Ethereum has never fit neatly into the boxes that finance uses to value companies. Its native token is not a share, a bond, or a legal claim on future profits, yet the network generates enormous economic activity and attracts vast capital. A new theoretical study argues that the missing piece is not a better single formula but a map of how the forces behind Ethereum’s value feed back into one another. Researchers at the University of Zagreb’s Faculty of Organization and Informatics in Varaždin have built an integrative causal-loop framework that treats the platform as a complex adaptive socio-economic system rather than as a conventional productive asset or monetary instrument.

The study, published in Frontiers in Blockchain, organizes the sprawling blockchain valuation literature into six principal perspectives: fundamental and on-chain models, network-based models, monetary and tokenomics models, scarcity-oriented models, security and staking-based models, and behavioral and reflexive models. Each stream emphasizes a different mechanism through which the token might acquire value. Fundamental approaches link price to observable protocol activity such as transaction fees, gas consumption, active addresses, and total value locked. Network models trace value to adoption and externalities, drawing on Metcalfe’s Law, which proposes that a network’s value grows with the number of connected users. Tokenomics models focus on transactional demand, velocity, and liquidity. Scarcity models stress supply constraints and the monetary premium that can arise when issuance is low or negative. Security models point to staking, in which validators lock up tokens to help secure consensus and earn rewards. Behavioral models highlight investor attention, sentiment, and reflexivity, the process by which market participants’ beliefs influence the very outcomes they try to predict.

The authors’ central claim is that none of these perspectives, taken alone, explains how productive activity, adoption, monetary demand, token scarcity, network security, and market expectations interact. They therefore translated the mechanisms identified in the literature into system variables and connected them with directional causal links, producing a diagram containing five reinforcing feedback loops, three balancing loops, and additional productive value-creation pathways that do not close into a loop. Reinforcing loops can amplify both increases and decreases. The balancing loops regulate Layer-1 execution-gas utilization, Layer-2 blob-gas utilization, and staking participation through endogenous fee and return adjustments. Crucially, the framework analytically separates three things that are often conflated: the economic utility the ecosystem generates, the mechanisms through which ETH captures value from that activity, and the market price at which the token trades.

The conceptual separation matters because holding ETH confers no legal residual claim on network revenues, unlike equity in a corporation. Discounted cash flow logic, a cornerstone of traditional valuation, assumes that an asset’s value equals the present value of expected future cash flows flowing to shareholders and creditors with legally defined claims. Applying that logic to a decentralized platform is problematic: Ethereum is not a corporation, and token holders have no entitlement to profits comparable to equity ownership. The framework therefore treats DCF-style reasoning as capturing only the productive dimension of value formation, one component within a broader adaptive system that also involves monetary utility, network effects, security participation, scarcity, and behavioral expectations.

One reinforcing loop, labeled R1, traces adoption and utility. More active users can increase the service utility available in the ecosystem, which may attract developer activity, support a wider range of applications, generate additional transaction activity, and draw in further users. A second loop, R2, runs through monetary dynamics: to the extent that greater Layer-1 and Layer-2 activity expands the actual use of ETH for fees, settlement, collateral, and liquidity, broader monetary acceptance may support adoption and subsequent activity. The authors stress that this link is conditional rather than mechanical, because fee abstraction, Layer-2 architecture, and alternative settlement assets can weaken the relationship between ecosystem activity and end-user demand for ETH balances.

A third loop, R3, connects staking to network security. Greater adoption can raise fundamental demand for ETH, part of which may be allocated to staking. A larger amount of staked ETH can, all else equal, increase the economic resources required to compromise consensus. A balancing loop, B3, counteracts unlimited staking growth: because the protocol reward rate per unit staked falls as total stake rises, a higher expected net return attracts additional stake after a delay, while a larger staked base reduces the return available to an additional staker. The authors note that validator compensation should not be equated with company-like cash flow, since the expected net staking return comprises issuance-based consensus rewards, priority fees, and potential MEV, net of penalties and slashing risk, while the EIP-1559 base fee is burned rather than paid to validators.

