The Geoeconomics of Venture Capital An Economic Complexity Approach to Emerging Technological Sovereignty

This paper applies economic complexity methods to venture capital data to quantify emerging technological sovereignty, introducing indices that rank countries and technologies based on specialization patterns and identifying optimal single-technology pathways for nations to enhance their geoeconomic positioning.

The Gist

Imagine the world's economy not as a giant marketplace, but as a massive, high-stakes card game. In this game, countries are the players, and their "cards" are the technologies they are developing—things like Artificial Intelligence, Quantum Computing, or Cybersecurity.

For a long time, people thought the country with the most cards or the flashiest cards (like AI) would win. But this paper argues that's a trap. It's not about how many cards you have; it's about which specific cards are in your hand and who else is holding them.

Here is the breakdown of the paper's ideas using simple analogies:

1. The Game: "Technological Sovereignty"

Think of "Sovereignty" as a country's ability to play the game without needing to borrow cards from other players. If you need to rely on someone else for your most important cards, you aren't truly independent.

The authors look at Venture Capital (VC) as the "dealer" distributing these cards. VC money is what helps startups turn wild scientific ideas into real products. By tracking where this money goes, the authors can see which countries are building their own decks of powerful, unique cards.

2. The New Scorecard: "Geoeconomic Complexity"

The paper introduces two new ways to score the game, using a method called Economic Complexity. Think of this like a "Rarity and Difficulty" rating in a video game.

• The "Common Card" Problem: Imagine "Energy Storage" (like batteries) is a card that almost every country is trying to collect. Because everyone has it, it's not very special. It doesn't give you a huge advantage over your neighbors.
• The "Rare Card" Advantage: Now imagine "Cloud Computing" or "Cybersecurity." The paper found that only a tiny group of elite countries (like the US, Israel, and China) are really good at these. Because so few people have these cards, holding them gives you a massive strategic advantage.

The authors created two scores:

• The GCI (Country Score): This ranks countries based on the quality of their hand. The US and Israel are at the top because they hold a mix of many different cards, but crucially, they hold the rare ones that other countries don't have.
• The ETGCI (Technology Score): This ranks the technologies themselves. It asks: "Is this technology held by the elite players, or is it common among everyone?"
  • Top Tier (High Score): Cloud Computing, Cybersecurity, and Medtech. These are the "Trumps" of the deck.
  • Lower Tier (Low Score): AI and Quantum Computing. Surprisingly, these rank lower in this specific game. Why? Because everyone is trying to get them. If everyone has an AI card, it doesn't make you stand out as much as having a unique Cloud Computing card.

3. The "High-Diversity, Low-Ubiquity" Strategy

The paper highlights a specific winning strategy used by the top players (US and Israel).

• Diversity: They have cards in many different categories (they aren't just betting on one thing).
• Low Ubiquity: The specific cards they hold are rare. Most other countries aren't holding them.

It's like a poker player who has a hand full of different suits, but every single card is a "Royal Flush" that no one else at the table has. Meanwhile, a country like Sweden might have a hand that is very common (everyone has those cards), so it doesn't give them a winning edge.

4. The "Simplest Next Move" (SSSET)

The authors also ran a simulation to answer a practical question: "If I'm a country currently in the middle of the pack, what is the easiest new card I can pick up to jump up the rankings?"

They call this the SSSET (Simplest Single Sovereignty Enhancing Technology).

• The Logic: You can't just magically invent a new card. You have to build on what you already know. If you are already good at "Robotics," it's easier to learn "Autonomous Systems" than to suddenly become an expert in "Nuclear Fusion."
• The Result: For countries like the UK or Australia, the simulation suggests that adding Cybersecurity Tools or Medtech to their portfolio would be the most effective way to improve their standing, because it fits well with what they already do and those cards are currently rare and valuable.

5. The Big Takeaway

The main lesson of this paper is a reality check for policymakers:

Don't just chase the "hottest" technology (like AI) because everyone else is doing it. If everyone is doing it, it becomes a commodity, and it stops being a source of power.

True power comes from finding the rare, difficult-to-copy technologies that only a few powerful nations can master. The US and Israel are winning because they aren't just playing the game; they are playing a different, more exclusive version of it.

In short: It's not about having the biggest pile of cards; it's about holding the cards that no one else can get their hands on.

Technical Summary

1. Problem Statement

The paper addresses the growing strategic competition among nations regarding technological sovereignty. While countries increasingly view control over emerging technologies (e.g., AI, quantum computing, fusion) as a determinant of geoeconomic power, existing assessments are predominantly qualitative and anecdotal.

Current frameworks fail to distinguish between:

• Intrinsic importance: A technology's inherent strategic value.
• Relative geoeconomic power: The advantage gained when a technology is controlled by a few powerful nations versus being ubiquitous (widely available) among many.

The authors argue that true "trump cards" in the sovereignty game are not just technologies that exist, but those where a country's venture portfolio is specialized in domains that are complex, rare, and concentrated among high-capability nations. There is a need for a quantitative framework to map national venture capital (VC) portfolios to emerging technological domains to assess this relative positioning.

2. Methodology

The authors adapt the Economic Complexity framework (originally developed by Hidalgo and Hausmann for trade) to the domain of Venture Capital.

