DeFi is changing how finance is working in many ways. One of the most interesting ways is the ability to power the financial market with liquidity pools. Collective liquidity or monetary pools are not DeFi's creation, but in traditional finance, pools like these often mean under-the-table operations and non-transparency. However, with blockchain technology, everything is coded in smart contracts and open to the public. Pools can be as transparent as they need to be, and they are a cost-efficient and effective way to boost liquidity.
There are different kinds of pools elaborately programmed for liquidity in DeFi in various scenarios, and their application is ever-growing. Lending pools, market-making pools and options collateral pools are among the best known.
TLTR, the table below covers the information described in this article.
Lending pools are the first decentralized liquidity pools in DeFi, designed to boost the borrowing and lending market. Unlike the peer-to-peer lending model --- where borrowers obtain loans directly from the lenders --- lending pools act as the counterparty directly and interest is set by coded algorithms automatically. Projects such as Compound and Aave are some of the best-known lending pools.
Compound is an open-source money market protocol that lets users lend or borrow assets against collateral. Users are contributing their assets to a large pool of liquidity (a "market") that is available for other users to borrow, and they share in the interest that borrowers pay back to the pool.
When supplying assets, users receive cTokens from Compound in exchange. As interest accrues to the assets supplied, cTokens are redeemable at an exchange rate (relative to the underlying asset) that constantly increases over time, based on the rate of interest earned by the underlying asset.
In Compound's lending model, the interest isn't distributed. Instead, simply by holding cTokens, users are earning interest automatically.
Aave is a decentralized non-custodial money market protocol where users can participate as depositors or borrowers. Depositors provide liquidity to the market to earn a passive income, while borrowers are able to borrow in an overcollateralized (perpetually) or undercollateralized (one-block liquidity) fashion.
Users can deposit assets into the Aave reserve pool of funds and earn interest. Users can also borrow some assets from the pool, provided that they have enough collateral.
All deposits in Aave Protocol have a corresponding "aToken," an Aave interest-bearing token pegged 1:1 to the value of the underlying asset.
Lending pools are quite straightforward. By pooling the liquidity of the lenders together, the borrowers are borrowing from the pool, and interest paid is shared among the pool contributors. These cTokens from Compound or aTokens from Aave are like the pool shares. Interest rate mechanisms may differ among various platforms in smart contracts, but the basics are similar. The more demands on the market, the higher the rate will be.
In addition, mining incentives are allocated to liquidity providers to strengthen the yield as in
Due to the fact that the borrowing activities are always fully collateralized, lending pools are claimed to be non-loss pools. (We assume that the codes always work and leave exploits out of this discussion)
In summary, the lending pool collects liquidity with lower risk preferences, and the rewards are not high compared with other pools. As the interoperable nature of DeFi allows, we have witnessed many other projects with yield chasing strategies to regard these lending protocols as the base layer of the lego, and build on top.
AMM stands for an automated market maker, which is a type of decentralized exchange that relies on mathematical formulas to price assets. Instead of using order books like in traditional finance, a pricing algorithm coded in the smart contracts prices that assets. Liquidity pools are then created for each token's trading paid.
Liquidity pools provide the liquidity needed for trading, in an automated fashion. In simplest terms, traders no longer need to find someone else in order to sell or buy their coins. The transactions fees are automatically distributed amongst all liquidity providers, according to their shares.
AMM pools provide liquidity for token transactions, and the token compositions in the pools are changing with each transaction. Transaction fees are charged on each trade and distributed to the pool participants.
Different mechanisms are applied among these platforms. One of the best-known is the XYK model, which is used by Uniswap and SushiSwap. Curve is designed for stablecoin and homogeneous token transactions, with a low slippage algorithm. Balancer allows for multiple coins in a pool while customizing token composition ratio and transaction fees. Bancor v2.1 introduces single-sided exposure and impermanent loss protection to AMM pools. Kyber Network builds liquidity with an open reserve architecture. DODO applies a proactive market maker algorithm that allows liquidity providers to deposit only one kind of token in the pool.
