Author: Andrew Hong Source: cryptodatabytes Translation: Shan Ouba, Golden Finance

Today, many people, including BlackRock CEO Larry Fink and Robinhood CEO Vlad Tenev, believe that everything can be tokenized. Although this statement may seem exaggerated and easy to overlook, the logic behind it is that on-chain assets can provide far greater liquidity, flow, and return rates than off-chain.

But if you do not truly understand the underlying logic of cryptocurrencies (instead of just staying on the surface of tokens), you may feel that the new wave of application fever is driven only by homogeneous stablecoins and yields from unknown sources. Even though all information on blockchain technology is transparent and traceable, it is still difficult to see the underlying patterns and strategies behind it; you might also mistakenly believe that aggressive token incentives are still driving industry growth. But in fact, the era from 2020 to 2023, which relied solely on mining arbitrage and capital rotation, has basically come to an end.

Decentralized Finance (DeFi) is constantly evolving in each round of the crypto cycle, condensing hundreds of years of experience in traditional finance and economics into on-chain protocols — these protocols are more efficient, trustless, and permissionless than their off-chain counterparts. Initially, assets were packaged into tokens, then tokens were pooled together to support lending and trading, and now it combines funding pools with portfolio management and complex structured products through treasury-like protocols.

The concept of the profit margin logic from the source of income is quite simple: you can deposit multiple tokens into a liquidity pool and vault, earning returns through aggregation. Whenever users conduct financial transactions using these tokens, you can earn transaction fees:

This type of interest rate differential has actually existed in traditional finance for a long time, but it is only open to large intermediary institutions (banks, funds), and ordinary users can only earn a negligible 1 cent in their Chase Bank accounts. In the cryptocurrency space, these earnings belong directly to the users/liquidity providers.

In the past five years, the opportunities and complexities of DeFi have grown exponentially — next, I will break down the core logic for you one by one. If you want to establish yourself in the DeFi space, you must understand these foundational concepts and their connections:

• Tokens: How to Build a Token Flywheel Worth Billions in Market Capitalization
• Capital Pool: How tokens access trading / lending markets to enhance capital efficiency
• Vault: How to manage liquidity across funds pools for safer high returns

This article will focus on the business logic behind these concepts (what and why), rather than the technical details at the code level (how). The three mentioned operate through smart contracts, with multiple sets of smart contracts working in collaboration to form a protocol.

Understanding the Token Flywheel

Tokens are essentially smart contracts that follow specific standards (ERC20, ERC721, ERC1155). For stablecoins and assets aiming to become blue-chip tokens, there are two core objectives:

1. Sustainably manage the issuance and redemption of a certain type of asset to give tokens value.
2. Build a liquidity cycle on top of the token wrapping layer to enhance token yield and ecological impact.

Any token that can achieve these two goals will ultimately form a powerful growth flywheel:

It should be noted that the “flywheel” here does not refer to inflating the market value to 1 billion dollars through means such as being listed on centralized exchanges or digital asset treasury, but rather to the actual application scale of the token within the DeFi ecosystem reaching this level.

Due to the need for tokens to circulate among numerous ever-changing protocols, most people view tokens as a single asset, only focusing on superficial trading discussions — this understanding is flawed.

Let's start with the issuance and redemption management of tokens. Tokens backed by actual collateral are usually managed through a code module that delegates the minting and burning functions to another contract or a whitelisted address.

The following are the main categories of stablecoins (including issuer / deployment time):

1. Fiat-collateralized stablecoin

• Represents: USDC (Circle, 2018), USDT (Tether, 2017), PyUSD (PayPal, 2023)
• Features: 1:1 pegged to fiat currency redemption, holding short-term government bonds (T-bills) and other high-liquidity quality assets off-chain as collateral. Most use off-chain redemption mechanisms, but some also support on-chain redemption (e.g., Bold, Liquity). Nowadays, anyone can launch similar stablecoins through integration tools — for example, issuing off-chain/reserve management through Bridge (acquired by Stripe), and on-chain liquidity/cross-chain token management through m0 (such as mUSD, MetaMask related stablecoins).

2. Over-collateralized Stablecoin

• Represents: DAI (Maker/Sky, 2019)
• Features: Users deposit collateral (initially ETH), borrow such stablecoins, and pay a lending rate (stable rate) for their positions. This model is known as Collateralized Debt Position (CDP). Maker also sets a savings rate for DAI (similar to the Federal Reserve's federal funds rate), and the protocol can balance supply and demand and maintain a soft peg to 1 USD by adjusting these two rates.