A. Data Sources and Taxonomy

• Data: Crunchbase firm- and deal-level data.
• Sample: The 17 countries with the highest aggregate VC funding.
• Taxonomy: 18 emerging technological domains (e.g., Cloud Computing, Cybersecurity, Medtech, Quantum Computing) identified via literature review.
• Classification: A probabilistic multi-label Large Language Model (LLM, ChatGPT-4o-mini) classifies startups into these domains based on their descriptions. The model was validated against a manually annotated set (90.9% precision, 78.8% recall).

B. The Specialization Matrix

The core object is a binary Country–Technology Specialization Matrix ($M$).

• Revealed Venture Advantage (RVA): Analogous to Revealed Comparative Advantage (RCA) in trade. A country $i$ is considered specialized in technology $j$ if its share of global VC investment in $j$ exceeds its share of global VC investment overall.
• Robustness: The authors employ sensitivity checks, including rounding investment amounts to the nearest $100M and using a two-year rolling window for specialization determination.

C. Metrics and Indices

From the matrix $M$, the authors derive two eigenvector-based indices:

1. Geoeconomic Complexity Index (GCI): Ranks countries. A country has high GCI if it specializes in technologies that are themselves complex (i.e., specialized in by other high-GCI countries) and rare (low ubiquity).
2. Emerging Technology Geoeconomic Complexity Index (ETGCI): Ranks technologies. A technology has high ETGCI if its specialization is concentrated among high-GCI countries.

• Ubiquity: The number of countries specialized in a given technology.
• Diversity: The number of technologies a country is specialized in.

D. Strategic Simulations (SSSET)

The authors simulate "Simplest Single Sovereignty Enhancing Technology" (SSSET) for each country.

• Mechanism: They toggle a country's specialization status from 0 to 1 for a non-specialized domain.
• Feasibility Proxy: Technological Relatedness, calculated based on how frequently a new domain co-occurs with a country's existing portfolio across other nations.
• Goal: Identify the single new specialization that offers the highest expected rank improvement in GCI while maintaining high feasibility (relatedness).

3. Key Results

A. Country Rankings (GCI)

• Leaders: The United States and Israel consistently occupy the top positions. They exhibit a unique "high-diversity / low-ubiquity" profile, meaning they invest in many domains, but those domains are rarely specialized in by other nations.
• Followers: China, France, Japan, and Germany follow.
• Laggards: Sweden and Australia rank lower, characterized by portfolios concentrated in ubiquitous, lower-complexity domains (e.g., Mobility, Raw Materials).
• Stability: Rankings have remained stable from 2021 to 2024.

B. Technology Rankings (ETGCI)

• Highest Complexity: Cloud Computing, Cybersecurity Tools, and Medtech rank highest. These domains are geoeconomically powerful because specialization is concentrated in a small set of leading countries (e.g., US/Israel for Cloud; US/Israel/China for Cyber).
• Lower Complexity (Paradox): Artificial Intelligence, Quantum Computing, and Energy Storage rank lower in ETGCI.
  • Interpretation: This does not mean they are unimportant. Rather, they are ubiquitous; many mid- and low-GCI countries specialize in them, diluting their relative geoeconomic leverage. It is "easier" for a broader set of nations to compete in these fields compared to Cloud or Cyber.

C. Strategic Simulations (SSSET)

• Top Tier: No single new specialization can improve the relative ranking of the US, Israel, China, or Germany (they are already at the ceiling of the distribution).
• Mid-Tier Opportunities: Countries like the UK, Canada, Australia, and Sweden have clear SSSETs (e.g., Cybersecurity Tools, Medtech) that could yield significant rank improvements (+4 to +7 positions) due to high relatedness to their existing portfolios.

4. Key Contributions

1. Quantitative Framework for Sovereignty: Moves the discourse on technological sovereignty from qualitative narrative to a rigorous, data-driven economic complexity model.
2. Differentiation of "Power" vs. "Importance": Demonstrates that a technology's geoeconomic power is a function of who controls it, not just what it is. Ubiquity reduces relative power.
3. VC as a Leading Indicator: Establishes venture capital allocation as a proxy for the formation of future industrial capabilities and strategic autonomy, distinct from public R&D or export data.
4. Actionable Policy Tool: The SSSET simulation provides a concrete metric for policymakers to identify feasible "next steps" for upgrading their national technological portfolios.

5. Significance and Implications

• Strategic Insight: The paper suggests that for nations to gain geoeconomic leverage, they should not merely chase "hot" ubiquitous technologies (like general AI) but should strive to specialize in domains where the barrier to entry is high and the set of competitors is small (e.g., specialized Cloud infrastructure or Cyber-defense).
• Policy Design: It highlights that for the EU, simple aggregation of member states' VC flows may mask specialization. A bloc strategy requires complementarity (members specializing in different high-ETGCI domains) rather than just summing resources.
• Limitations: The authors acknowledge that the model focuses on VC flows and headquarters location, potentially overlooking foreign ownership of domestic firms or non-VC state funding. It is a structural analysis, not a causal one.

In summary, the paper provides a novel lens to view the "technological sovereignty game," arguing that true power lies in the composition of a nation's investment portfolio—specifically, its ability to dominate rare, complex domains that are inaccessible to the broader global community.