Why do non-fungible token (NFT) prices diverge from fundamental value? A conditional asset pricing framework for digital art and virtual real estate

Non-fungible tokens (NFTs) are heterogeneous digital assets whose prices may reflect both asset-specific ownership benefits and changing compensation for market, liquidity, and institutional risks. This study proposes a conceptual and theory-building asset-pricing framework that integrates two channels. The fundamental-value channel represents the expected utility-equivalent benefits generated by ownership, including aesthetic experience, social signalling, community participation, functional access, and virtual location. The discount-rate channel represents the required return associated with systematic cryptocurrency risk, illiquidity, institutional and governance uncertainty, platform dependence, legal enforceability, and ownership concentration. The framework is developed through integrative conceptual literature synthesis and draws on no-arbitrage asset-pricing logic, the Campbell–Shiller present-value approach, Gordon-type valuation, liquidity-adjusted asset pricing, and evidence from recent NFT hedonic, behavioural, network, and forecasting studies. A conditional three-state structure is proposed to distinguish stable fundamental-oriented periods, risk-dominated stress periods, and high-attention speculative periods. Importantly, the regime structure is presented as a testable theoretical extension rather than an empirically established classification. The framework also addresses cross-sectional dependence within NFT collections. Seven propositions and an illustrative numerical example translate the framework into an empirical research agenda. The study provides a coherent basis for future testing of NFT price formation while avoiding the stronger claim that any single determinant universally dominates NFT pricing.