3. Algorithmic Stablecoin

• Represents: Fei/Tribe (Fei Protocol, 2021), UST/LUNA (Terra, 2021)
• Features: Attempts to maintain value and prevent runs through a complex dual-token economic model, but all such stablecoins have ended in failure; however, their design logic is still worth studying.

4. Delta-Neutral Stablecoin

• Represents: USDe (Ethena, 2023), Neutrl (Neutrl, 2025)
• Features: Users can deposit specified collateral (such as ETH, USDT, and other tokens) to mint USDe. Ethena will conduct hedging operations on the public market for the collateral — profiting from the funding rate by shorting perpetual contracts (most of the time the short side will gain funding fees), while hedging against the risk of collateral price declines. Currently, Ethena has launched a stablecoin issuance service, capable of managing token issuance (in collaboration with Anchorage) and on-chain yield.

Overall, the evolution trend of tokens is: the types of collateral are becoming more diversified, and the over-collateralization rate is gradually decreasing. For example, the collateralization rate of delta-neutral strategies like Ethena is only 101%, while the collateralization rate of DAI is about 270%. Due to the high unpredictability of complex economic models, the era of algorithmic stablecoins has officially ended.

Next, let's talk about wrapped tokens and their role in the flywheel. The most typical wrapped token is WETH, which converts the native Ethereum asset ETH into an ERC20 token. This conversion is essential because most contracts are built on the ERC20 standard, and WETH simplifies the interaction logic. You can unwrap WETH back to ETH at any time to redeem the underlying asset; the amount of ETH is only linked to the amount of WETH and is unrelated to the wallet that initially performed the wrapping operation.

Protocols can achieve various functions through wrapped tokens, such as:

• Interest / Reward Accumulation: The balance of wrapped tokens will automatically increase over time (similar to the aTokens in Aave), and these tokens are sometimes referred to as rebase tokens.
• Modular splitting of underlying tokens: for example, splitting Aave's yield token into principal tokens and yield tokens (such as Pendle protocol); or splitting NFTs into multiple fungible tokens to achieve fractional ownership (such as Fractional.art).
• Wrapped Tokens for Batch Packaging: For example, in a Uniswap liquidity pool, wrapped tokens are needed to represent the two assets in a trading pair; or in a vault, wrapped tokens represent the user's share of all assets in the vault contract (such as MetaMorpho).

Each function mentioned above only provides one protocol example, but these are common patterns used by hundreds of protocols on Ethereum.

Since wrapped tokens follow the ERC standard, they are often reusable in other protocols. Taking USDe as an example (which is a typical case of the current strongest token flywheel, with a market value of 12 billion dollars as of October 22, 2025):

1. Stake USDe to earn Ethena protocol revenue sharing and receive sUSDe;
2. Split sUSDe into Principal Tokens (PT-sUSDe) and Yield Tokens (YT-sUSDe) through Pendle.
3. Deposit the principal tokens into the Morpho market to earn returns through the PT-sUSDe/USDC trading pair;
4. Or deposit into the Morpho vault (which holds a diversified position, where the PT market position may account for 40%), to receive MC_PTS.

sUSDe - (Splitting)

You can view all these wrapped tokens as extensions of the underlying USDe token – they possess additional value and functionalities that the underlying token does not have. When analyzing a token, it is necessary to sort through all its associated wrapped tokens to fully understand its value and growth potential (although this is easier said than done).

The Operating Mechanism and Profit Model of the Capital Pool

In the above case, the returned wrapped tokens represent a share of the “funding pool,” which is the contract that receives deposits of underlying tokens. Nowadays, the “funding pool” has become a general term referring to any contract that can hold deposits of tokens from multiple users simultaneously and is convenient for other users to call. It is a peer-to-peer (P2P) system managed by protocols responsible for all ownership records, calculation logic, and order matching rules. The core purpose of the funding pool has always been to maximize capital efficiency for a series of financial orders/operations and to allow users to profit through the acquisition of usage fees.

