TL;DR: AMMs have played a huge role in driving DeFi and democratizing liquidity provision but they can come with heavy limitations. A new wave of innovation will be needed for mainstream users to better manage their risks and continue participating in this revolution.
For most, the first point of entry into the crypto world was likely through a centralized exchange (CEX) such as Coinbase or Binance. It’s easy to understand their popularity given their sufficient liquidity, ease of use, and the familiarity of the classical order book mechanism utilized.
However, relying entirely on CEXs in a decentralisation-first blockchain/crypto world always seemed a little odd, especially since your funds are held in their custody, and the assets available for trading are subject to their discretion.
Thankfully, decentralized exchanges (DEXs) built on Ethereum, such as 0x and DDEX, then emerged, with non-custody of funds and smart contracts that could be scrutinized. However, these DEXs still employed the same order book model as CEXs, and along with poorer UX/UI, often suffered from a severe lack of liquidity.
In a Web3 world where DeFi is the engine that democratizes financial services, to still be relying on mechanisms from the “traditional world” was very much fitting a square peg in a round hole. A fundamental rethinking of exchanges and liquidity was needed.
Led by Uniswap, Automated Market Makers (AMMs) were introduced into the crypto world and soon became the poster child for DEXs and liquidity in DeFi. In essence, instead of utilizing order books, an AMM exchange pools together liquidity provided by others and makes market according to a deterministic algorithm. Traders gain access to the increased liquidity provided, while regular users can reap benefits in the form of trading fees simply by providing liquidity of their idle assets.
Combined with the incentives provided through liquidity mining, AMMs saw explosive growth as liquidity could now be “crowdsourced” and market making has been democratized. In September 2020 alone, Uniswap’s trading volume was $15.4 billion, almost $2 billion higher than that of Coinbase. Accounting for over 93% of the market share occupied by DEXs, the AMM model made a strong case for being the innovation to bring decentralized trading to the masses.
AMM isn’t exactly a perfect solution and does come with several limitations such as low fund utilization, additional risk exposure, and also the widely discussed issue of impermanent loss. Liquidity providers (LPs) have realized that market making isn’t risk-free and that impermanent loss can occur, resulting in “losses” as their assets could hold higher value simply by sitting in their wallets instead.
The core of DeFi has always been the simple yet ruthless quest for efficiency and yield, for capital to never be idle, but instead constantly utilized, be it through providing loans on decentralized lending platforms or providing liquidity on-chain.
While AMMs by themselves have helped democratize liquidity provision, the current model still requires further innovation for it to align more closely to the ethos of DeFi.
Further readings on concepts mentionedWhat Is an Automated Market Maker (AMM)?Liquidity Mining: A User-Centric Token Distribution StrategyBeginner’s Guide to (Getting Rekt by) Impermanent Loss
With rapid innovation in the DeFi space proving to be the rule, not the exception, alternatives to the AMM model soon appeared, where the downsides of liquidity provision could be better mitigated. For instance, projects such as DODO, CoFiX, and Bancor are now offering new, innovative solutions to tackle existing problems.
One of the major issues with AMM is the way in which provided liquidity is allocated. For instance, Uniswap allocates funds uniformly across the entire price range, meaning that only funds which are allocated near the market price can actually be effectively utilized, while the rest sits idle. This gives rise to high slippage and high capital inefficiency.
In this case, market maker algorithms have to become “smarter” and act more like human market makers in CEXs who are able to continuously adjust their buy and sell orders according to the changing market price.
DODO serves as an example of this with their next-generation solution to liquidity provision in the form of their Proactive Market Maker (PMM) algorithm. Distinctly different from the AMM model, PMM leverages price oracles and seeks to emulate human market making by aggregating liquidity near the market price.
With the flatter PMM curve, traders can expect to benefit from lower slippage. And even as the market price shifts, the PMM will proactively shift the price curve accordingly to ensure that sufficient liquidity is still available. This ensures that fund utilization rate is kept high, providing slippage on par with Uniswap, even with just a tenth of liquidity volume.
In the DeFi world where liquidity is oil, and everyone is competing for LPs’ assets through all sorts of incentives, a defining advantage of liquidity provision solutions would be how capital efficient they are. The models that will be successful in the long-run, even after the incentives from liquidity mining are over, will be those that can provide competitive liquidity and low slippage rates to traders even without as much assets locked.
Another barrier that mass retail users face in liquidity provision is the fact that they have to take on additional risk exposure (which they are often not willing to) since AMMs such as Uniswap require the supply of two assets in equal ratio.
For example, if a user happens to have $100 worth of LINK tokens sitting idle in their wallet and is considering depositing liquidity into Uniswap to earn a portion of the trading fees accrued, they would also have to deposit $100 worth of another asset such as ETH. This results in additional risk exposure as the value of the user’s holdings is now impacted by the fluctuations of ETH price as well.
However, alternatives to AMM are now built such that users can simply deposit a single asset, thus limiting to single risk exposure. This can be achieved in several ways.
With DODO’s PMM, two separate pools are utilized for a single trading pair of assets for instance LINK-USDC, resulting in two parts of the price curve, the bidding side and the asking side, which are determined by the respective amount of tokens in each pool. This is as opposed to a single larger 50/50 pool in Uniswap’s AMM model. Hence, LPs need only deposit the asset that they are comfortable taking on the risk for, as opposed to buying an unwanted asset, just for the purpose of market making.
