The DeFi Primer — 2024 Edition

With Decentralised Finance entering its fifth year of existence, it is about time to take one step back and examine what it actually brings to the table.

But before we dive in, we need to define what we actually mean by the term Decentralised Finance. While there are many attempts to define the term, I prefer to use my own interpretation from a DeFi user’s perspective.

My mental model

I like to understand the set of DeFi applications in a specific crypto ecosystem as an important part of its value layer.
Grasi’s Mental Model (GMM) for DeFi looks somewhat like this:

Grasi’s Mental Model (GMM) for DeFi ™️

Let me explain:

The Building Blocks

Most blockchain systems like e.g. Ethereum have found a structure like this. They care for the core infrastructure, issue a native token and — in the case of proof-of-stake blockchains — offer benefits for direct or indirect stakers who secure and validate the system. So, staking and collecting staking rewards can be seen as the built-in value layer for these chains.

Blockchains aren’t that interesting when we understand them as closed systems that just focus on their built-in native token and cryptocurrency. Blockchains can do so much more if they are designed in an open, permissionless and extensible fashion.

This extensibility can be used on the asset layer when some third party issues their own token (e.g. based on the ERC20 standard) — or when someone creates an NFT. I like to understand both as crypto-native assets. Creating these assets is still very close to the initially intended use on the asset layer.

Stablecoins are similar, but somewhat different. They also play on the asset layer, but they bring in an important external dependency: Price.
Most stablecoins are designed to be pegged to a certain fiat price like USD — and that price is determined outside of the blockchain ecosystem (by a piece of software called “oracles”) — which sounds harmless, but brings in a lot of complexity and attack surface. More on that later.

DeFi services are built on top of this foundation.

DeFi services use tokens of all kinds from the asset layer and build various financial use cases on top of them. Done right, DeFi services inherit the core characteristics from underlying assets: They are open, permissionless, extensible, transparent. Their logic is coded in smart contracts that can be inspected and validated by users and third parties (like auditors and regulators). But they can also be copied, augmented and extended — even by competitors.

DeFi Services come in many Flavours

Borrowing and lending

Initial DeFi players like MakerDAO focused on borrowing and lending use cases. The main idea is similar to mortgages in traditional finance: Instead of using real estate property as collateral, DeFi clients use their cryptocurrency (like e.g. ETH) to get a loan which is usually paid out in stablecoins. The main value proposition is that DeFi users want to hold their crypto long-term because they deeply believe in its price appreciation. These users lock their crypto in a DeFi service’s smart contracts (usually called “vaults”) and borrow a loan against it. They can use this stablecoin money, extract and spend it in the real world and pay the loan back over time.

While this may sound simple, original DeFi players had to innovate on many layers to make this simple concept work — some of them even created their own stablecoin, like DAI in the case of MakerDAO.

The value of borrowing and lending via DeFi has subtle, but important additional benefits: In many jurisdictions borrowing against your own collateral is not seen as a taxable event — while selling your crypto often is. Advanced users use borrowing repeatedly to increase leverage. They e.g. use their stablecoin loan to buy more cryptocurrency and increase their exposure. Certainly, these investment strategies come with significant risk if we consider the volatility of these assets.

DeFi borrowing and lending has also significant limitations: DeFi loans are in most cases over-collateralised — meaning the value of your locked crypto collateral is usually (much) higher than your loan. Depending on the system this might be 120%–500%. If the collateral falls below a threshold, positions get liquidated to repay the debt. In this case: the users’ crypto collateral is sold to the market. The nice thing about this process is that everything is coded into the DeFi services’ smart contracts. The rules are clear and fully automated. If the market crashes heavily, loans get liquidated.

If we compare this setup to traditional finance and advanced credit scoring, it seems very much inferior. In DeFi, you can’t bring your real-world reputation and financial history to the table to get better conditions. DeFi services usually rely on your on-chain solvency. Not even your on-chain history is taken into consideration — at least for now. While this might seem like a big limitation and a “step back” to financial professionals, it also has its beauty. Every user is treated equally, only facts count. That’s also the reason why some insiders like to see DeFi as the “democratisation of finance”. Romance aside, this will change once these systems get more mature. I expect that the solvency of users will certainly get considered in the future and lead to better conditions.

