author: Manny Reimi (@mannyreimi )
- Liquidity Mining (LM) is the best way to align incentives between investors and users while launching and boostrapping a new DeFi protocol.
- LM rewards drive adoption and retention, but LM programs must be designed carefully in order to avoid failed launches and other problems down the road.
- When designing LM programs, tokenomics and value-capture mechanisms play an important role in resisting short-term shocks.
- There is plenty of data regarding best practices in token allocations and speed of distributions in DeFi nowadays to craft a competitive LM program.
- LM programs with notable modifications have encountered important success recently and important lessons can be drawn in LM design from these very novel examples.
What is Liquidity Mining?
Liquidity Mining (LM) programs are an effective growth hack and marketing tool for blockchain dApps looking to make quick inroads in product adoption and ecosystem development.
In the recent past (2015-2020), nascent blockchain-based applications would mint a new token and sell it to wiling investors, either privately or publicly, and sometimes in a major event termed an Initial Coin Offering (ICO). This trend reached its apex by 2017 ($2.6B raised in ICOs), and increasingly became more scrutinized by regulators in several jurisdictions.
By contrast, in an LM, a protocol would distribute tokens in exchange for liquidity or market-making to bootstrap the price of the new token as well as provide incentives for its effective use. Usually one or more of the following conditions are present:
- the protocol running the LM has a product (usually a dApp) where the token can be readily-used, either for its utility, to vote in altering some protocol parameters (i.e. a governance token), or a combination of both.
- the LM takes advantage of permissionless, decentralized exchanges – e.g. Uniswap , Curve , or Balancer – to facilitate the participation of non-sophisticated but enthusiastic potential users in the provision of liquidity.
- the LM operates with a schedule of emissions, which can be either front-loaded, linear, or in tranches. This is so that LM participants have a clear expectations and build trust from the onset.
Compound: the original Liquidity Mining program
Compound Finance launched its decentralized lending protocol in early 2019 to little fanfare except a small circle of mathematicians. The protocol hard-fought for users, touting its advantages (autonomous lending, earn interest in Crypto, no credit checks or KYC/AML), and roughly 18 months later had a little over $100M in deposits, a great victory, no doubt. Yet, Compound was still small in the DeFi market, dominated by the more mature Maker, which was at least an order of magnitude larger.
COMP codecode with an LM on June 2020 and DeFi had never looked back. Growing to $900M in deposits just two months afterwards, Compound now has $3B in total value locked (TVL).
What the Compound LM looks like:
COMPcodecode are distributed every day, for 4 years (still ongoing as of Jan 2021).
- 50% of
COMPcodecode go to lenders, 50% to borrowers.
COMPcodecode is given to markets where more interest if being accumulated. This is calculated daily.
COMP launched at ~$80 and quickly reached an all-time-high of $372 on the launch mania. After a severe correction to $87,
COMP has seen a new recovery cycle and currently trades at $240 codespan.
Given that during these four years only 42% of
COMP codecode will be distributed (the rest is to be kept by the team, investors, and advisors), one could say that the Compound LM, the first in history, is amongst the greatest growth hacks of all-time, turning Compound from an experimental lending protocol to a $2.4B DeFi unicorn in six months.
Tokenomics in the era of Liquidity Mining
Naturally, many protocols have tried to replicate the success of Compound, some with more success than others. At present, the market has a plethora of governance tokens, and several LM programs are announced weekly.
Notably, some protocols like Synthetix , which got started during the ICO days, remodelled their token models as they understood the power of the LM to encourage users to participate in its marketplace.
A few trends can be seen by teams in DeFi regarding tokenomics and LM design:
- Multi-year programs have seen more lasting success, particularly for novel markets that need time to mature, subsidizing participation/liquidity should not be done overnight.
- Inflation can be used to reward active over passive users, and many protocols have introduced perpetual stable inflation as part of their tokenomics. However, given crypto-users fixation with hard-supply caps (inherited from Bitcoin's marketing), this should be handled with care with the community.
- Even without front-loading, new token prices can fluctuate wildly, causing new users significant stress and putting a strain on community management resources. To avoid such scenarios, it is important to subsidize liquidity to the trading markets of the LM token itself.
In order to make a competitive LM in today's environment, a protocol must also consider the utility of the token itself. A few mechanisms have tried to address this:
- Treasuries & Community Development Funds are on-chain wallets, usually administered by a multi-sig elected by governance. Treasuries support the LM token price because they implicitly back tokenholders. However, their real utility comes in the ability to deploy funds to invest in growth (hiring engineers, designers, and product talent; rewarding contributors in the community; fostering the development of ecosystem projects and tools). Many protocols nowadays launch with a Treasury, usually a set-percentage (up to 10%) of protocol emissions will be directed towards Treasuries and Funds controlled by governance.
- Vote-Locking was developed by Curve
when it launched its
CRVgovernance token. In order to participate in governance,
CRVholders must lock their tokens in a voting escrow contract. To get their full-voting power, the tokens must be locked for 4 years, but options like 1 week and 1 month are also available, which give a pro-rata amount of
veCRV, the non-tradeable, on-chain representation of the vote-escrowed
CRVcodecode. To tie the system together, Curve also offers boosted rewards, or up to 2.5x the LM rewards to fully vote-locked holders. Boosts decay over time, and as more users enter the system getting a boost becomes more and more costly, encouraging people to lock early and for long. Curve also shares 50% of its trading fees with its governance-locked holders, compounding the incentives to lock, which takes tokens out of circulation and results in positive price-pressure.
