Distributed ledgers are the embodiment of the vision of a decentralised notion of money. The key idea is to harness a network of self-interested ‘validators’ to perform the record keeping role of trusted intermediaries that underpin our current monetary system. This column examines the economic trade-offs in designing such systems, focusing on the interplay between decentralisation, finality, and scalability. A well-designed decentralised system aligns the interests of market participants with ledger maintainers. However, replicating validation across numerous validators can be cost-effective only in low-trust environments.
Money is a social convention, accepted in payments in the expectation that others will do so in the future. In monetary exchange, holding money is evidence of past transactions of goods sold or services rendered. In effect, money can be seen as a record-keeping device. Kocherlakota (1998) and Kocherlakota and Wallace (1998) demonstrated that money can function as a record-keeping mechanism nearly as effectively as a complete ledger of all past transactions, encapsulated in the motto “money is memory”.
Maintaining a complete history of past transactions is a ‘brute force’ way of keeping such a record of each individual’s debt to every other member. Lugging around the master paper ledger was an elegant theoretical benchmark, but hardly a practical proposition of recording economic activities in the real world. Instead, trusted intermediaries like modern central banks have maintained centralised transaction ledgers to ensure accurate and reliable balances. Commercial banks perform the same role with respect to retail users.
However, with the advent of digital ledgers and the emergence of blockchain and other distributed ledger technologies (DLTs),the fanciful notion of a master ledger of all transactions stretching back to the beginning of time looms as a potentially realistic proposition. The traditional monetary system based on trusted intermediaries has a new challenger that makes the ‘brute force’ way of having a master ledger a tantalising prospect. Key to the model are decentralised consensus mechanisms that update and maintain transaction records through a network of self-interested record keepers (‘validators’) for whom being part of the decentralised consensus is a matter of self-interested response to incentives.
While the technology has brought forward the prospect of having digital master ledgers as being a potential challenger to the current monetary system, there are fundamental questions about who should have the authority to update these ledgers and how. The decentralisation agenda, exemplified by blockchain technologies underpinning cryptocurrencies like Bitcoin, advocates transferring this authority from commercial and central banks to dispersed and self-interested network members, where truth is determined by the collective consensus. How realistic is this proposition?
The value of decentralisation
The fundamental question, which we examine in Auer et al. (2025), is whether and under what conditions it is optimal (in a sense to be made precise below) to update monetary ledgers in a decentralised manner. It turns out from our analysis that the economic gains from decentralisation are far from self-evident. 2
Decentralisation introduces a key economic friction – the inefficiency cost of maintaining consensus and ensuring that validators act honestly. The specific features of the decentralisation mechanism that exclude coordination on all but the true state of the world hence sets a high bar on the strength of the consensus necessary to update the ledger.This high bar is conducive to the ledger being more secure, and tamper-proof. However, the high bar for consensus means that agreement is often not possible, and takes more time and effort when it is. The larger the ledger, the harder it becomes to update it in a timely manner. Put another way, the safeguards necessary for the integrity of the ledger may make decentralised consensus unattainable for many cases, and slow and cumbersome for those cases where decentralised consensus is feasible.
The more nodes there are in a network, the higher the bar and the longer it takes to reach broad agreement. In short, the ledger is hard to scale and fails to provide the necessary foundation for the monetary system. This conundrum introduces trade-offs sometimes known as the ledger’s ‘scalability trilemma’ (Buterin 2021). The trilemma is posed in terms of the impossibility of attaining a ledger is simultaneously decentralised, secure and scalable (see Figure 1).
Figure 1 Buterin’s trilemma
What is the optimal level of decentralisation?
In view of the trilemma, the task is to take decentralisation at face value, and identify the ‘sweet spot’ between decentralisation, security and scalability. What are the conceptual limits? In our paper, we explore this question in a credit economy, where some agents produce early for others, expecting those beneficiaries to reciprocate later (see Figure 2).If these ‘late’ producers cannot commit to reciprocating, trade cannot be sustained in equilibrium.
Figure 2 Timeline
Research by Kocherlakota (1998), Kocherlakota and Wallace (1998) and Rocheteau and Nosal (2017) has shown that trustless exchange can emerge as an equilibrium when agents have access to a record-keeping system that tracks the beneficiaries’ default history.
However, the previous literature had sidestepped the question of who should be responsible for ensuring the integrity of this ledger. Our contribution is to examine the economic principles underlying the validation protocol. We hence revisit Juvenal’s analysis of who – or what mechanism – should guard the guardians of the ledger in such a setting.
Our approach abstracts from the details of the computing or cryptographic implementation and focuses, instead, on the economic incentives underlying the consensus process. Concretely, we focus on the incentives that economic agents face in fulfilling their tasks as validators whose role is to agree on the unique set of executed trades that will be written into the ledger. The elements of our monetary economy correspond to the three vertices of the triangle in the scalability trilemma.
