One of the most interesting recent developments in cryptocurrency is the emergence of decentralized liquidity pools.
Algorithmic-based smart contract liquidity pools such as Ethereum’s Uniswap, or privacy-focused, off-chain decentralized exchanges such as Starkware’s StarkDEX are just two examples of projects leading the charge.
Inbound/outbound liquidity is essential for the creation and growth of financial markets. Price discovery, and the ability to move in and out of trade positions, whether they’re from a big institutional firm, or a small-time trader, remains key if crypto is to reach maturity; where its aggregate daily volume could sustain at levels comparable to the legacy financial system.
It is not exactly a secret that the blockchain and cryptocurrency industries have a liquidity problem. Large trades in all but the most popular assets move the market to an alarming degree. This volatility then causes a cascade of ills.
First, it decreases the credibility of the markets due to the reality or appearance of manipulation.
Second, it makes people nervous about holding assets, meaning that applications dependent on low volatility have trouble getting off the ground.
Third, it harms the viability of decentralized exchanges and other decentralized token economies because insofar as they depend on slow mainnets, they lag badly behind the price information available on faster, more efficient centralized exchanges.
Decentralized payments are just one piece of the puzzle of what it really means to be decentralized, as you’ll also need the assistance of decentralized liquidity to build and extend additional functional financial layers on top of your blockchain-related protocol/application. Liquidity is king, and it can make or break your protocol if you cannot rally sufficient liquidity to aid in your project’s growth and enable the use cases you sought out to provide your end-users.
With the proliferation of decentralized lending, borrowing, and more, the current decentralized landscape appears to be grasping the basic essentials necessary for the financial instruments we’ve grown familiar with in traditional legacy markets (Compound Finance is but one example). To better understand where we’re at, let’s first go deeper into what solutions the industry has concocted thus far.
To begin, liquidity pools could help address a key problem faced by new token-based projects: the need to arduously bootstrap a liquidity-providing network before the project has real utility. Liquidity pools can mitigate this by providing a unique, less-speculative reason for people to hold tokens that do not have a large user base yet (i.e., to provide liquidity for a fee).
Moreover, the existence of decentralized liquidity pools provides added reassurance to large investors in young projects who do not want to get stuck trying to unload their tokens in an illiquid market. The pools thus function somewhat like insurance for token holders (we’ll cover this idea more below).
Second, liquidity pools should be considered an impressive achievement in decentralized institution-building. Liquidity has long been a central concern, not only for cryptocurrency and blockchain projects, but for financial markets in general. It is a prerequisite for the growth of a whole range of other institutions, financial and otherwise.
And, decentralized liquidity provisioning is emerging through a mechanism that does not exist in traditional financial markets — automated smart contracts. This is a totally new vector of provisioning liquidity, which opens up the possibility of broader, more competitive involvement in market-making. Liquidity pools are thus a bellwether of maturation for decentralized cryptocurrency markets.
The total quantity of liquidity in these decentralized pools remains small by the standards of conventional markets (which can trade daily with volumes exceeding hundreds of billions in USD), but it is growing at a fairly impressive pace.
It is also worth taking a closer look at a few of the leading liquidity pool providers. Their mechanics vary, and are not always terribly straightforward. However, they represent important opportunities for investors to analyze. And if they continue to grow, they could alter the calculus for large investors interested in cryptocurrency markets but concerned about liquidity risks.
Uniswap has emerged as a leader in the decentralized liquidity space. Their contracts are simply pools of 50% ETH and 50% some target asset. Traders buy either asset directly from the contract, causing the prices to move algorithmically. When differences emerge between the algorithmically-determined price offered by the contract and the market price, arbitrageurs close the gap. Anyone can replenish liquidity in the contracts by contributing an equal amount of ETH and the target asset. Doing so entitles them to a pro-rata share of the trading fees (0.3% per trade) that accumulate in the contract.
This piece provides an excellent starting framework for understanding the basic bet implied by supplying liquidity to a Uniswap contract. A deeper dive can be found here as well.
Bancor built the first meaningful decentralized liquidity solution. However, it has been losing ground because it suffers from several technical disadvantages, and is also dependent on its own token (making its solution less elegant than Uniswap’s architecture).
The biggest problem faced by liquidity suppliers to pools like Uniswap is the risk of major relative price movements between the paired assets; if the price of an asset in a trading pair surged suddenly, it could cause a ripple effect of negative counter-trades from a lack of proper liquidity. It is therefore ideal to supply liquidity in terms of a stable asset, instead of a volatile one like ETH. This problem is exacerbated by Bancor’s dependence on its native token, BNT, which is even less stable/liquid than ETH, while also adding the complexity of another abstracted token to maintain.