DEXes, Automated Market Makers and Liquidity Mining

Similarly important to borrowing and lending are DeFi use cases for trading. In the early crypto days until 2017, most trading happened on centralised exchanges like e.g. Coinbase, Kraken or Binance. These services offer convenience, but also impose a certain risk because users give away control over their assets — you might have heard about the philosophy “not your keys, not your crypto”. Centralised exchanges don’t really match the decentralised ethos of crypto assets and self-hosted wallets.

That’s why a lot of energy was put into eliminating these dependencies on centralised actors: Projects like Uniswap, Balancer and Curve created Decentralised Exchanges (DEXes) that work very differently from traditional financial marketplaces. Instead of relying and paying intermediaries like market makers to find a fair price between sellers and buyers and fulfil orders to make the market “work”, DEXes use algorithms to enable trading along bonding curves. Automated Market Makers (AMMs) were born. Initially, many market observers had strong doubts if this simple model would ever work. The huge success of DEXes like Uniswap has proven these experts wrong.

Again, smart contracts and standardised token contracts are the secret sauce to make this work: DEXes invite their users to provide liquidity for individual trading pools. In return, these Liquidity Providers (LPs) earn a part of the trading fees that get collected when a trade happens via their pools. Each pool usually consists of a trading pair, e.g. ETH and DAI. LPs usually have to deposit an equal amount of tokens on both sides of the pool, in our example 50% ETH and 50% DAI. The rest happens automatically — controlled by smart contracts.

What sounds simple, has some interesting implications. Due to fluctuating prices, liquidity providers never get back the exact same distribution of assets as they put in. Trading fees have to outweigh these risks, otherwise nobody would provide liquidity. Pools with deep liquidity are attractive, that’s why projects fight for capital for DEX pools that trade their token.

An example: Let’s say we planned to issue our own token XYZ and wanted to create a market for our new token. Because of the permissionless nature of DeFi, we could simply use Uniswap and create two new liquidity pools: one to trade XYZ against ETH and one to trade XYZ against a stablecoin like DAI. We would provide both pools with a basic initial liquidity and then invite our users and token holders to do the same. It would take us 5 minutes to create a market for our token. No coding involved. Isn’t that fascinating?

Yield Farming

These possibilities led to interesting phenomenons like projects “funding” their liquidity pools with additional incentives. Instead of relying on regular trading fee incentives by the DEX, many projects paid liquidity providers additionally with their own crypto currency to attract LPs and ensure deep liquidity for their pools.

Advanced DeFi users regularly screened the market and put their capital in DEX pools that offered the highest yield. “Yield farming” was born and led to a lot of hype — and obviously several problems when farmers suddenly sold profits or shady projects offered incentives in tokens that didn’t carry any true value.

Vaults

While all of this might sound fascinating, it also leads to a lot of complexity. It is extremely hard to stay ahead of the DeFi game and find the most promising investment opportunities across all DeFi services. Users constantly have to balance between interesting yields (usually offered by new projects) and maturity and safety of more established DeFi services. While all these systems are usually open and can be examined for quality, most users are not tech-savvy enough to actually “verify” these services properly for themselves.

High-level DeFi services like Yearn, Convex or Enzyme offer vaults that adhere to sometimes complex investment strategies or simple fund-like structures that re-balance their positions regularly.

Aggregators

It is similarly hard to find the best place to make a trade. Prices, slippage and trading fees might vary significantly between DEXes and trading pairs. Additionally, the market is full of trading bots that run tirelessly in search of arbitrage opportunities. Here, the standardisation has some downsides for naive users. Those who simply use DEXes like Uniswap for convenience rarely get the best possible price these days. Even worse, sometimes their trading transactions will fail because the trade slippage exceeds the pre-defined default limits.

That’s where established aggregating services like 1Inch or Matcha come into play. They try to offer a simple user interface and automate the complex stuff like finding the best place to trade — even splitting bigger trades between different DEXes and more advanced features. New DEXes like CoWSwap innovate even further and create mechanisms like “Coincidence of Wants” to bring buyers and sellers together in a confidential off-chain fashion — avoiding the interference of arbitrage seeking bots.

Liquid Staking and Re-Staking

While I don’t consider staking in itself as a DeFi service, liquid staking certainly is. Again, third parties improve the native platform service. Speaking for Ethereum, native a stake of 32 ETH — which means investing ~80.000$ at the time of writing. Liquid staking services like Lido, Rocketpool and others take away this burden from users.

