Recent market dynamics have highlighted the vulnerability of supply chains to various disruptions. Accelerated artificial intelligence investment by hyperscalers, reduced housing turnover, and persistent tariff uncertainty illustrate how shocks to downstream demand and operations can rapidly propagate to upstream suppliers. Consequently, identifying signals that indicate how customer performance transmits through the supply chain has become increasingly valuable.
Customer momentum addresses this challenge by linking a supplier’s outlook to the recent return momentum of its key customers. Although the basic approach with an equal-weight average of customer momentum is transparent and straightforward, it may introduce sector biases and fail to account for the differing influence of individual customers within the network.
To better reflect economic reality and manage unintended exposures, two enhancements are evaluated. First, a centrality-based weighting scheme is implemented, assigning greater importance to customers occupying more influential positions within the supply chain network. Second, a sector-neutral portfolio is constructed to minimize broad industry exposures and sharpen the intended signal.
Backtests indicate the sector-neutral, centrality-weighted customer momentum approach outperforms the simple average, boosting total return and the Sharpe ratio. These findings suggest that incorporating network structure and disciplined portfolio construction significantly enhances the effectiveness of customer momentum in the current market environment.
We use FactSet Supply Chain Relationships as our primary database for analysis. The database sources disclosures from more than 53,000 companies worldwide and provides 13 types of entity relationships to understand each company’s ecosystem. All relationships are manually verified and compiled by FactSet analysts. The database also includes useful metadata, such as relationship keywords and subsidiary names involved in the relationship, as well as derived metrics, such as relationship relevance ranks and centrality measures.
The customer momentum factor is determined by averaging the customers’ momentum factor, typically based on stock price returns. The simplest method for computing this factor is to take the average of customers' one-month stock returns.
We conducted a backtest on the iShares Russell 1000 constituents over the past 20 years (from January 2006 to December 2025). We excluded financial companies and companies with fewer than 3 customers that have the momentum factor. We created equal-weighted quintile portfolios based on the average of customers' one-month stock returns and reconstructed them monthly. We monitored their performance, ignoring transaction costs.
Figure 1 displays the cumulative returns of these quintile portfolios. Q5 represents companies with the highest Customer Momentum factor exposure, while Q1 consists of companies with the lowest. As expected, Q5 outperformed Q1 by 86.0%.
Figure 1: Quintile portfolio cumulative returns
On the other hand, Q4 exhibits slightly better total return than one for Q5. Also, the spread return is quite volatile, resulting in a annualized Sharpe ratio of 0.38 (assuming a 0% risk-free rate return) for the long-short portfolio.
Figure 2: Cumulative spread return between Q5 and Q1 portfolios
The result indicates that customer momentum in its basic form remains effective overall, despite multiple spikes observed in the historical spread return.
Next we look into how the strategy can be enhanced, focusing on two aspects.
The customer momentum factor above was calculated as a simple average of customers’ stock return momentum, implicitly assuming that all customers have an equal impact on a supplier’s business. In reality, this is often not the case. Ideally, supplier revenue exposure to each customer could serve as a more accurate weighting scheme, but such data is rarely reported and is therefore often unavailable.
To address this limitation, we use a centrality measure as a proxy weighting scheme for the customer momentum factor. Centrality, derived from network theory, quantifies the importance of nodes or links within the network. Various centrality metrics exist, but for this study, we focus on Katz centrality.
Katz centrality accounts for both direct and indirect connections, including links to higher-tier customers/suppliers, by discounting longer paths so that indirect relationships have less influence. In supply chain networks, a high Katz centrality score indicates a company has more downstream customers (in a supplier-to-customer graph) or more downstream suppliers (in a customer-to-supplier graph). In this case, we assume customers with broader downstream customers have a greater impact on their suppliers.
FactSet Supply Chain Relationships data offers a variety of pre-calculated centrality measures derived from a point-in-time supply chain relationship network. We will use this information to define the weighting schema for the customer momentum strategy.
The customer momentum strategy aims to capture alpha signals from customer performance, using these signals as leading indicators. However, it often reflects industry momentum as well. For example, cyclical (defensive) companies tend to have a significant portion of their performance in Q5 during bull markets and in Q1 during bear markets.
Similarly, thematic sector trends such as the AI boom can lead to concentrated exposure in a particular sector. Figure 3 illustrates the sector allocation for Q5 and Q1 as of December 2025. In the top quintile group (Q5), the major sector is Healthcare, while the bottom quintile group (Q1) shows a large exposure to Industrials and Technology.
Figure 3: Sector allocation by RBICS L1 sector for Q5 and Q1 portfolios as of December 2025
To prevent portfolio bias towards specific sectors and the potential volatility that could be introduced, we will construct a sector-neutral portfolio. The Q5 group includes the top 20% of stocks in each sector by factor exposure, while the Q1 group includes the bottom 20% of stocks in each sector by the same criteria.
We are now conducting a backtest with the enhanced customer momentum factor for the same universe and period. Figure 4 displays the cumulative returns for sector-neutral quintile portfolios constructed with centrality-weighted customer momentum. The portfolio results are clearly stratified: Q5 delivers the highest returns, while Q1 underperforms relative to all other groups.
Figure 4: Cumulative returns of quintile portfolios based on enhanced strategy
Figure 5 shows the cumulative spread returns of quintile portfolios under the initial and enhanced strategies. The cumulative spread return improved by 35.6%. Moreover, with lower volatility in the spread returns, the Sharpe ratio improved by 0.36 (from 0.35 to 0.71). The results show that the factor return was notably improved and stabilized.
Figure 5: Cumulative spread returns comparison between a simple average vs sector-neutral and centrality weighted customer momentum
Our analysis indicates the customer momentum factor, derived from supply chain relationships, continues to generate meaningful alpha signals. Moreover, incorporating sector-neutral and centrality-based adjustments substantially improves both the magnitude and stability of outperformance.
While this article explored a select set of enhancements, there remains a broad range of future research possibilities. For example, alternative weighting metrics can improve the accuracy of measuring customers’ impact on suppliers.
In the realm of centrality measures, numerous other types exist, such as edge betweenness centrality, which could provide a more accurate estimate of the materiality of relationships.
Additionally, while we incorporated customers’ stock price returns as momentum factors in this study, we could also consider other factors, like suppliers’ sustainability sentiment, to enhance the accuracy of predicting future returns for the focal companies. There are many more ways to enhance the predictive power of the supplier/customer factors for the focal companies’ future performance.
This blog post is for informational purposes only. The information contained in this blog post is not legal, tax, or investment advice. FactSet does not endorse or recommend any investments and assumes no liability for any consequence relating directly or indirectly to any action or inaction taken based on the information contained in this article.