Financial Intermediation and Disintermediation: What Coase's Theory of the Firm Tells Us About Crypto Lending

Ronald Coase studied the concept of firms in economics in his 1937 article. The central question in Coase's investigation was whether firms would be necessary if market forces could effectively connect buyers and sellers. Coase's hypothesis was that the structure of a firm is determined by the transaction costs associated with conducting business in an open marketplace. When it is less expensive (in terms of time, resources, and money) to conduct business through a corporate (firm) structure rather than the open marketplace, then a firm is formed. Banks provide an example of this situation. A person cannot easily find a potential lender, verify that the borrower is capable of repaying the loan, and enforce repayment all by themselves. Crypto loan protocols argue that the logic of Coase does not apply any more. They claim that smart contracts can provide a more cost-effective and efficient way of connecting lenders and borrowers without the use of a bank or any intermediary.

The Three Jobs a Bank Is Built to Do

Most economists have classified bank intermediation into three categories: maturity transformation, risk pooling, and monitoring. Maturity transformation is the process by which a bank takes its customers' short-term, liquid deposits and converts them into long-term, illiquid loans. Risk pooling is the process of combining the funds of many depositors into a single pool to minimise the effects of the default risk from any one of those depositors on the entire pool of loans.

Monitoring is the process by which banks assess their borrowers both before making a loan and after it has been made, to ensure that borrowers fulfil their loan obligations. By monitoring their borrowers, banks reduce the information asymmetries between lenders and borrowers.

The functions of these three categories of bank intermediation are carried out by the code in cryptocurrency lending platforms. The process of pooling deposits occurs algorithmically. Interest rates on loans are set by the market through supply and demand, rather than at the discretion of a loan officer or banker. The function of monitoring is replaced by over-collateralisation (i.e. requiring more collateral than the actual amount of the crypto loan).

Read also: - Cryptocurrencies and Inflation Hedging

A Trillion-Dollar Question Being Tested in Real Time

 As of April 2026, DeFi (Decentralised Finance) lending has transitioned from the fringe to widespread utilisation, with just under $54 billion in total deposits across just over 380 DeFi lenders in operation today. The largest of those lenders is responsible for approximately $19 billion of that amount. All of the funds held in deposits at those lenders are transacted without a loan officer's approval and instead through the creation and evaluation of the cryptocurrency lending practice, and thus the situation provides a unique testing ground to determine whether the Coasean 'make or buy' market economics will be a viable paradigm in the overall market for finance.

Where Trust Actually Goes When You Remove the Bank

This idea is critical to the discussion. The traditional Diamond and Dybvig Model of 1983 explains how banks are inherently fragile because they use depositors' liquid deposits to fund illiquid loans. In such a scenario, when a large number of depositors fear that other depositors will withdraw their deposits first, it is in the depositors' best interests to do so as well; the depositors will collectively exit the system, leading to a 'bank run'.

Many proponents of cryptocurrency lending touted the opportunity to remove discretion from intermediaries in their lending methods; however, as evidenced by recent events, there is still the potential for fragility in this market, but it has simply migrated to a different location within the ecosystem.

In April 2026, a malicious actor exploited a $292 million security gap on a bridge connected to a popular restaking token and then utilised the stolen tokens as collateral to borrow against on an industry-leading lending platform. Although the malicious act did not directly involve any of the lender's contracts, the overall TVL (Total Value Locked) of the lender's platform dropped by an estimated $6.6 billion within 24 hours of the attack, and the lending platform's governance token subsequently lost 16% of its value. Over the course of 48 hours, the entire DeFi ecosystem suffered through over $13 billion of net redemptions as users withdrew their funds across DeFi platforms out of fear of instability, thus creating a bank run via the use of their wallets instead of via a teller.

The trust associated with lending was not eliminated. Rather, it simply moved from one type of entity to another: the bridge provider, price oracle, and governance token holders, who vote on which collateral is deemed acceptable.

See also - What it means to have a cryptocurrency business account. 

What This Means for How We Think About Financial Firms

If trust moves rather than disappears, then calling the process of lending with cryptocurrency 'trustless' is more accurately a marketing term than an economic one. The true question, as per Coase, is not whether cryptocurrency lending is cheaper than traditional forms of lending; it is which part of the current model is cheaper and where the costs are now going. For example, while monitoring costs (the costs associated with the risk of lending) do not go away simply because a cryptocurrency loan is overcollateralised, they now reside with liquidation bots and regular depositors who have to evaluate the risk involved with smart contracts, the design of price oracles, and the cross-chain risk independently from the way a bank's risk department did previously.