Discussion: Liquidity Mining Impacts

I know there was also talks of adding different vault types for existing collateral as well. Example: ETH-B with a collateralization ratio of 135%, but higher stability fee and liquidation penalty. As it stands, Compound has a minimum collaterization ratio of 135%. Has there been an development on new vault types? I also know there was talks about implementing a new on-chain credit scoring method

@RepoTactics We might want to wait for the new liquidation system before doing that.

And it’s hard to compete with Compound on that front. With their liquidity mining.

1 Like

They mainly use the imagination of COMP tokens to attract users, not just interest rates.

BAT is currently earning ~23% APY (~100%+ APY including COMP) on Compound. Clever yield farmers realized that it is easy to max out utilization on small cap coins, which yields more COMP than simply borrowing/depositing USDT. Similar dynamic is happening with ZRX (11% base APY). If these conditions persist it will be uneconomical to maintain BAT or ZRX vaults.

Dharma has submitted a proposal to decouple Compound’s DAI interest rate model from Maker’s interest rates. If this passes in ~2 weeks, the likely result will be Compound’s DAI deposit and borrowing rate trending higher than Maker’s rates. This could potentially incentivize minting DAI to deposit to Compound.


They choose the interest rate to decouple from MAKER, which is a correct choice.

Because our interest rate has seriously deviated from the market average interest rate.

We do not respect the market, but choose to walk alone.

You can’t talk about market rates with compound. They’re subsidizing all their rates.

Yes, what I want to say is the combined average interest rate of the Defi and Cefi loan markets before compound liquidity mining COMP. At that time, we have deviated.
Compound’s liquidity mining COMP model is not suitable for us.
Compound has a brave heart.


Do we think this is sustainable long term for compound? How long can they realistically keep up negative effective rates? When does the music stop? Surely we need to do something if we think this will continue for an extended period of time, but not necessarily if we think this environment will change rapidly.

Perhaps we can add the technical capacity for negative rates, competing with borrow rates at least vs compound. After we have switched to the new more robust auction system we can probably lower collateralization ratios across the board as well.

Preliminary thinking:
We use the interest of daily income to repurchase MKR from the secondary market, and then distribute MKR to lenders in proportion. Achieve a balance between income and expenditure, and can form a virtuous circle.

As for how long this lasts, I believe there are a few factors at play. The first factor being the strength of positive reflexivity in this market. As it’s currently playing out, I see this as some 2017 stuff at play. If Bitcoin’s volatility remains low, I could see COMP staying at elevated prices and encouraging borrowing across supported assets for some time (remember float is low). As it stands, 57.71% has been distributed to team members, founders, investors, and partners. I have yet to find if there is a lockup period for this group. 42.29% will be distributed over the next 4 years. If there is no lockup period for the early founders group, the potential for a massive price decline is present. If there is a lockup period, I see more stability with the price as there is a more predictable rate of inflation. My final point is that as new incentives become increasingly common (Synthetix, Ren, Balancer), liquidity will flow back and forth between these protocols. This could lead to volatile interest rates between the protocols. In this case, I believe Compound may drain its customer acquisition budget (42.29% of the protocols value) by arbitrageurs.

The thing that concerns me the most is the size of loans against less liquid assets such as BAT. As it currently stands, there is ~$129,000,000 in BAT loans on the platform (about 36% of all BAT outstanding :face_with_raised_eyebrow:). I sincerely hope their liquidation engine is able to perform in the event that market volatility across the globe increases (109,000 BTC Options Expire by Friday & concerns about a second wave of virus outbreaks causing market turbulence in active markets)

I’m grateful @monet-supply pointed out the proposal submitted by Dharma. If that passes, we benefit even more so. Coumpound is borrowing from its future value to incentivize supplying and borrowing. DAI is one of the assets that is included in this incentivized program, yet we are not diluting our MKR supply. We are already seeing the benefit (I believe about 8 million DAI has been minted since the program launched). As such, I don’t see a need to introduce negative effective interest rates. However, I am but one man with one opinion. I encourage anyone with a different opinion to chime in!



DAI from BAT has declined from over 500k to ~125k. This is understandable due to the high marginal returns for transferring the position to Compound.