- Revenue-Sharing, in the form of staking rewards, or together with more complex mechanisms like the above vote-locking, is a powerful incentive for users to partipate in an ecosystem. It is also, an important milestone, that a protocol has reached profitability and can offer some hard-currency (
wBTCcodecode) to its long-term holders. Some protocols like Maker and Aave go further and turn their holders (stakers, in the case of Aave) into implicit backers of the protocol (their tokens or stakes are slashed in price or amount if the protocol suffers a shortfall event). Notable, these are very mature protocols (and currently king and queen of DeFi).
Team, Investor, and Advisor Incentives
Besides not-funding a healthy treasury, a harder and more intractable problem to solve after an ill-thought, sudden, or botched LM is carving out a size of the pie for the founding team.
Ideas do not become reality by entrophy or intertia (quite the contrary, in fact), yet token-holders and community members will be very frustrated with any change to the token supply post-launch, particularly one made for the sole purpose of fairly compensating the original founders and backers.
A very heated and long discussion is currenty happening in Yearn , the yield farming OG, due to a lack of foresight during launch. The proposers of new tokenomics have produced this document , which shows most competitive teams in DeFi have launched with 20-30% reserve for team, investors, and advisors. Uniswap kept 40%. Certainly, the days of keeping 60% like Compound may be long gone.
Newer Trends in Liquidity Mining: the Good, the Bad, the Ugly
Yearn: the fair launch
enjoys a great asset above all others: its founder is superstar DeFi developers architect Andre Cronje. The native South African is to DeFi what Elon Musk is to electric cars. Unfortunately, Andre is known for his penchant for experiments in live conditions (I test in prod is his motto). Andre launched Yearn and was quickly overwhelmed with the user demand and loud response, however he had no treasury and no funding. He decided to launch only 30,000
YFI tokens in an LM so quick it barely made an impact, and soon the
YFI codecode mania started, except Andre had excluded himself from any YFI as he wanted a fair launch. Needless to say, except for the memes, this has backfired horribly. To this day, this poor tokenomic design harms Yearn's competitiveness and is an enormous source of frustration for a team that delivered a 10,000x token and many innovations in DeFi without being able to capture much of that value.
Don't be Yearn.
Uniswap: UNI-versal basic income
Uniswap has been a darling of DeFi users everywhere, with its humble origins, unassuming near-accidental founder Hayden Adams, and its amazing UX/UI. DeFi's first unicorn was one-upped by a fork SushiSwap with better tokenomics just as soon as it claimed the DeFi crown. In fact, Uniswap had no token, so ... zero tokenomics.
Not to be outdone, the launch of
UNI was a massive airdrop that put $1,200 (400
UNI codecode @ $3 each) in the hands of all the 250,000+ that ever used Uniswap, including those that only ever sent a failed transaction.
This, obviously, was great publicity, and was followed by a 2-month LM on some of the biggest pools in DeFi that cemented Uniswap's position as a premier DEX.
Short, sweet, and probably something that cannot be easily replicated by others. A great success, while it lasted.
As soon as the short LM ended, competitor SushiSwap has made significant inroads with its more persistant LM and revenue-sharing design, while Uniswap v3 has been shrouded in secrecy. In DeFi, more often than not, short and steady wins the race.
The algo-dollars: bootstrapping with hard money
After all, these zero- or partial-collateral tokens had never been tried before, and some of the mechanics (particularly for zero-collaterals) seemed funky. However, for two weeks or so, the LM schedule of these protocols offered solid rewards (some 5% of the total supply) for staking tokens of renowed value like
After all, it worked. The LM did, the protocols are still up in the air. Turns out, what seems like an exercise in futility for the protocol (there's no market created with the bootstrapping liquidity), is actually proof that people are willing to risk valuable assets in order to get their hands on something they see value in as well. Overall, a good lesson, and one that brings smart money early on in the LM program's progress.
1inch: the staggered approach
DEX aggregator has been running since mid-2019. With innovations like DEX limit orders, cheap-gas
CHI codecode tokens, and a trusted network of miners to avoid front-running, it's been a favourite of DeFi traders for quite sometime.
When 1inch decided to upgrade and launch its own automated market-maker LPs, in competition with Uniswap
, and others – it decided to go the
UNI-route and airdropped all its old users, which were of course, very happy. A small, smalll caveat, of course: the
1inch codecode airdrop represented only 6% of the total supply.
In total, 30% of the
1inch codecode supply will be unlocked for incentives, but most of the remaining 24% of expected to be in LM incentives. Now, however, it's already in the hands of governance. 1inch decided to empower its community on how to distribute the remaining rewards, and the community has followed up with a 1% tranche of LM rewards, currently ongoing. The token reached a capitalization of $1.5B, only since Christmas Day 2020, and now has a sizeable war chest to mount an incentives offensive against its more-established rivals.
StakeDAO: precision strikes
The most curious and probably best thought-of LM in recent memory has been the newly-launched StakeDAO
SDT codecode drop.
The newly-launched product arrived took 2021 by storm by airdropping a mere 1.5% of its token supply, and having had no users before, they did the best they could: selectively targeting the users they wanted.
Through a clever feat of engineering, StakeDAO sourced the addresses of DeFi's most active participants: stakers in novel projects like LIDO , Gitcoin donors, and addresses that voted in the governance of the main DeFi protocols, most of which use the Snapshot app.
How would StakeDAO know these addresses wouldn't just dump the token? Well, for one, its product is essentially an improve fork of Zapper , the favourite dashboard of DeFi users. Zapper has held off on launching a token for too long, so like Uniswap before it, a competitor popped up to fill that niche.
More importantly, StakeDao made it clear that their
SDT codecode tokens would be eligible for staking shortly after the airdrop was over, with triple-digit yields! Way to separate the weak from the diamond hands!