Due to the differences in the algorithms in these platforms, the financial performance in these pools may vary much. Here in this analysis, we will mostly concentrate on the XYK model.
The gains in the pools are mostly transaction fees and mining incentives. For example, Uniswap collects a 0.3% fee and distributes pro-rata to all LPs in the pool at the moment of the trade.
However, there is a hidden risk that is becoming more widely known, impermanent loss (IL).
IL is the loss suffered by the liquidity providers in an AMM liquidity poo. Impermanent loss happens when you provide liquidity to a liquidity pool, and the price of your deposited assets changes compared to when you deposited them. The bigger this change is, the more you are exposed to impermanent loss.
In Uniswap and SushiSwap, both of which apply the XYK model with 50/50 deposit in token pairs, the IL risks can be quite high. While in Curve, with stable coin transactions and little price slippage, IL risks are low. Other projects like Bancor, DODO provide different solutions to this risk.
Uniswap has been sitting on the throne of DEXes for quite some time, the IL is not neglectable, and it is important to how the financials will perform after jumping in the pool.
Suppose one is holding two assets X and Y with equal value, where Y is a stable asset. He/she intends to contribute X and Y into the 50/50 AMM liquidity pool. The black straight line below is the total value of holding X and Y when the price of X changes. The grey curve is that financial performance when the user contributes the equal value of X and Y into the liquidity pool without any transaction fee rewards. You can see that he/she will always be better off holding the assets than in the pool.
With some transaction gains as compensations, the grey curve will shift up to the red one. But he/she may still suffer and end up in a more unfavorable condition than just holding and doing nothing if X's price moves further away from the original input point.
*(Supposing the transaction fee provides a 5% return, and the protocol provides a 30% incentive in X mining) *
The chart below is a simulation of joining in the ETH/USDC pool on Uniswap on Nov. 27, 2019 and staying for a year. Details of the simulation above can be found here.
Therefore, providing liquidity in Uniswap and SuShiswap can be risky for volatile assets. The contributors do not always end up with profit. Especially when a volatile token price collapses, the AMM algorithm can even eat up the majority of the value of the stable assets one contributes to the pool. Mining incentives can be helpful as they can push the curve up a lot (in blue), provided with a 30% return in token X in the same period as in the blue line of the example above. But still, it is an unstable and short-term plan for most projects.
In summary, the risks in AMM pools vary much in different models. The IL risks could be significant in the classic XYK model in Uniswap and SushiSwap.
But for stablecoin trading pools, risks are much lower. Some others are innovating with other algorithms to mitigate IL. Many projects are creating pools on these DEXs to boost their transaction liquidity, and rewarding liquidity providers with high mining incentives as compensation. At present, the mining compensation is crucial to contribute to the major part of the gains for the participants.
An option is a binding contract that allows you (as the buyer) to sell or buy an underlying asset (goods, stocks, indexes, etc.) at a predetermined price within a set time frame. You, as the buyer of an option contract, have the right, but not the obligation, to buy or sell the underlying asset.
With options, in order to get the right to buy or sell a particular asset at a predetermined price, one has to pay the option seller a price, which is called the option premium.
The option buyer pays the option premium and is entitled to rights only. The option seller receives the premium as consideration for giving up those rights. Therefore, it is usually required that the option seller needs to deposit collateral for performance of the contract.
The potential risks exposed by options sellers are theoretically unlimited. In traditional finance, the sellers are usually professional institutional players, with sophisticated hedging tools to mitigate the risks. Also, there are professional dealers and market makers to provide order-book liquidity for different option trading pairs.
However, if we make these option contracts decentralized on-chain, there may be some problems. A lack of professional decentralized hedging instruments will keep the sellers away, as on-chain order-books can be costly and have low efficiency.
Decentralized options with liquidity pools as collateral are great innovations in DeFi --- rarely seen in traditional finance --- with Hegic and FinNexus as representative projects.
