A vault in DeFi is a smart contract that pools deposits and puts them to work in yield-generating strategies, issuing each depositor a token that represents their share of the pool.
In under five years, this structure has moved from a niche convenience to the primary way capital is deployed onchain. This article traces how that happened, across three distinct generations of vault design, and explains why the shift has the shape of a structural change rather than another passing crypto cycle.
First Generation: Single-Strategy Yield Farming
The first widely adopted vaults emerged from Yearn Finance during 2020 and the yield-farming period of 2021. These were single-strategy products built to capture the token incentives that protocols were offering at the time. Annual rates regularly exceeded 1,000% on certain tokens, so the opportunity was clear. Yet, capturing it was operationally demanding. A user had to claim reward tokens, swap them back to the underlying asset, and redeposit, often paying $300 to $500 in gas per transaction on Ethereum.
Yearn solved this by pooling capital and automating the harvest, convert, and redeposit cycle. A user could deposit once and let the vault handle the rest. At the scale of a shared vault, the gas cost per user dropped sharply, which made strategies profitable that would have been uneconomical for an individual acting alone.
The vault's original value proposition was narrow and concrete: gas optimisation and auto-compounding at scale. Pool capital, automate execution, return proceeds proportionally.
The significance of the first generation was the mechanic it proved rather than the rates it captured. That mechanic survived the end of the farming period that produced it. When the high emission rates faded, the pooling-and-automation pattern remained useful for any strategy that involved recurring onchain actions, which is why the design persisted into conditions very different from the ones that created it. The first vaults are best understood as the moment the industry learned that capital could be pooled and managed programmatically at scale, a lesson that outlasted the specific opportunity it was built to capture.
Second Generation: Multi-Strategy Architecture
Yearn's v2 release introduced multiple strategies inside a single vault. A deposit could now gain diversified exposure across several pools and protocols at once, rather than tracking a single farm. Users described this as a set-and-forget approach, where one deposit produced exposure to multiple yield sources and the vault handled rotation between them.
This design created a new role: the strategist, the person who decides which strategies the vault deploys into, how to weight allocations, and when to rotate out of declining opportunities.
The complexity of running a vault increased substantially. And so did the appeal, because that complexity moved off the depositor and onto a specialist. For anyone without the time or technical depth to manage positions across a fragmenting landscape, delegating to a multi-strategy vault was a clear improvement in quality of life. This is the point at which the vault stopped being a tool and became a managed product.
The strategist role also changed who carried the analytical burden. In the first generation, a depositor still had to choose which farm to enter. In the second, that choice moved to a specialist who monitored opportunities and rotated between them. The depositor's decision narrowed to selecting a vault and a manager rather than a position. That separation between the capital and the decision-maker is the thing a depositor ultimately relies on, because the quality of the manager now determines the quality of the outcome.
Third Generation: Institutional Infrastructure
The current generation has moved well beyond farming. Two developments in particular reshaped it.
First, modular lending protocols introduced curated vault architectures that allow isolated markets with distinct risk profiles. Instead of one shared pool, a protocol could host many separate markets, and a vault could allocate across a chosen subset of them. These were the first vault designs that genuinely suited institutional capital, because they offered granular control over exposure, transparent governance, and non-custodial withdrawal, meaning the depositor keeps ownership of the assets throughout. An institution could see exactly which markets its capital touched and choose accordingly.
Second, the vault-as-a-service model matured. Providers such as Veda, Concrete, and UpShift built infrastructure that lets any protocol or ecosystem launch a multi-strategy vault quickly. This catalysed a wave of themed vaults built around specific assets and ecosystems.
EtherFi vaults deployed into restaking strategies.
Lombard vaults generated yield on Bitcoin.
Berachain and Plume used vault infrastructure to attract pre-deposit capital before their ecosystems fully launched.
Across these examples, the vault became the standard distribution mechanism for DeFi yield, and the standard way a new ecosystem bootstrapped liquidity.
Two Signals of Scale
The first signal is supply. Vault creation has become close to commoditised. A protocol, an ecosystem, or an institution can spin up a vault in days, and many have. The result is a large and growing population of vaults across protocols and chains, with the vault now serving as the default way DeFi yield reaches depositors. Breadth of supply is itself evidence that the pattern has won, because builders reach for it by default rather than as an experiment.
The themed vaults that resulted were built around specific assets and ecosystems rather than generic yield. A vault might exist to deploy into one ecosystem's restaking strategies, to generate yield on a specific form of Bitcoin, or to gather pre-deposit capital for a chain that had not yet launched. Each of these is a distribution decision as much as a financial product. The fact that builders reached for the vault format by default, across very different goals, is itself evidence of how standard the pattern had become. The vault stopped being one option among several and became the assumed container for onchain capital.
The second signal is the composition of the capital. In the early period, vault deposits were overwhelmingly retail. Today, institutional participants represent the majority of deployable capital: crypto-native funds, treasuries, and exchanges building yield programs for their users. This change represents a significant change in volume, but also raises the bar for risk management, transparency, and operational rigor in ways the first generation of builders never had to consider. Capital that answers to investment committees and auditors brings expectations about disclosure, governance, and accountability that retail capital did not impose.
Why the Change Happens at the Infrastructure Level
Growth driven by a single incentive cycle fades when the incentive ends. The vault's growth has a different shape. It tracks a durable preference: capital wants exposure to onchain strategies without carrying the full operational burden of running positions directly. Modular structures that allow isolated risk and configurable parameters serve that preference better than shared pools where governance sets the risk profile for everyone.
The direction of travel has stayed consistent across bull and bear conditions, across the collapse of farming rates, and across the arrival of institutional capital. Consistency across regimes is the mark of a structural change rather than a cyclical one.
The composition shift reinforces the same conclusion from a different direction. Retail capital chases rates and leaves when rates fall. Institutional capital allocates against a mandate and stays while the mandate holds. A market that has moved from the first kind of capital to the second has changed its base of demand. Such change is far harder to reverse than a change driven by a single incentive.
The commoditisation of vault creation carries a consequence worth holding onto. When launching a vault is trivial, the supply of vaults grows faster than the supply of disciplined risk management. The interface looks uniform across products, but the quality underneath varies widely, and the depositor cannot read that quality from the surface. A clean front end can sit on top of careful curation or careless curation, and the two are indistinguishable until conditions turn.
The shift is best understood as a change in how DeFi works. Capital now reaches onchain strategy through a pooled, managed, tokenized container, and that container has become the assumed starting point for new products and new ecosystems. The open question it leaves is one of quality. A structure that is trivial to launch produces vaults of widely varying discipline, and telling a well-run vault from a fragile one has become the central skill for anyone deploying capital in this market.
About this article: This article is based on the data and analysis in Sentora's research report, The Vault Economy: Architecture, Risk, and the Rise of Professional Curation in DeFi. Read the full report for the complete figures, sources, and detail.