The following are the main types of liquidity pools you need to understand, arranged in the order of their development and popularity in the ecosystem:

1. Lending Fund Pool (Money Market)

• Operations: Deposit, Borrow, Circular Staking
• Pegged Token: A rebase token issued based on floating interest rates, targeting the deposited or borrowed amount.
• Development History: Collateralized Debt Positions (CDP, such as over-collateralized DAI lending) initiated the foundational on-chain lending model; in 2019, Aave and Compound launched aggregated money markets, allowing users to conduct deposit and loan operations in a shared liquidity pool. These funds pools each have an interest rate curve, with the curve setting deposit and loan rates around the “target utilization rate.” Before reaching the target utilization rate, the interest rate curve rises slowly; once it exceeds the target, the curve steepens sharply until the utilization rate reaches 100%. Each token is also equipped with a customized risk profile, which determines the Loan-to-Value (LTV) and Liquidation Loan-to-Value (LLTV) parameters - the risk profile and LTV/LLTV parameters are set based on the collateral, clarifying the amount of funds the user can borrow by pledging the asset. Between 2020 and 2023, various attempts emerged in the industry to optimize collateral management and liquidation auction efficiency (such as Rari Isolation Pools, Alchemix self-repaying loans, B-protocol aggregate liquidation). “Flash loans” also originated from this, allowing users to borrow and repay large amounts of funds without collateral within a single block.
• Key Trends: Tokens must be manually added to such markets through governance proposals (with few exceptions), leading to frequent controversies surrounding high-risk assets like CRV, and attempts by some malicious entities to manipulate the liquidity pool system. As more institutional users enter the crypto space and physical assets (RWA) are included as collateral, the entry requirements for liquidity pools and counterparty risk management will become stricter. From an institutional perspective, KYC and non-KYC funds must be strictly differentiated, resulting in the emergence of whitelist liquidity pools like Aave Horizon RWA. To cover a broader range of institutional borrowers and facilitate under-collateralized lending, some products adopt a model of “liquidity pools receiving regular deposits, then lending to borrowers through private management and credit scoring/seizure mechanisms” (represented by Maple and 3Jane, respectively). In the future, as thousands of stablecoins and physical assets (RWA) launch, the industry will face a set of contradictions: issuers want their tokens included in liquidity pools to generate returns, but liquidity pool protocols must meet institutional requirements, imposing strict limitations on asset classes and participants. This dynamic is worth monitoring.

2. Trading Fund Pool (Liquidity Pool)

• Operation: Add / Remove Liquidity, Exchange
• Wrapped Tokens: LP tokens (Liquidity Provider tokens), which can be redeemed for an equivalent share of the underlying liquidity pool tokens, plus accumulated trading fees (the fee rate is usually set between 0.05% and 1% of the trading volume). Since the balance of liquidity pool tokens may change after the deposit, users may face impermanent loss (which is actually a permanent loss).
• Development history: Starting with the dual-token trading pair based on the x*y=k curve of Uniswap, automated market makers have gradually improved issues related to price impact and idle liquidity in pools by optimizing mathematical models and the granularity of price curves (exchange rates). The industry has seen the emergence of stablecoin-specific curves, 8-token weighted curves, and centralized curves designed for managing liquidity within specific price ranges; recently, the “Liquidity Pool Hook” feature was introduced, allowing for greater composability with other protocols before and after exchange operations. Unlike lending pools, liquidity pools always maintain a permissionless characteristic, allowing anyone to add any token (due to the lack of collateral-related connections between trading pairs). Over the past year, token issuance platforms (launchpads) have also become a focus, heavily influenced by trading curves tied to initial liquidity. New protocols such as Pump.fun, Clanker, and Doppler (adopted by Zora) have introduced unique liquidity guiding or binding curves, providing initial momentum, price discovery mechanisms, and anti-sniping strategies for tokens.
• Key Trends: Both lending pools and liquidity pools aim to reduce idle funds, therefore the new generation of protocols introduces extended functionalities to enable seamless fund transfers between the two. For example, Balancer's enhanced pools deposit idle liquidity into interest-bearing pools (like Aave); Uniswap hooks allow protocols such as Eulerswap to import lending pool liquidity into Uniswap for Just-In-Time (JIT) Swap. Such rebalancing mechanisms are also becoming increasingly popular within individual pools, such as Arrakis managing price targets for concentrated liquidity, and Morpho's PublicAllocator used to balance target interest rates in the lending pool market.

There are many controversies regarding the order matching mechanisms of trading pairs (such as AMM, Central Limit Order Book (CLOB), Intentional Market Makers, Perpetual Contract DEX, etc.), but these differences are not closely related to the core theme of this article and can be discussed separately later.

