Multi-layer customer relationship weighted momentum factor - Multi-layer customer relationship momentum effect based on supply chain network
factor.formula
Multi-layered customer relationship weighted momentum factor:
Normalized weight of edge betweenness centrality:
In the formula:
- :
Represents the weighted momentum factor value of the multi-layer customer relationship of target customer $i$. This value is obtained by summing up the momentum returns of all customers of target customer $i$ (including multi-layer relationship customers) and weighting them by the corresponding edge betweenness centrality weight.
- :
Indicates the number of levels of customer relationships in the supply chain network. For example, $l=1$ represents a direct customer, $l=2$ represents a customer's customer, and so on, up to the preset maximum number of levels $L$.
- :
Represents the graph of the $l$th level of customer relationships in the supply chain network. Where $(m, n)$ means that in the $l$th level of relationships, customer $m$ is the supplier of customer $n$.
- :
Represents the edge betweenness centrality of customer $m$ and customer $n$ in the supply chain network. This indicator measures how many paths pass through the edge connecting customers $m$ and $n$ among all the shortest paths. It can be understood as the importance of the connection between customers $m$ and $n$ in the entire supply chain network.
- :
represents the normalized weight of the edge betweenness centrality between customers $m$ and $n$. This weight is obtained by normalizing the edge betweenness centrality between all customer relationships in all levels of the supply chain network. Its essential meaning is to compare the centrality weights of different customer relationships relatively, so as to more accurately measure the influence of each edge in the network. This weight plays a key role in calculating customer momentum, which gives higher weights to more important relationships in the supply chain network.
- :
Represents the return of customer $n$ in the past month, as a measure of momentum signal. The momentum here can be the change in stock price or other related indicators.
factor.explanation
When calculating customer momentum, this factor not only considers the momentum information of direct customers, but also further explores the incremental information contained in multi-layer customer relationships (such as customers of customers). Compared with the traditional method that only considers direct customers, this factor can more comprehensively capture the momentum transfer effect between customers by introducing the edge betweenness centrality of the supply chain graph network as a weight and considering multi-layer customer relationships. Empirical research shows that considering deeper customer relationships (such as customers of customers) helps improve the factor's predictive ability and provide more reliable signals for quantitative investment.