Abstract

International business scholarship has long documented that the benefits of globalization accrue unevenly: large multinational enterprises (MNEs) capture most cross-border value, while small and medium-sized enterprises (SMEs), which constitute the overwhelming majority of firms worldwide, remain largely excluded from global value chains. This exclusion is primarily a matter of constraint rather than ambition or product quality. The prevailing internationalization models-stage theories, the Uppsala model, and born-global paradigms-assume access to market intelligence, foreign-market capital, multilingual talent, and regulatory-compliance capacity that resource-constrained SMEs often cannot supply. The Value-Yield Optimization (VYO) Framework (Yevsiukova, 2026) addressed this problem domestically by ranking AI-driven optimization opportunities according to the return achievable within a firm's real capital, labor, and technical limits. This paper introduces the Global Value-Yield Optimization (G-VYO) Framework, which extends VYO to the cross-border domain. G-VYO differs from VYO in three respects: it adds three cross-border factors (market distance, currency and macroeconomic volatility, and regulatory friction) to the constraint-weighting function; it adds a sixth stage, a systemic resilience audit of the international portfolio; and it introduces a three-axis readiness-headroom-distance decision cube for entry-market planning. The framework specifies five foundational principles, a six-stage operating sequence, and a Global Constraint-Weighted Return on Investment (GCW-ROI) function with explicit measurement rules and a worked numerical example. Four country-entry archetypes are illustrated with five published empirical studies of SME internationalization from Eastern Europe, Southeast Asia, Latin America, and Sub-Saharan Africa. The paper contributes a practitioner-oriented, platform-agnostic methodology for more inclusive participation of resource-constrained SMEs in the global economy and identifies four directions for empirical calibration and longitudinal validation.

This work is licensed under a Creative Commons Attribution 4.0 License.