MakerDAO, a brief history and a troublesome future

An opinionated analysis on the past and future of MakerDAO which concern me.

Introduction

If there's a project that every DeFi fan is obsessed about, MakerDAO takes that spot any day. The project itself has humble origins starting in 2015/16 by Rune and many other Ethereum die-hard believers in the vision of creating a permission-less, decentralised stable coin on the Ethereum blockchain.

You have to remember, this was a time before Tether was on Ethereum, USDC/USDT/USDC were even around and the idea of a stable coin was even possible. Having any notion of USD inside smart contracts meant you had to rely on Oracalize (I think they've rebranded to something else now). Early 2018 when the officially launched it was groundbreaking in the community, my own DeFi startup (on-chain recurring payments protocol) was going to rely on it heavily as well since merchants could accept real US dollars rather than volatile Ether. Throughout 2018 the tech progressed although marketing was severely lacking and it took close to 8 steps to open up a CDP. Many people in the industry today may or may not remember the space however DAI was an obscure coin that you maybe had heard about as the 'edgy' thing at the time. 2019 the maker team changed that. They went very aggressive with marketing, promoting the project, ensuring 3rd party developers to build interfaces (InstaDApp was born out of this need as well) and revamping marketing materials overall.

The question on everyone's mind was whether an overcollateralised system would be able to retain the value of the synthetic asset even after the price tanked significantly. 2018-2019 proved this was possible as Ether crashed from its heights of $1000 all the way to below $100 over the course of a few months. People thought that the system worked and we had a winner. I would imagine that many VCs in the space such as a16z and Polychain had invested on a winner takes all thesis with Maker's MOAT being integrations and liquidity of DAI and of course - the supposed elegance of the system. Many more funds probably purchased in by doing proxy due diligence. However the cracks in Maker were starting to show only a few months after.


The Red Pill

Setting context: it's probably early-2019, MakerDAO has been doing well with a couple of hundred million locked up. The community is wondering what Maker's future roadmap is going to look like now that they've proved out p/m fit for their

core product. During this time we then see the following report come out:

https://www.coindesk.com/darkest-days-yet-purple-pill-tell-all-details-years-long-rift-at-heart-of-makerdao-stablecoin-project

In a 10,000 word story known as "Zandy's story", we learn about the political turmoil going on internally at MakerDAO where two ideological struggles are beginning to play out. One side is know as the red pill and the other is known as the purple pill. Unfortunately the story ends with most of Maker's core believers leaving the project as the "red pill" prevails. The idea around the red pill is to bridge the gap between legacy and traditional financial systems through introducing securities as collateral and leading Maker to be more dependent on the existing financial system. This was the beginning of what we know as multi-collateral DAI (~2 years ago). Based on various parameters and factors, BAT was the second collateral type added to MCD even though its on-chain liquidity is sub-par to say the least. If you have the time to read it, I’d highly recommend it.


Governance

MakerDAO's ideological pitch is a way for users of volatile assets to hedge against inflation. Turns out that's not really the use. It's actually for large ETH whales to go long on ETH and use the DAI to buy more ETH. In early 2019 when the price of ETH was super low, optimism around future price action was high and the stability fee was 0.5% - it led to one direct outcome: create CDPs as much as you can since the cost of borrowing/going long is super cheap. That's literally it. If you have 10,000 ETH worth $10m and you only need to pay $50k/year for betting on the price increasing by just 0.5% to be in the green, that's a no-brainer move. So of course, degen crypto people do what they do best, be degens and bet the house. They did this until people had generated so much DAI to buy ETH that there was too much DAI on secondary markets causing the price of DAI to dip below the actual threshold.

This was the point where MakerDAO governance actually came into play and became a larger talking point. Important note that a16z, Polychain and a few other funds had purchased tens of millions of dollars of MKR so their say has weight at the table. Going back, since DAI was deviating away from the peg, MakerDAO announced an executive vote to increase the stability fee. The idea around this was to increase prices which would force CDP holders to buy DAI off the market in order to close out CDPs (to avoid paying higher interest rates). Unfortunately the first increase to 4.5% wasn't enough. They had to repeat this a few times until the stability fee was over 15% from what I can vaguely remember. It was at this point we saw the rise of on-chain money markets such as Compound shine as DAI lender could earn over 10% for lending.

What must go up, must come down. DAI wasn't an exception to Murphy's law. Now that the stability fee had been increased and DAI was near its peg, it then started trading at a premium. More executive votes were held to change the stability fee to match secondary market demand. In a world where majority of MKR is held by Rune, a16z & Polychain, voter apathy is very real thing. The incentive for people to care isn't really there because any large whale can turn things around, and ensuring the fine tuning of the system isn't always in a large MKR whale's interest.