3. Staking Fund Pool (Insurance Fund)

• Operations: Staking, Locking
• Wrapped tokens: Tokens that usually have an “s” prefix (e.g., sUSDe), holders are generally entitled to receive a proportional share of the protocol's revenue.
• Development History: In the staking scenario, users typically need to pledge tokens. If a black swan event or negative incident occurs, the staked assets may face the risk of confiscation (loss of staked shares). Specific scenarios include: staking ETH on validation nodes, where if the block proposal fails, penalties will be imposed; absorbing bad debts for lending pools like Aave. Sometimes, staking serves merely as a locking mechanism to prevent tokens from being immediately sold off to maintain supply stability—such as Ethena's USDe being converted to sUSDe. In all these scenarios, protocols usually allocate a portion of the income directly to stakers to compensate for the additional risks or liquidity losses they bear.
• Key Trend: Staking has been, and will continue to be, a core component for maintaining protocol security and providing initial returns to users. It is the simplest type among all liquidity pools, yet it is a critical link in the token flywheel. You may hear the term “Restaking,” which essentially means using existing blue-chip tokens to stake in other protocols (rather than issuing the protocol's native tokens) — its mechanism is consistent with ordinary staking and does not require excessive interpretation.

4. Incentive Fund Pool (Meter)

• Action: Vote
• Wrapped Tokens: Typically add a “ve” prefix before staking tokens (starting with Curve's veCRV). Holders can earn protocol tokens as rewards in addition to the usage fees from liquidity/borrowing pools.
• Development process: Suppose you are a token issuer facing multiple lending/liquidity pools with similar returns. How do you attract users to choose the liquidity pool where your issued token is located? The answer is naturally to add obtainable token incentives. Many protocols (whether they belong to token, liquidity pool, or treasury types) issue protocol tokens for governance or ownership rights. These tokens can then be used to continuously provide incentives to liquidity providers (referred to as “emissions”). For example, liquidity pool protocols like Aave can issue tokens to users who deposit in their pools; Ethena can additionally add ENA incentives in Aave's aEthUSDe pool to provide higher returns. The emergence of incentive liquidity pools is essentially a way to help protocols compete for user liquidity through targeted token incentive allocation. Curve pioneered the “voting escrow (ve)” mechanism, where token holders can lock CRV and vote to decide the direction of incentives (to enhance their own returns). This type of mechanism is known as “gauges.” Subsequently, a large number of aggregation layer protocols (such as Yearn, Convex, Tokemak, Butterfly, Olympus) emerged to compete for voting dominance over the underlying protocol liquidity pools. This trend coincided with the airdrop and “vampire attack” era — teams like Sushiswap would fork other protocols and attract competitors' users and funds through high incentives. The previously seen 1670% super high returns originated from this type of mechanism.
• Key Trends: Although incentive liquidity pools represent a chaotic yet noteworthy chapter in the history of DeFi liquidity pools, their current scale is much smaller than before. Aerodrome and Pendle are typical examples of modern incentive liquidity pool voting mechanisms. Fortunately, the technology developed for such protocols has successfully transitioned to “treasuries” for more professional DeFi strategy management. Yearn and Tokemak (Auto Finance) remain as representative projects from the post-“mining arbitrage” era.

5. Strategy Fund Pool (Vault)

• Operation: Deposit, Redeem
• Wrapped Tokens: Vault Token Shares (calculated as a percentage). This percentage ensures that users can automatically redeem any earnings generated by all tokens within the vault.
• Development History: The vault mainly originated from the aforementioned “mining based on emissions and incentives” era, and it now uniformly follows the ERC4626 standard, with a core focus on efficient allocation and rebalancing across liquidity pools. Details about the vault structure will be elaborated separately in the following text, so it will not be reiterated here.

6. Structured Fund Pool (Structured Finance)

• Operations: Split, Divide
• Wrapped tokens: Diverse in form, the core represents a certain component of the underlying tokens (and their associated rights/benefits), with the goal of creating wrapped tokens that are more liquid. Common prefixes include “L”, such as Liquidity Staking Tokens (LST), Liquidity Yield Tokens (LYT), Liquidity Re-staking Tokens (LRT), Liquidity Locking Tools, etc.
• Development History: After receiving assets, this type of funding pool restructures them into smaller components. Token splitting can create more reuse scenarios for wrapped tokens, with Pendle being the most successful case — it splits principal tokens (PT) and yield tokens (YT) from yield-bearing tokens and is widely used in the DeFi ecosystem. During the NFT boom in 2022, the industry also made multiple attempts at “fragmentation”, such as splitting NFTs into fungible shares to achieve partial ownership.
• Key Trend: As DeFi matures, it is expected that more protocols will convert traditional structured financial products into crypto-native products. The technical implementation of such liquidity pools is the most challenging, and they often face difficulties in user communication and collaboration.

Vault Management and Revenue Strategy

From a protocol perspective, the yields of many liquidity pools are independent of each other, but they are interconnected through wrapped tokens. How can users easily manage all these wrapped tokens and liquidity pools to optimize their yields? The answer lies in “strategies.”

