Introduction: A Three-Stage Model of FDI System Evolution and Definition of Core Variables
Over the past two decades, the operational logic of global foreign direct investment (FDI) has been undergoing fundamental changes. Traditional FDI theories primarily explain capital flows based on cost differences, market size, and resource endowments, but the current global investment system has entered a dynamic system stage driven jointly by efficiency, risk, ecosystem, and strategic security.
From the perspective of systems economics, FDI is not a simple migration of capital between countries, but a dynamic feedback system formed by the interplay of corporate behavior, industrial networks, policy environments, and global risks.
This system can be divided into three evolutionary stages:
Phase 1 (2010–2016): Cost Optimization Era
Enterprises primarily allocate manufacturing bases based on labor costs, land costs, and export convenience. After the rise of manufacturing costs in China, Southeast Asia became a typical receiving region, with the logic of FDI flows centered on "seeking low-cost production nodes".
Phase 2 (2016–2020): Supply Chain Diversification Era
Trade frictions, geopolitical risks, and the exposure of supply chain concentration have led enterprises to reduce dependence on a single country. FDI has shifted from "lowest cost selection" to "balance of cost and risk".
Phase 3 (2020–present): Resilience Restructuring Era
The pandemic, energy security, technological competition, and industrial policies have driven global capital into a reconfiguration cycle. Enterprises no longer pursue a single optimal location but instead establish functionally complementary networks across multiple regions.
Therefore, the core variables of the modern FDI dynamic system include:
- C (Cost): Cost advantage
- R (Risk): Supply chain and political risks
- E (Ecosystem): Industrial ecosystem maturity
- M (Market): Market proximity
- I (Institution): Institutional stability
- S (Strategic Resource): Strategic resource control capability
The outcome of FDI flows is not determined by a single variable, but by the dynamic balance among these variables.
I. Reality Anchoring Layer: Industrial Structure Changes Behind Global FDI Flows
According to the long-term observation logic of UNCTAD, global FDI flows are shifting from a past high concentration in traditional manufacturing and resource industries to digital infrastructure, advanced manufacturing, new energy, and supply chain node construction.
In reality, Southeast Asia has become an important region for verifying this change.
2010–2016: Low-Cost Manufacturing Reception PhaseDuring this period, China's manufacturing costs rose, and labor-intensive industries began to shift to Vietnam, Cambodia, and Indonesia.
Major industries include:
- Textile and apparel
- Electronic assembly
- Home appliance manufacturing
- Basic components
Investment logic:
Enterprises are looking for "alternative production bases".
Vietnam became an important destination for electronic manufacturing due to labor costs, export trade agreements, and open manufacturing policies.
2016–2020: Supply Chain Regionalization Phase
With the increase in global trade frictions, enterprises began to build a "China+1" model.
During this stage, FDI targets changed:
From:
"Where is the lowest production cost?"
To:
"Where can supply chain risks be reduced?"
Thus emerged:
- Vietnam's electronic industry cluster strengthened
- Malaysia's semiconductor back-end manufacturing expanded
- Thailand's new energy vehicle supply chain layout
- Indonesia's resource-based industry investment increased
2020 to Present: Industrial Resilience Restructuring Phase
Current FDI competition has entered "functional node competition".
Enterprises are no longer just looking for factory locations, but:
- Chip packaging capabilities
- Battery material supply
- Engineering talent system
- Port logistics network
- Policy support capabilities
For example:
Electronics industry forms:
China R&D and core components
↓
Vietnam assembly manufacturing
↓
Malaysia semiconductor back-end
↓
Global market export
New energy vehicle industry forms:
Indonesia nickel resources
↓
Battery materials
↓
Regional manufacturing bases
↓
Global new energy supply chain
FDI is forming a networked structure, not a simple relocation.
2. System Drive Layer: Variable Compensation Mechanism of FDI Flow
The problem with traditional investment models is that they treat costs, markets, institutions, infrastructure and other factors as independent variables.
But real FDI decisions are not like that.
Enterprises adopt:
A dynamic model where "insufficient advantages are compensated by other system capabilities".
1. Weak Cost Advantage → Institutional Stability Compensation
Even if a country has higher wages, as long as its institutions are stable and property rights are well protected, it may still attract FDI.
For example:
Malaysia's manufacturing costs are not the lowest in Southeast Asia, but its decades-long accumulation in the semiconductor industry has formed a stable ecosystem, making it an important node in the electronics industry.
Enterprises accept higher costs because:
Risk is reduced.
2. Weak Infrastructure → Geographic Cluster Compensation
Some countries have inadequate infrastructure, but if they are close to mature industrial networks, they can still receive investment.
For example:
Vietnam's early infrastructure was not perfect, but it was close to the East Asian supply chain system and grew rapidly through export manufacturing mode.
Enterprises focus on:
"The efficiency of entering global networks", not a single infrastructure score.
3. Weak Market Size → Export Capability CompensationVietnam and Malaysia have limited market sizes, but they attract global enterprises through trade agreements and export manufacturing systems.
Therefore:
Small market ≠ low attractiveness.
The key lies in:
Whether it can connect to global demand.
The FDI system is thus formed:
Cost + Risk + Ecological Dependence + Institutional Capability
a comprehensive optimization model.
The variables are not in an equal relationship but involve substitution and compensation.
III. Transmission Mechanism Layer: How Enterprises Make FDI Decisions
FDI is ultimately driven by enterprise behavior.
