indices, based on the normalized country-product matrix, are further used in correlation analysis to understand the relations between the three trade systems and the identification of extreme values. 5. **Correlation Analysis:** Spearman's correlation coefficient was used to analyze the relationships between ECI and PCI values across the three market perspectives. 6. **Extreme Values Mining:** The differences in ECI and PCI values between each market and the world average were calculated to identify extreme values, highlighting significant disparities in economic development and product dependence across markets.

The study's key findings are as follows: 1. **Product Proximity:** The world trade market shows a relatively even distribution of product proximity. China shows increasing proximity among textile-related products (category 8), while the US market shows a declining trend in overall product proximity. This indicates that China offers more opportunities for product diversification in textile industry than the US market does. 2. **Correlation Analysis:** The correlation between China's ECI and PCI and the world's is high, indicating consistency with global trade patterns. The correlation between the US and the world is lower, suggesting greater divergence. The correlation between China and the US is the lowest, signifying significant differences in economic trade patterns. The correlation between the China and the world increases over time, while the US and China correlation decreases, reflecting changes in trading dynamics. 3. **Complexity Trends:** The overall trends in ECI and PCI across the three markets are broadly consistent. However, significant variations are observed among specific countries and products. For instance, China shows lower complexity values for raw materials like coal, whereas the US exhibits higher complexity values for certain products. This underscores distinct economic dependencies and opportunities in each market. 4. **Extreme Values:** Analyzing extreme values of ECI and PCI differences between each market and the world average revealed significant disparities. Countries with high ECI values in China but low values in the world market suggest that these countries have better trading positions in the Chinese market compared to the global market. Conversely, for products with high PCI values in China and low values in the world, the Chinese market is less dependent on them, providing opportunities for specialized exporters.

The findings highlight the importance of considering economic complexity when selecting trading partners. The contrasting trends in product proximity between China and the US suggest different strategic approaches to product diversification. The stronger correlation between China's economic complexity and that of the world market indicates a more conservative and less volatile trading environment compared to the US market. The identification of extreme values emphasizes the need for a nuanced approach, considering the specific characteristics of products and the unique dynamics of each market. The results provide a valuable framework for countries to make data-driven decisions about trade partner selection, accounting for both current conditions and future development possibilities.

This study provides a novel framework for selecting trading partners based on economic complexity, offering a more nuanced approach than traditional methods. The findings highlight the differences between the Chinese and US markets, indicating different opportunities and challenges for product diversification. Future research could explore the impact of specific policy interventions on economic complexity and trade partner selection. Expanding the analysis to incorporate other economic complexity measures and a wider range of trade markets would further enhance the model's scope and applicability.

The study's findings are based on data from 2001 to 2015. Changes in the global trade landscape since then, including the impact of the COVID-19 pandemic and ongoing geopolitical shifts, could affect the relationships observed. Further, the study focuses on two major markets (China and the US) and a general world market benchmark. Including more detailed regional markets could provide a richer picture. The focus on the classical economic complexity index and the product proximity method may also limit the scope of conclusions, and future research could explore alternative metrics and methodologies to improve the model.