Reliably extracting valuable information from economic data is notoriously difficult. They are the result of highly subjective actions driven by contingency and strong non-linearity, driven by high-dimensional influences. These characteristics have long posed a challenge for conventional econometric, regression-based models. To tackle this challenge, we propose using knowledge graph embeddings to analyze economic trade data, with a focus on predicting international trade relationships. We introduce KonecoKG, a knowledge graph representation of economic trade data with multidimensional relationships, built using SDM-RDFizer. We then transform these relationships into knowledge graph embeddings using AmpliGraph. Our results show that the method performs much more accurately when compared to baseline regression models.