Other AMM alternatives that solve for this include CoFiX, where you can also deposit any amount of a single asset. However, that alone can be risky as LPs could lose value due to ongoing changes in the liquidity pool ratio, which is why a hedging program is encouraged. By mirroring the proportion changes in the liquidity pool in another market, users will effectively be able to hedge their position, thus balancing out the negative fluctuations, and locking in their profits from market making.
While the mechanism of balancing pools by requiring users to deposit two assets has thus far achieved the intended outcome, this will not be ideal going forward as more retail users come into the space. Thus, allowing for just single asset risk exposure will become a basic requirement for most liquidity provision solutions.
The concept of impermanent loss is one that many LPs learn for the first time when they realize that they could have actually made more money simply by holding onto their assets, as opposed to providing liquidity into an AMM. While there are already many pieces that have elaborated extensively on this topic, it essentially boils down to the fact that simple AMMs are simply too simple by design.
An AMM such as Uniswap follows the Constant Product Market Maker (CPMM) model, relying on the equation of x*y=k to calculate asset prices and rebalance proportions in the liquidity pool. This is because Uniswap by itself has no way of knowing what the true equilibrium market price of an asset is, hence it requires arbitrageurs to swoop in to buy assets at a discount when the price on Uniswap differs from that of the markets outside, rebalancing the pools in the process. But for LPs, this means that they are essentially paying arbitrageurs to facilitate the price discovery and verification process since AMMs are unable to do so by themselves.
There are a couple of solutions to this however, and they often include the use of decentralized oracles, in order to minimize the amount of impermanent loss incurred by LPs.
For example, with CoFiX, impermanent loss is theoretically gone since there is nothing to arbitrage away to begin with. Instead, CoFiX utilizes Nest Protocol as the oracle to provide market equilibrium prices for its assets, while adding a slight premium on top of the market price called, thus pricing in computable risks in each trade, and compensating LPs for the risks they take. However, this solution is only effective when done in conjunction with the hedging program mentioned above, which will allow LPs to lock in their profits gained.
With DODO’s PMM, Chainlink is used to provide decentralized price feeds, substituting the use of CPMM’s formula for price discovery. However, there is still the issue of pool proportions changing, and LPs not being able to withdraw the same amount of tokens that they deposited. Hence, in this case, DODO encourages arbitrage trading through their PMM which adjusts the market price slightly to make it profitable enough for arbitrageurs to come in and rebalance back the pool.
Another alternative to mitigate impermanent loss is the approach taken by Bancor, which essentially introduces insurance to ensure that LPs can at least receive back 100% of their initial capital, as well as any fees accrued. This is facilitated through BNT’s elastic supply that allows for the minting of new BNTs tokens as needed, and is dependent on long-term collective gains from market making being greater than the impermanent loss suffered by the pool. Details on this mechanism can be found here.
The one key takeaway from these various experiments: while liquidity provision has been democratized, market making is certainly not for everyone.
There is no free lunch but there are complex risks involved, so individual LPs will have to decide for themselves if they are willing to take on the risk to begin with. What we may see happening as liquidity mining rewards begin to decline is that retail LPs will no longer be as incentivized to contribute liquidity, thus having an impact on the liquidity and performance of DEXs.
There are several ways in which this may play out in the future.
The first would be the increased presence of professional market makers on DEXs, who would be willing to bring their liquidity over to DeFi as profitable opportunities clearly exist. For this to happen, there needs to be more freedom available to them, be it in forming their own parameters or in their own pools.
The second would be allowing users to manage their risks in more innovative and effective (but ideally also simpler) ways. This is a core theme in DeFi and will naturally be crucial for DEXs. The goal should be to maintain the lower barriers to entry that LPs currently have while also reducing unnecessary, complicated risks, through methods such as insurance, risk tokenization, or even just a suite of supporting tools.
From this perspective, an optimal solution may be the one that DODO is currently working on, where professional market makers are increasingly onboarded to DEXs, by allowing them to customize their on-chain market making strategies, thus incentivizing them to move their liquidity over. In the future, retail LPs could then follow these same strategies to further reduce the complexity of liquidity provision.
Such a scenario will result in traders being able to still enjoy the same depth of liquidity as that of CEXs, while also benefitting from the inherent advantages of DEXs such as non-custody of funds and permissionless onboarding of new assets. At the same time, individual LPs can still continue to provide liquidity and share in the success of the platforms that they are a part of, as long as they are willing to understand and take on the risks, though ideally there will be innovations in place to help manage said risks.
The world of DeFi is wild yet imaginative, an entirely different playground where innovations are given ample space to thrive (or fail spectacularly). The current growth that we’ve seen with DEXs and AMMs is one part of it, and will undoubtedly continue to play a crucial role in driving DeFi’s mission of democratizing financial access and services. But there’s a lot more to come.
So strap up, the journey has just begun.
If you would like to learn more about how DODO is driving a revolution in on-chain liquidity, their Medium is a great place to start.
Disclosure: LongHash Ventures is an investor in DODO.