The Question
Natural disasters, pandemics, regulatory crackdowns, and policy shifts hit places, not just sectors. When Shanghai locked down in spring 2022, the impact wasn't limited to Shanghai-based companies—it rippled through supply chains, disrupting suppliers nationwide and customers globally.
Traditional risk models classify stocks by industry, size, and style factors—but not by geographic exposure. Two companies in the same sector can have radically different regional concentration. One sources 80% of inputs from a single province; the other has a diversified national supply base. When regional shocks hit, the concentrated player suffers disproportionately.
Can you systematically map geographic risk exposure—not just where a company's headquarters is, but where its suppliers and customers are concentrated—to predict which stocks will be hit hardest when regional disruptions occur?
The Approach
We build a multi-layer geographic risk map combining company domicile data with supply chain networks:
Layer 1: Direct Exposure
Identify all listed companies headquartered in a specific region (e.g., Shanghai). From Financial Notes and public filings, extract headquarters location and major production facility locations. This gives the first-order exposure: companies physically located in the affected region.
Layer 2: Upstream Exposure (Suppliers)
Using Customer & Supplier data, identify suppliers to Shanghai-based companies. If a component manufacturer in Jiangsu derives 40% of revenue from Shanghai customers, it has high indirect exposure. We weight supplier exposure by the % of their revenue tied to the affected region.
Layer 3: Downstream Exposure (Customers)
Similarly, identify customers of Shanghai-based companies. If an automaker in Guangdong sources critical parts from Shanghai suppliers, it faces disruption risk even though it's geographically distant. We quantify this through product dependency—using SAM to identify which products are bottlenecks (few alternative suppliers) vs. commoditized (many substitutes).
Layer 4: Extended Supply Chain
Map second-degree connections: suppliers' suppliers and customers' customers. A Sichuan raw material provider selling to a Jiangsu manufacturer who sells to Shanghai assemblers has third-order exposure. We construct the full supply chain graph and run centrality metrics to identify companies highly embedded in Shanghai-centric networks, even if they're not adjacent.
The model produces a geographic risk score for every listed company relative to any specified region, quantifying both direct and supply-chain-mediated exposure.
The Finding
During Shanghai's 2022 lockdown, companies with high geographic risk scores (combining headquarters location + supply chain exposure) underperformed the market by 8-12% on average over the 60-day disruption window. Importantly, the supply chain exposure component was more predictive than simple headquarters location.
Case study: Semiconductor supply chains
Shanghai hosts critical semiconductor equipment and material suppliers. Chip manufacturers nationwide (Shenzhen, Wuxi, Hefei) depend on these suppliers. During the lockdown, even companies with no physical presence in Shanghai saw production delays and margin compression due to bottlenecked inputs. The model correctly flagged these exposures weeks before earnings warnings.
Case study: Automotive supply chains
Shanghai is a major auto manufacturing hub (Tesla Gigafactory, SAIC, local suppliers). But the bigger exposure was to tier-1 and tier-2 suppliers scattered across Jiangsu and Zhejiang provinces who depended on Shanghai-based customers or sub-suppliers. Mapping the extended supply chain revealed hidden vulnerabilities that simple "Shanghai headquarters" screens missed.
False positives and resilience:
Not all Shanghai-based companies underperformed—some had diversified supplier bases or flexible production networks that absorbed shocks. Companies with high SAM product diversification (multiple product lines sourced from different regions) showed resilience. Geographic risk mapping needs to be combined with supply chain flexibility metrics, not just concentration.
Pre-event positioning:
The real alpha comes from identifying concentrated exposures before disruptions hit, allowing hedging or de-risking in advance. The framework applies to any regional risk: natural disasters (earthquake zones), policy changes (carbon regulations hitting specific provinces), or geopolitical tensions (Taiwan-adjacent exposure).
Try It Yourself
Building a geographic risk map requires integrating company location data with supply chain relationship networks—then running graph algorithms to quantify direct and indirect exposures.
Implementation steps:
-
Curate location data: Company headquarters, major facilities, supplier/customer locations. This data is often scattered across disclosures, registrations, and third-party databases—consolidating it is the hard part.
-
Build supply chain graphs: Use Customer & Supplier datasets to construct directed graphs (supplier → customer). Enrich with SAM product data to identify bottleneck relationships (critical inputs with few alternatives).
-
Run exposure calculations: For any specified region, calculate multi-hop exposure scores using graph traversal algorithms (depth-limited search, weighted by revenue dependencies).
-
Monitor triggering events: Set alerts for regional disruptions—weather events, policy announcements, outbreak reports. When triggers fire, immediately score your portfolio's exposure and stress-test positions.
-
Dynamic hedging: Use exposure scores to construct region-neutral portfolios or hedge via index futures when regional risks rise.
This framework is particularly valuable for event-driven strategies and risk management overlays. It doesn't generate continuous alpha (you're not rebalancing weekly based on geography), but it protects against tail risks that traditional risk models miss entirely.
Ready to map geographic exposure across your portfolio's supply chains? Book a call to discuss data integration, graph infrastructure, and stress-testing frameworks.
Source: 兴业证券《办公地为上海的企业产业链》 (2022-04-16).