Moreover, transactions on Bancor are structured in such a way that they can incur high gas fees, and they are not presently planning to utilize layer 2 scaling technologies to alleviate those pains.
Bancor is working to address these issues instead by introducing a new stablecoin to replace BNT as the basis of its liquidity pools, as well as a few other upgrades. It remains to be seen whether the effort will succeed. It would appear though that an algorithmic approach such as Uniswap, paired with a tightly-coupled asset like ETH or another ETH-built stablecoin, is probably the best approach until further improvements emerge in the industry.
[Other projects like Kyber Network and the 0x Project focus on cross-chain liquidity and possess their own ERC20 asset — but they fall outside the scope of this post.]
Balancer is only a whitepaper for now. But it details a protocol that would allow people to easily instantiate new liquidity pools, backed by larger, more flexible sets of assets, with more precisely calibrated algorithmic incentives and user-determined transactions. If it takes off, it could encourage much broader participation in providing liquidity.
Described in a recent whitepaper, the Convexity Protocol may turn out to be an important accelerant of decentralized liquidity provision. By allowing anyone to easily write collateralized options contracts, and sell those contracts in the form of an ERC20 token (oTokens), it will allow more sophisticated forms of hedging to occur without an intermediating institution. While the Convexity Protocol could have a nearly infinite range of possible uses, one of the most obvious is as liquidity insurance. Would-be liquidity providers in new markets will have a bit less to fear when they can use relatively stable assets as their base pair, insuring against collapses in the liquidity of their target market.
Obviously, there is no free lunch. In a sense, it is possible to think of Convexity Protocol as “spreading” risk from riskier markets into more stable ones. Nonetheless, if the fundamentals of participation continue to improve, tools such as this could speed the arrival of meaningful decentralized liquidity in a wider range of assets.
Unipig and StarkDEX
It is important to notice the connection between liquidity and network capacity/throughput. The inability of the main Ethereum chain to rapidly settle high volumes of transactions is a fundamental impediment to liquidity providers, because the ability to quickly remove liquidity is a driver of willingness to provide liquidity in the first place. (For a deeper dive into the nuances of this dynamic, take a look at the CFTC’s report about the 2010 “Flash Crash”).
Therefore, one of the most important fronts in the battle to unlock decentralized liquidity is the development of Layer-2 and off-chain solutions for the rapid settlement of a high volume of trades. Two of the most interesting projects in this space are Unipig and StarkDEX. They both promise vastly increased network capacity and execution time, but take different routes to get there.
Unipig, currently live in demo form, allows transactions to be posted in real time and high volume to aggregators running fully functional Uniswap contracts, which are then “rolled up” and posted to the main chain. Parties’ trust in the veracity of these aggregators’ reporting is backed by a bond that dishonest aggregators stand to lose. This is a simple scaling solution whose success with real money at stake will depend upon effective auditing of aggregators. We suspect that the Unipig team will get the auditing and verification mechanisms right, but there is still some uncertainty about whether large institutional players will ever feel comfortable supplying liquidity through this channel. Still, we think their approach of scaling Uniswap via layer 2 technologies and optimistic roll-ups remains one of the most adept approaches yet seen; without the use of SNARKs/STARKs, more developers will be able to learn how to utilize their setup more quickly.
StarkDEX, on the other hand, uses state-of-the-art cryptographic STARK proofs to take on-chain transactions, process them off-chain, and then batch them back on-chain in order to increase throughput. The challenges of this method are purely technical, rather than social, as with Unipig (where you just have to get others to supply more liquidity over time). Running on testnet it appears to increase transaction volumes by more than 100x versus the main chain, with correspondingly decreased gas costs.
It is not obvious to us how the trade timing constraints it imposes will interact with the needs of major liquidity providers, or how quickly their solution will be adopted by other major players. That said, it is a very promising step towards orders-of-magnitude better throughput and major new opportunities for decentralized liquidity provision, and could likely play a pivotal role in the creation of scalable dark pools — with few or no trust assumptions
The Liquidity backbone
Plenty of parties have tried (and will continue to try) to “shortcut” the liquidity problem by providing liquidity from some concentrated or centralized venue. But this only highlights the deep connection between liquidity and the decentralization ethos itself.
Any financial system is, in a sense, only as decentralized as the sources of its liquidity. After all, if there are no central banks, but instead a handful of whales acting as central banks, what exactly has improved?
When provided from a wide range of parties whose behavior is deeply uncorrelated, liquidity is fundamentally more robust: it is less likely to evaporate in a crisis and more indicative of a healthy market.
Therefore the health of DeFi is largely identical to the health of decentralized liquidity venues. We are excited to see so many great teams attacking this key problem and striving to unlock a new phase of maturity and innovation in the space.