Deposit your available ETH (current minimum: 0.01 ETH) into their service (read: smart contract), and the DeFi magic begins: You get back staked ETH tokens: stETH (Lido) or rETH (Rocketpool). Your ETH gets bundled with deposits from other users and staked in the background. As soon as you want to stop staking you simply withdraw raw ETH in exchange for your staked ETH tokens.

EigenLayer offers users additional yield who lock their stETH or rETH into Eigenlayer’s re-staking service. This collateral is used to secure third party services, like e.g. other blockchains or data availability layers like their own EigenDA.

These examples show again, how DeFi plays out its powerful composability features.

Advanced DeFi Use Cases

Beyond these mainstream DeFi use cases described above, there are many niches with more advanced stuff.

• One interesting area are derivatives ecosystems like DYDX or Synthetix with its adjacent Kwenta marketplace.
• NFT owners can use lending marketplaces like NFTX to take loans against their Bored Ape or Cryptopunk NFTs.
• Cautious users can use decentralised insurance services for DeFi like Nexus Mutual.
• Flash loans are a very special DeFi building block — a mechanism to get a collateral-free loan, do something with that borrowed money and pay it back in the very same blockchain transaction.

Again, these systems and building blocks profit heavily from DeFi’s composability and openness.

DeFi Frontends

Finally, there is the category of DeFi Frontends that try to improve the user experience across different DeFi protocols. While in the beginning most DeFi protocols came with their own, often clunky user interface, there are now applications like Summer.fi, Zapper or DeFi Saver to make the users’ lives more convenient.

These tools try to strike a balance between user friendliness and enough transparency to be clear about investment risks involved. As you can imagine, this is no simple task. While DeFi-savvy users might prefer a user interface that is very much clear about what is happening behind the scenes, DeFi newbies will prefer a lot more abstraction from that complexity. I expect several winners in this race. We will see specialised offerings for more advanced users, and simple ones for new joiners.

Beyond Ethereum

There is a plethora of DeFi ecosystems. I deliberately focused on Ethereum based projects in this article, because it is the original root of DeFi. Every blockchain ecosystem (Solana, Cosmos, Polkadot, Avalanche, Optimism, Arbitrum, Base, and many more) has its own landscape of DeFi applications — which makes it very hard to keep up with the developments.

Most mature DeFi projects now cover more than one blockchain. But there also appear new DeFi apps on freshly launched blockchains. Many of these are clones of older Ethereum projects with similar traits. Often, they bring some new element to the table which is great for the speed of innovation. No DeFi project can lean back — there is fierce competition to offer the best financial blockchain services.

Spreading DeFi across different blockchains brings some additional challenges:

• It is hard to compose DeFi services across blockchains.
  DeFi plug and play works best on one chain.
• Users need to use so-called bridges to move assets between blockchains. These bridges have an increased attack surface — we saw many bridge hacks in the past. Complexity rises even more.
• Some technical mechanisms like flash loans don’t work cross-chain.
• Deep liquidity is very important to make financial markets work efficiently. When DeFi protocols span several blockchains, liquidity gets fragmented between these blockchains.

Summary

To summarise: We defined the usual scope of DeFi and described foundational building blocks in typical blockchain ecosystems like Ethereum.

We then analysed different DeFi projects and clustered them into use cases like:

• Borrowing and Lending
• DEXes, Automated Market Makers and Liquidity Mining
• Yield Farming
• Vaults
• Aggregators
• Advanced DeFi Use Cases
• Liquid Staking and Re-Staking

With this article, I wanted to show that DeFi is a fascinating field of innovation that rightfully catches more and more attention. With its five years of age, it is still in its early days. Yet, it already handles billions of dollars. It also survived many critical events like the Terra/Luna and FTX frauds and scandals in 2022 which involved dramatic price drops. For the most part, surrounding DeFi services worked just as intended: Prices of collateral fell by 50%, liquidity providers got liquidated, and everything came back to normal. If you witnessed situations like these first-hand, it is very hard to imagine how these innovations might not succeed on a broader scale.

DeFi concepts are simply the next stage of digitalisation and automation of financial services.

They are here to stay.

Further reading: You can find more articles about Blockchain and DeFi on my blog posts on Medium.

Disclaimer: This article is not intended to be an investment advice of any sort. Do your own research and search for professional support if you intend to invest in one of the projects mentioned in this article.

You are also welcome to follow me on Twitter or get in touch via LinkedIn (and please tell me your reason to connect and how you found me).