From 2019 to 2022, users had to manually discover and manage strategies across tokens and liquidity pools. Some popular strategies include “leverage stacking”—for example, depositing $100 worth of ETH in a lending pool, borrowing $80 in USDC, converting the USDC back to ETH and depositing it again for additional returns. This strategy carries the risk of liquidation if the ETH price falls below the LTV threshold, but as long as the USDC borrowing rate is lower than the ETH deposit rate, it can be profitable. Entering various mining vaults usually requires completing multiple cross-transaction conversions/deposits/staking operations. To address this issue, Solidity developers aggregate these function calls into a single contract, allowing users to establish a position with just one transaction—this mechanism is known as “zap” (first seen in DeFi Zap). Subsequently, products like Zapper, Instadapp, and Defisaver developed numerous zap contracts for each newly launched protocol, akin to “whack-a-mole” in response to rapid iteration. Given the security risks, deploying such contracts is also extremely challenging—you can glimpse their complexity by understanding the architecture of Defisaver. As expected, no single team can keep pace with all new DeFi protocols/versions/vaults. Today, products aimed at retail investors (such as the aforementioned leverage management tools, or combination/ETF management tools like Reserve and Glider.fi) have improved user experience by tenfold compared to 2020.

Therefore, allowing users to deposit assets into the “self-managed various zaps” vaults (and referring to zaps as “strategies”) becomes a natural choice. Currently, most vaults follow the ERC4626 standard, which includes the following core elements:

1. User: Deposit a certain type of token into the vault to receive a wrapped token that represents a percentage share of all tokens in the vault - this share includes all earnings that can be redeemed at any time from the vault.
2. Strategy management function: Functions used to manage “strategies” within the treasury, which are responsible for transferring tokens and adjusting positions (these functions are sometimes referred to as “adapters” because they require custom logic to fit specific protocols and liquidity pools).
3. Curator: Responsible for defining and executing strategies long-term, and adding/removing tokens.

Like all ERC standards, numerous extension standards (such as ERC7540) have been derived from the basic standard of ERC4626. Some extensions have not formed a unified standard but are specific to protocols, such as the adapters in Morpho Vault V2.

The current vault strategies are quite similar: the vault deposits tokens into a set of lending/liquidity pools, with most strategies not using or only using a small amount of leverage. The best case for manually managed vaults is Morpho—managers implement strategies by controlling the limits and capital flow across different markets; Yearn and Lagoon are also quality examples. In addition, there are automated vaults like Sommelier and Beefy, as well as products like the Hyperliquid vault that support trading strategies. The industry is currently in an exploratory phase, with teams dedicated to creating flexible vaults that “cover the most capital pools/tokens and can execute and manage strategies within them.”

Taking the Morpho vault as an example, the yield structure for users (depositors) is as follows:

Morpho treasury's “native fees” come from the utilization rate sharing of underlying lending pools in different configurations; MORPHO and SEAM tokens are additional incentive emissions; the treasury manager collects a high performance commission of 15%.

There are transparency issues with the current vault: users often cannot understand the actual strategies being executed underneath. The Herd team is working to address this issue and will launch relevant features in the future. Please exercise caution when depositing into high-yield vaults.

Combining the content learned earlier, taking the Ethena stablecoin USDe as an example, all elements in the token flywheel can be linked together to form the following flowchart:

Every liquidity pool, wrapped token, and vault adds value to the underlying USDe token; as Ethena evolves, the ecosystem will drive the creation of more complex liquidity pools and vaults. All of this will generate more demand for the minting of USDe.

Clearly, the token issuers want to control the profits generated by this flywheel as much as possible. Therefore, it has become increasingly common for issuers to launch their own treasury, such as Perena and Midas (whose security remains to be seen, and is currently mentioned only sporadically).

Summary Thoughts

If you can read this, first, thank you for your willingness to engage in deep reading in the AI era — you should now have a clearer understanding of the development history and core logic of DeFi. In the future, when you hear others discussing tokens or liquidity pools, you will know how to organize the relevant connections between tokens, liquidity pools, and vaults, and engage in more in-depth discussions.

It should be noted that this article does not cover many underlying concepts and details related to the risks associated with tokens, liquidity pools, and treasury operations. These risks are often related to cross-protocol token pricing and oracle issues (aggregating off-chain price data / calculation results as a reliable source for on-chain token pricing) — it is recommended that you refer to relevant materials to understand existing oracle mechanisms. Additionally, attention should also be paid to the issue of maximum extractable value (MEV) and front-running bots.