As in traditional finance, options are normally peer-to-peer bilateral contracts. But Hegic and FinNexus move the option contracts on-chain with peer-to-pool models. Liquidity is collected together in the collateral pools. The pools are the counterparties to all the options with different terms, while providing collaterals to them. Risks and premiums are shared equally across the entire group of liquidity providers so that no individual participant is at high risk and all participants can share in the rewards.
For example, the USDC liquidity pool in FinNexus operates as follows:
Measuring the financial performance in the collateral pool is not as straight-forward as for the other liquidity pools. Options are playing with odds, and the well-known BS options pricing model is based on the theory of probabilities.
Based on the mathematical statistics, the writers are more likely to win in general if the options are properly priced, but in real application scenarios, things could be more complicated. The pool is the writer of all options with different terms: it is a mutual fund for writing options. It could take months or even years to prove the reliability and stability of the returns in the options collateral pool --- but there are still ways to run some tests on what the financial picture will look like in the pool.
It is widely accepted that cryptoassets like Bitcoin (BTC) and Ethereum (ETH) are extremely volatile. Therefore, it would be reasonable to comprehend that options could be more friendly and beneficial to the holders rather than the sellers, as they only have rights without obligations, and the cost is capped to the premium.
Is this true?
Before we jump into the mathematics, it is important to understand that volatility is an important parameter for pricing options. Options will be more expensive in more volatile assets. It is often known as implied volatility in pricing options in the BS model. Here in the following analysis, we use the historical average volatility to price options.
First, for BTC options, we calculate the historical average volatility in 2.5 years between Mar. 29, 2017 and Nov. 26, 2019, including a volatility of 1 day, 2 days, 3 days, 7 days, 15 days and 30 days.
Second, we calculate the price of ATM options with each expiration date with the volatility above, in the past 12 months, from Nov. 27, 2019 to Nov. 27, 2020.
Third, suppose the same amount of puts and calls is written by the pool, we can derive the P&L for the sellers in each expiry in the same period.
Last, we allocate different weights to the options in different expiry, referring to the transaction volume on Deribit. We can derive the pool's average expected APR and the maximum drawdown.
The details of calculation can be found here.
The pool as the collective option seller will be more likely to win in the long run, despite the occasional drawdowns. In the past 12 months, the pool's net value increased by over 60%, which means a 60% APR in return. Of course, this measurement is using the hypothesis that the pool as collateral is fully utilized. For an average 50% utilization ratio, the expected average APR will be 30%, also with smaller drawdowns.
From the chart above, in case of a relatively long time of unilateral price movement, the pool will suffer losses, especially from March to May 2020 and from October to November 2020 in the test period.
Similarly, with the same methodology, we can simulate the finance performance of a pool for ETH options in the recent 12 months as below. The details of simulation can be found here.
Collateral pools for options behave differently from better known lending and AMM liquidity pools. The pools provide liquidity and collateral for options, while rewarded with option premiums and undertaking risks. It is possible to lose money in cases of the unilateral market price movement of the underlying assets. But in the long run, it could bring some favorable rewards for the liquidity providers.
There are other pools --- like yield chasting vaults (protocols) such as YFI, YFII, Harvest, as well as staking pools for the purpose of voting, farming or boosting, like with CRV and Hegic. The returns in these pools can be measured in different base currencies and also be diversified with different risk-return profiles. Vaults claim to apply no-loss on-chain strategies, with constant farming and cashing mining rewards, and they are developing into more complicated systems as the more decentralized tools are integrated. Staking lots in some platforms can provide a boost in other mining activities and be rewarded in gains from transaction fees collected by the protocol.
Decentralized liquidity pools are among the greatest innovations in decentralized finance. They successfully gather enough liquidity to maximize efficiency and lower costs, while maintaining an open-source and censorship-resistant manner with blockchain technology.
Moreover, as DeFi projects continue to provide users with more diversified pooled liquidity models, participants with different risk preferences are getting more and more choices. Of course, it is highly recommended that you always do your own research before depositing money into pools, and never put in anything that you cannot afford to lose.
(This article was originally published on CoinMarketCap on December 18, 2020)