Currently Compound market info:

COMP price = 208
COMP speed = 1440 per day for borrowers and lenders
BAT supply rate = ~25%
DAI borrow rate = ~1.5%
total daily interest = $215,000

COMP rebate percentage = COMP price * COMP speed / total daily interest
~140% = ~$208 * 1440 / ~$215,000

Take an example vault with tab = 1000 DAI and LR = 300%. Maker’s DAI borrow rate = 0%, so the total position return/cost on $2,000 capital ($3,000 BAT - $1,000 DAI) is 0%. If you switch it over to Compound, the effective returns jump:

Total return = BAT return - DAI cost / (BAT amount - DAI amount)
BAT return = BAT amount * BAT supply rate * (1 + COMP rebate percentage)
DAI cost = DAI amount * DAI borrow rate * (1 - COMP rebate percentage)

Total return = ($3,000 * 25% * (1 + 140%)) - ($1,000 * 1.5% * (1 - 140%)) / $2,000
Total return = ~90%

Similar dynamics will apply to ZRX if/when it gets onboarded. If Compound allows WBTC to be used as collateral, we may also see strong incentives for those assets to move.


you are right. makerdao have to do something to compound Liquidity mine, or more and more assets will move to compound. and dai supply will be less, Dai liquidity also will be bad, finally makerdao’s market share will be smaller and smaller.

1 Like

The current 2 biggest liquidity mining distribution is Compound and Balancer.

Compound is currently distributing 20,160 COMP (~$4.8m) a week and Balancer 145k BAL ($2.5m) a week.

This is attracting a ton of liquidity into these protocols. And ctokens and Balancer liquidity tokens could make for some great collateral on Maker.

Would need some engineering work so that users could still earn COMP and BAL on their collateral.

ctokens have exposure to compound protocol risk. They can also be used as collateral to borrow stuff on compound with rebates. But users might trust maker more.

Balancer liquidity tokens are also interesting. But the oracle team might have trouble with pricing them.

These are both new liquidity incentives, they’re still evolving. But it’s still a good idea to keep an eye on them, possibly make plans. They’re attracting a lot of liquidity that could be used as collateral on Maker. The COMP distribution is slated to last 4 years and the Balancer distribution 8.6 years.

ctokens and Balancer liquidty tokens were also interesting options for collateral before the liquidity incentives, they’re just a bit more interesting now.

1 Like

I just became a Compound user and am experiencing Compound.

This is happening, some collateral has flowed to the compound, we may need to continue to observe…

I am also became compound user. more and more user are moving their asset to compound or balancer.
but if makerdao also design a liquidity mining program. user could mine by making Dai, then use dai to mine through compound again. that’s will be a good way to expand Dai.

When the same people are taking out loans against the same collateral like this iteratively the growth is illusory. I think the greatest benefit could be more widespread distribution of governance tokens, but we’ve already seen that at compound this just means more optimization for comp distribution. It all kind of amounts to a sugar high and ultimately the value has to come from somewhere.

So where does that value come from? Control over the assets in compound? It is growing, but if borrowers are getting negative rates and lenders are also getting paid then even the assets locked in compound aren’t worth anything per se, they are losing cashflow every day rather than producing cashflow. This is unlikely to change also, since comp holders are incentivized to vote for continued payouts over rational policy.

It seems like the only place this value can really be derived from is the comp token itself, causing the token to reduce in price as more people cash out. Maybe we’ve already seen that begin. But trying to emulate something like this would at least be a disaster for mkr holders, and might ruin the entire risk balance for the platform.


Compound is currently voting on an update to the distribution calculation for COMP. This change would make COMP distribution proportional to total $ amount borrowed, instead of the current method of allocating based on interest paid. Discussion can be found here.

So far, it looks like this update is likely to pass almost unanimously. This will greatly increase the allocation of COMP directed to the DAI market.

Assuming 2% DAI borrowing rate on Compound, the total cost of borrowing is ~$55 per $1M borrowed per day. The chart above indicates distribution of 3.4886 COMP per $1M borrowed per day, which is equivalent to over $800 in rebates per day at the current COMP price of ~$240. Netting these figures against each other, $1M DAI borrowed on Compound yields ~$750 in gains per day, for an annualized APY of around 30%.

The proposal has around 2 days left for voting, and then a 2 day time-lock before it can be executed.


Yeah, incentives will be quite strong on the supply side too as I can see the DAI utilization rate being high, especially compared to other stablecoins.

There is a question of how much DAI will be locked up on compound. If the utilization rate is very high, compound will have less DAI locked.

There might be increased demand for USDC-A.

If I understand correctly there will be an incentive to mint DAI through maker to deposit in compound? They will be basically operating like our DSR correct?


We should focus on observing the growth of compound token users, the growth of collateral assets, the growth of loans, the growth of collateral of a single token type and how they balance revenue and expenditure, not the price of COMP tokens.

If the compound can maintain the double growth of users and mortgages for a long time, the continuous increase in quantity may lead to a leap in quality.