Citrini Research has urged a move away from the largest cryptocurrencies and toward DeFi protocols and tokenization infrastructure that collect fees from automated on-chain activity, arguing that AI agents could turn blockchains into the operating rails of a continuously active financial system.
In a 79-page report dated Oct. 8, Citrini said portfolios built around Bitcoin, Ether and XRP may offer less direct exposure to the cash flows created by agent-driven finance than protocols handling derivatives, lending, staking, payments and settlement. Its model crypto basket excluded Bitcoin and Ether, instead assigning roughly 10% each to Derive, Lighter and ether.fi.
The report’s central bet is that AI programs acting for users and businesses will increasingly move money, seek yield, trade assets and settle transactions around the clock. Blockchains, with programmable transactions and always-open markets, could supply the infrastructure for that activity. Under that scenario, the largest gains would not necessarily accrue to the tokens that sit beneath the networks, but to platforms that charge fees each time an automated agent uses them.
Citrini’s position challenges the familiar strategy of holding major cryptoassets as broad exposure to the sector’s growth. It places greater weight on whether a project can capture and distribute revenue generated by actual usage.
A portfolio built around protocol revenue
Derivatives-related positions made up about 39% of Citrini’s crypto basket, according to the report. Yield and staking-related allocations represented about 27%, while blue-chip infrastructure accounted for about 26%.
Alongside Derive, Lighter and ether.fi, the portfolio included Solana, Hyperliquid’s HYPE, Aave, Ethena, Aerodrome, Maple, Pendle and Ondo Finance. The mix leans heavily toward platforms designed to earn transaction fees, facilitate leveraged trading, route liquidity or package on-chain yield.
Citrini said Derive gained 1,300% over the preceding year, while most other tokens in its selected basket recorded double-digit gains. Bitcoin, Ether and XRP lagged that group over the same period, according to the report.
Past performance alone does not establish a durable advantage, particularly in thinner or newer token markets where prices can move sharply. Citrini acknowledged that higher on-chain activity does not automatically translate into token appreciation. A protocol’s ability to retain fees, manage incentives, compete for liquidity and navigate regulatory constraints remains central to whether activity becomes value for token holders.
That distinction explains the report’s focus on projects with buybacks, revenue-sharing mechanisms or other forms of token-linked economic distribution. In practical terms, an AI agent making frequent swaps, borrowing against collateral or shifting assets between yield venues could create recurring fees for the protocols processing those transactions. A base-layer token may benefit from network demand, but it does not always provide holders with a direct claim on those fees.
AI agents could accelerate competition for deposits
Citrini also outlined an “agentic bank run” scenario, in which millions of AI agents systematically transfer idle cash to the best available yield venue. Traditional banks rely heavily on relatively stable, low-cost deposits to fund lending and manage liquidity. Automated systems able to compare rates and move money continuously could make that funding base less predictable.
The report did not present the scenario as an imminent event. Instead, it used the example to show how agent-based software could alter the speed of competition between banks, tokenized cash products, stablecoin platforms and DeFi lending markets.
Citrini argued that today’s financial infrastructure was not designed for persistent, automated reallocation of deposits at machine speed. An agent acting within set risk limits could monitor yields far more frequently than a retail customer, then shift funds without waiting for banking hours or manually navigating multiple applications.
Such behavior could favor systems offering fast settlement, transparent pricing and programmable transfers. It could also expose weaknesses in protocols that advertise high yields without durable revenue, sufficient liquidity or clear risk controls.
Tokenization shifts attention to trading and settlement platforms
The report said tokenization is moving the debate beyond whether real-world assets can legally come on-chain and toward which platforms collect revenue from their trading, lending, settlement and collateral use.
RWA.xyz data showed Ethereum hosted $16.9 billion of the $39 billion in tokenized assets tracked as of Oct. 9. That concentration gives Ethereum a major role in the tokenization market, though Citrini argued that owning Ether is largely a broad bet on ecosystem growth rather than a direct claim on every transaction involving tokenized assets.
Citrini cited an episode during the Iran conflict in which it said a decentralized exchange offered the most accurate weekend oil price, calling it a “Hyperliquid moment.” The report linked such episodes to growing interest in tokenized assets that can trade outside traditional market hours and be used across multiple applications.
Tokenized stocks and other assets can be made composable, meaning they may be traded, pledged as collateral, borrowed against or incorporated into other on-chain financial products. That structure could give specialized trading and lending protocols multiple ways to generate fees from a single tokenized asset.
Equity picks favor market infrastructure
Citrini applied a similar approach to listed equities. Its equity basket gave Securitize the highest weighting, followed by Circle and Coinbase. It excluded MicroStrategy and cryptocurrency mining stocks.
The choice suggests Citrini sees greater potential in companies that issue, custody, trade or support tokenized financial products than in businesses whose fortunes are more directly tied to Bitcoin’s price. The report’s broader warning is that rising crypto usage should not be confused with uniform gains across all digital assets.
Bitcoin rebounded from below $60,000 in July to above $80,000, according to Citrini, even as the report described broader digital-asset prices as under pressure. The firm warned that a market-wide selloff could still alter Bitcoin’s trading behavior and correlations.
For traders, the report frames the next stage of crypto markets as a question of business models: which protocols can convert automated activity into durable fees, and which tokens merely provide exposure to a growing system without capturing much of its economic output.
Explore how AI and DeFi reshape trading: dive into web3, AI, and crypto for the next wave of on-chain finance.
Disclaimer: The content on this page is provided for general informational purposes only and does not represent the views or financial advice of Toobit. We make no guarantees regarding the accuracy or completeness of this information and shall not be held liable for any errors, omissions, or outcomes resulting from its use. Investing in digital assets involves risk; users should independently evaluate their financial situation and the risks involved. For further details, please consult our Terms of Service and Risk Disclosure.