The logic of enterprise decision-making can be expressed as:
Investment choice = cost-benefit + risk control + ecological dependence
Apple's Supply Chain Split: Risk Diversification Mechanism
Apple is not simply looking for regions with the lowest manufacturing costs.
Its supply chain adjustment logic includes:
- Reducing dependence on a single region
- Maintaining supply continuity
- Leveraging manufacturing advantages of different countries
Therefore, regions such as China, India, and Vietnam assume different functions.
This is not supply chain migration but supply chain modularization.
Intel's Layout in Malaysia: Ecological Dependence Mechanism
The semiconductor industry is highly dependent on:
- Engineering talent
- Packaging and testing capabilities
- Supplier networks
Therefore, when enterprises choose locations, the weight of the industrial ecosystem exceeds labor costs.
Malaysia's important value is not low cost, but the long-established semiconductor back-end capabilities.
Tesla's Dependence on Indonesian Nickel Resources: Resource Lock-in Mechanism
In the new energy vehicle industry, key competitive factors have expanded from manufacturing capability to resource control.
An important reason Indonesia attracts new energy investment is:
Nickel resource advantages.
Enterprises lock in key material supply through local investment to reduce future resource risks.
IV. National Dynamic Structure Model: Internal Conflicts of Southeast Asian FDI Nodes
Vietnam: Export Manufacturing Upgrade Node
Current Function
A receiving hub for global electronics and manufacturing supply chains.
Internal Conflicts
The low-cost advantage is declining.
Rising wages and increasing land costs are putting pressure on the traditional manufacturing model.
External Pressures
India and Mexico may compete for some manufacturing segments.
Evolutionary Path
Vietnam needs to transition from:
"Assembly base"
upgrade to:
"Engineering manufacturing and supply chain coordination center."
Indonesia: Resource-Driven Industrial Upgrade Node
Current Function
A center for new energy materials and resource-based FDI.
Internal Conflicts
Resource advantages are clear, but the depth of the industrial chain still needs improvement.
External Pressures
Changes in global new energy technology may reduce the value of some resources.
Evolutionary Path
From:
Raw material exporting country
shift to:
New energy industry chain node.
Malaysia: High-Tech Manufacturing Stable Node### Current Functions
Semiconductor packaging, electronics manufacturing hub.
Internal Conflicts
Higher costs compared to some Southeast Asian countries.
External Pressures
India and Vietnam are competing for electronics manufacturing investment.
Evolution Path
Relying on technology ecosystem to maintain competitiveness, rather than cost competition.
V. Global Alternative Networks: Functional Slicing Competition in Southeast Asia, South Asia, and Latin America
Future FDI will not see a single region fully replace another.
The competition model will be:
Functional slicing substitution, not overall national substitution.
Southeast Asia
Advantages:
- Manufacturing experience
- Export networks
- Asian supply chain connectivity
Main roles:
Electronics manufacturing, new energy components, mid-range manufacturing.
South Asia
Advantages:
- Population size
- Software talent
- Domestic demand market
Main roles:
Digital services, consumer markets, partial manufacturing relocation.
Latin America
Advantages:
- Proximity to the U.S. market
- Abundant resources
Main roles:
New energy vehicles, energy, nearshoring manufacturing.
Global FDI is forming:
Asian manufacturing network
+
Indian market network
+
Latin American nearshoring network
a multi-center structure.
VI. Dynamic Feedback Loops and System Stability Analysis
Loop A: Cost Erosion Loop
FDI inflow
↓
Employment increase
↓
Wage rise
↓
Cost advantage decline
↓
Partial low-end manufacturing relocation
↓
New low-cost regions activated
This loop explains why manufacturing bases keep spreading to new regions.
It exhibits:
Divergent trend.
Loop B: Resilience Enhancement Loop
Supply chain dispersion
↓
Increased management complexity
↓
Enterprises reconcentrate on key nodes
↓
Formation of core supply chain ecosystem
↓
Improved system stability
This loop indicates:
Enterprises will not disperse indefinitely; they will form new concentrations.
It exhibits:
Convergent trend.
VII. FDI System Stability Classification
Stable Structures
Southeast Asia Manufacturing Cluster
Reasons:
- Multi-industry entry
- Supplier network formation
- Talent system establishment
Has sustained attractiveness.
Semiconductor Regional Ecosystem
Reasons:
Technological accumulation creates path dependency.
High relocation costs for enterprises.
Unstable Structures
Single Low-Cost Dependency Model
Risk:
Competitiveness declines rapidly after wage increases.
Resource-Dependent FDI Structure
Risk:
Affected by changes in global technology pathways.
Conclusion: Future Directions of the FDI Dynamic System Model
Modern FDI has shifted from "capital seeking cost advantages" to:"Enterprises seek optimal industrial network positions under global risk constraints."
An explainable dynamic system model of FDI must include:
Structural drivers:
Reorganization of global industrial chains, technological competition, supply chain security.
Behavioral mechanisms:
Enterprises invest based on cost, risk, and ecological dependencies.
Reality boundaries:
Institutions, infrastructure, resources, and markets define system constraints.
Dynamic feedback:
Cost erosion drives diffusion, while resilience demands drive re-concentration.
In the coming decade, global FDI competition will not be a simple contest between countries over investment scale, but rather a competition for key functional positions within industrial networks.
A truly stable investment system does not necessarily possess the lowest costs; instead, it possesses the systemic capacity to continuously absorb, upgrade, and connect global value chains.