We model the degree of decentralisation as the size of the committee that is tasked with updating the ledger. The system is fully centralised when just one node has the authority to update the ledger, and the system is fully decentralised when all nodes in the system take part in updating the ledger. In between these two extremes, the system is ‘permissioned’ when only a pre-selected set of nodes can update the ledger.
We model the strength of the security around the integrity of the ledger as the supermajority threshold for agreement that is needed to validate a new block in the chain. The most secure rule is to insist on unanimity, so that everyone agrees. However, insisting on unanimity could be a recipe for gridlock and delay. Deviations from unanimity increases efficiency but at the cost of security. The approach we take is to model the decision as a vote on whether a block in the blockchain reflects the truth. ‘Truth’ is a matter of consensus. Truth is whatever is deemed to be true by the consensus, whether it be objectively true or not .Achieving a clean, reconciled ledger that everyone accepts is a public good that benefits all. The voting game is therefore a public good contribution game (e.g. Morris and Shin 1998, 2003), and the supermajority threshold for the vote is a parameter of our model which can be adjusted to achieve the social optimum. Greater security demands a higher supermajority threshold for the successful provision of the public good – a clean, reconciled ledger that everyone accepts.
Finally, we formalise the scalability of the mechanism used to update the ledger as the transaction size that can be implemented with a given ledger design. The mechanism is considered scalable whenever trades are at their Pareto efficient levels.
The central dimensions of the trade-offs of our monetary economy reflect the essence of Buterin’s trilemma. In the public good contribution game, each validator needs to perform his or her assigned task of verifying the transactions. This task entails a small cost for each validator, and this cost is assumed to vary slightly across validators. These validation costs inject strategic uncertainty and set up a global game for the public good contribution game. Unless the rewards that accrue to the validators are sufficiently high, they may not follow the validation protocol in equilibrium. This is especially so when the supermajority threshold is set very high. Thus, a direct consequence of aiming for stronger security is that the validators need to be given a bigger piece of the social surplus in the form of rents. In turn, the high rents that the validators extract reduce the overall size of the pie in terms of the economic gains that arise from monetary exchange. Rents undermine scalability, but rents are necessary for security. We thus end up with a trade-off along one dimension of Buterin’s trilemma.
Rents to validators can be reduced (and scalability enhanced) by moving away from full decentralisation. If the single validator node can be trusted with managing the ledger, then security and scalability can be achieved. However, if the single validator cannot be trusted, better governance calls for the checks and balances that only a larger committee of validators can deliver. That nodes need to coordinate on a single ledger acts as a discipline device for each one of them. However, this brings us back full circle, as the robustness of the voting equilibrium requires nodes be incentivised with sufficient rents.
Designing an optimal ledger system
Our analysis presents a general framework for assessing the desirability of ledger systems from the perspective of the functioning of the monetary system. As a theoretical matter, these ledgers could be anonymous or identified, or indeed permissioned or permissionless. The optimal validation protocols balance the number of validators, their compensation and the supermajority threshold needed to update the ledger. For the purpose of updating the ledger, truth is whatever is deemed so by the requisite supermajority of the validators.
We hence model Buterin’s scalability trilemma formally within our monetary economy. Scalability is only achieved by abandoning decentralisation, while decentralisation and security comes at the cost of scalability. Achieving all three goals is ruled out in equilibrium. The optimal solution that balances decentralisation, scalability, and security depends on the specific frictions that operate in the economy and their severity. We lay out the considerations that determine the limits of decentralisation encapsulated by the triangle. Our main result is the derivation of a mapping from the trade-offs involved to the socially optimal point on the Buterin triangle.
Our model of ledger governance highlights the importance of the strength of intertemporal incentives. When long-term rewards matter more, a single entity managing the ledger turns out to be optimal, as it reduces inefficiency costs and ensures consistency. This single entity could be interpreted as a commercial bank or the central bank. However, when the future matters less and short-sighted incentives dominate, decentralisation can be a way to mitigate the costs associated with any single entity gaining excessive control.
We also find that a stakeholder economy – where those who participate in transactions also validate the ledger – can lead to better governance outcomes. This form of governance aligns incentives and ensures that validators have a vested interest in maintaining ledger integrity. Additionally, different ledger architectures can achieve socially optimal outcomes by adjusting the level of anonymity, permissions and competition between ledgers.
In sum, our research underscores that decentralisation is no end in and of itself, highlighting the trade-offs between centralised and decentralised ledger governance. Even if we leave aside considerations such as crypto’s role in fraud and money laundering, there are definite limits to decentralisation. These findings have important implications for policymakers and financial institutions exploring digital currencies, including new payment systems, novel settlement solutions and digital currencies such as stablecoins or other forms of digital infrastructure.
Source : VOXeu