A fourth loop, R4, captures scarcity dynamics through the interaction of activity-dependent ETH burning and consensus issuance. The burn rate depends on Layer-1 execution-gas utilization and the execution base fee, and on blob-gas utilization and the blob base fee for rollup data, rather than on transaction counts alone. When protocol burning exceeds issuance, net supply can fall; when burning is weak, net supply can rise. The study emphasizes that Ethereum’s supply dynamics are state- and protocol-dependent rather than structurally deflationary, citing evidence that after the Dencun upgrade, lower fee pressure and reduced ETH burning following EIP-4844 produced a net inflationary environment during parts of 2024 and 2025. Holding demand constant, lower supply or lower expected supply growth may place upward pressure on price, but the authors describe this effect as conditional rather than mechanical.

The fifth reinforcing loop, R5, is the speculative one. Rising prices can attract media attention, strengthen expectations of further appreciation, draw in capital, and increase speculative demand, amplifying the initial move; falling prices can reverse the same sequence. Empirical work on cryptocurrency markets documents momentum, investor-attention, and media-coverage effects consistent with individual links in this chain, but the authors caution that such evidence does not by itself establish the complete loop for Ethereum. They frame R5 as a market-price amplification mechanism that may operate without any contemporaneous change in ecosystem service utility, transaction activity, security, or supply conditions.

The framework also includes balancing loops B1 and B2, which represent Ethereum’s two fee markets. When Layer-1 execution-gas utilization exceeds its protocol target, the EIP-1559 mechanism raises the execution base fee in the subsequent block, which can reduce, postpone, or redirect marginal demand. Similarly, when Layer-2 blob-gas utilization exceeds its target, the blob base fee rises, encouraging rollups to batch more efficiently, compress data, or postpone publication. The authors illustrate these dynamics with EIP-4844, activated on 13 March 2024, which introduced blob transactions and a separate blob-gas fee market for rollup data. Reported findings that the upgrade increased rollups’ use of Ethereum while reducing rollup transaction fees, and that cost-minimizing batching strategies can delay transaction inclusion, are consistent with the framework’s distinction between ecosystem transaction activity, execution-gas utilization, and blob-gas utilization. The authors present this as an event-based consistency check, not a validation of the full feedback structure.

The study is explicitly conceptual. The authors describe it as a theory-building structural hypothesis rather than an empirically validated pricing model, and they list limitations including the absence of empirical estimation, interpretive judgment in classifying studies, an Ethereum-specific focus, and the exclusion of macroeconomic conditions, regulation, institutional flows, and market-wide shocks that are not endogenously represented. Latent constructs such as ecosystem service utility and monetary acceptance depend on imperfect proxies, and rapid protocol development may make some links time- and regime-dependent. The framework also does not attempt to represent social, ethical, political, or normative values shaping the ecosystem. Its intended contribution is foundational: by making the direction, polarity, and closure of hypothesized causal relationships explicit, the causal-loop diagram provides a basis for subsequent stock-and-flow simulation, empirical calibration, and comparative application to other blockchain platforms, turning a fragmented valuation debate into a testable set of propositions about how a decentralized economy actually holds itself together.

Subject of Research: A mechanism-based causal-loop framework for Ethereum value formation

Article Title: Ethereum value formation as a feedback system: an integrative causal-loop framework

Article References: Kovšca, V., Lacković Vincek, Z., & Keglević Kozjak, S. (2026). Ethereum value formation as a feedback system: an integrative causal-loop framework. Frontiers in Blockchain, 9, Article 2002131. https://doi.org/10.3389/fbloc.2026.2002131

Image Credits: AI Generated

DOI: 10.3389/fbloc.2026.2002131

Keywords: Ethereum, blockchain valuation, tokenomics, system dynamics, causal loop diagrams, complex adaptive systems, staking, EIP-1559, EIP-4844, network effects, scarcity, reflexivity

News Source: Denise Maddox. (November 8, 2026). Ethereum’s value may emerge from feedback loops, not a single formula. Scienmag.