The existing blockchain system mostly adopts the equal mining mode. All bookkeepers (entities) record the ledger books on a single master chain. The data storage is random. Moreover, in complex or classified financial scenarios, master chain data is difficult to be correlated or stored regularly, resulting in low efficiency of storage and query; At the same time, in the existing blockchain system, event traceability is mostly only found in the source block, and the implicit association between entities cannot be identified. To solve these problems, this paper proposes a composite blockchain associated event traceability method. This method firstly constructs the blockchain composite chain storage structure model, proposes the concept of private chain and alliance chain, and realizes the adaptive data association storage in complex or classified scenarios. Secondly, on the basis of obtaining the entity block of event source, auxiliary storage space is established to transfer storage relevant data. A query method of associated entity block based on Apriori algorithm is proposed, and the obtained traceability entity block is constructed to construct the source event association diagram, so as to describe the association relationship between event entities. Finally, a risk assessment system based on reinforcement learning is proposed to realize the risk assessment of traceability entity. Experiments show that the composite blockchain associated event traceability method proposed in this paper can reduce 60% of the storage overhead, improve 90% of the query accuracy and 50% of the security.