March 12

This is the day where we really saw MakerDAO as a system starting to break down. Global markets were facing a massive liquidity crunch and many longs were getting liquidated. All of this resulted in the price of ETH tanking from $200 to $80 over the course of 12 hours. As prices were dipping three key behaviours became observable:

  1. ETH holders selling into DAI to hedge against tanking prices

  2. Gas prices on the Ethereum network spiking

  3. CDPs being liquidated and DAI being purchased to ensure the system remains over-collateralised

Unfortunately, the combination of the 3 factors resulted in the following situation:

  1. MakerDAO liquidation bots weren't configured to dynamically adjust gas prices in the case of network congestion

  2. As a result, when CDPs were under-collateralised, they weren't any participants in the auction to purchase DAI to pay back the debt

  3. Due to low competition in the auctions, one lucky liquidator walked away with 30,000 ETH for $0.

The final outcome of all of the above a user lost 30,000 ETH and the Maker system is in debt for ~$4m. From here, the protocol had two choices. Trigger global settlement (effectively "shutting down" MakerDAO) or dilute MKR holders by minting more tokens and selling the proceeds to cover the debt. An executive vote was launched and the results were overwhelmingly towards diluting MKR holders rather than shutting down the system.

This was probably the right call to make to ensure the long term success of MakerDAO, although this isn't full picture. What comes next is still kind of crazy to see play out.


Enter USDC

In order for the MKR dilution auction to be a success, the MKR has to be sold for DAI - meaning DAI needs to have sufficient liquidity. Unfortunately, there just wasn't enough DAI liquidity at the time so the unthinkable (to me) happened, the proposal to introduce USDC as collateral for DAI was proposed. So basically a "decentralised" stablecoin backed by reserves of a centralised stablecoin which is itself backed by reserves in a USD bank account (shoutout to Alex Martell for this).

I posted this tweet which became fairly controversial but was interesting to see responses:

If I was to summarise many rebuttals to this I could summarise it as the following points:

  • MCD had been in the works for 2 years and people were aware this was going to be in the roadmap

  • More collateral types should be added to improve liquidity and resilience

  • We need more off-chain assets to grow DAI and rely on less-volatile assets

Overall I think these are all valid points, however I think perspective matters in how you interpret this.

  1. MakerDAO was promised to be a stable coin that relies on being 100% trust-less. A way to hold US dollars with no chance of censorship.

  2. The price to mint 1 DAI is high (150% ETH collateral), but is worth it because of the guarantee no censorship

Because point 1 is not true, point 2 does also not hold. To put it bluntly, why would I hold/mint a stable coin where underlying collateral can be seized and I still need to over collateralise it? Why not just use USDC directly?

On the flip side, let's assume that diversification is a good thing and we should sacrifice decentralisation for the sake of growth (bridging liquidity to trusted assets/securities). Okay but that raises the point that why was BAT added as the first collateral type while USDT isn't even talked about? To me this is a weird ideological decision that really doesn't make sense. BAT vs USDT volume compared below.

What I find fairly shocking or odd, is to see parts of the ETH community saying this is a good thing and this is good for the system. Coincidentally many also hold MKR. I think outside of ideological reasons, Allison Lu from UMA raises some good additional points:

Note: I don't hold any SNX or MKR. However, what I find almost comical to see is MKR shills saying Synthetix is self-referential when MakerDAO now holds that very same trait. The idea of having a global debt pool shared amongst all CDP holders could have been much more interesting although I'm not well versed enough in the Maker community to know if this actually was a good/viable option. On this topic, I also find it interesting to see how ICOs run by people close to the Ethereum foundation are never scrutinised while others are labelled as scams. There's perfectly logical reasons for it although just something worth thinking about even if you are a massive ETH/DeFi fan.

For me, the main take-away is that the idea of a fully decentralised stable coin is probably not going to be possible for a while. We may be able to get something close but the oracle problem is still a very hard one to solve. In the meantime, a much more promising path is acknowledging that centralised stable coins are here to stay and figure out how to leverage those assets to more exciting and useful things for the entire ecosystem.


The Future

Believe me when I say that I was one of the biggest fans of MakerDAO when I first heard about it two years ago. However, the following factors in aggregate concern me about its future and possible viability:

  • Large MKR holders increasingly participating on doing whatever they want. VC interests dominate community interests.

  • All core developers have left. Having been an engineer myself, whenever this happens it takes a tremendous amount of work from the leaders to turn the ship around. I do not have confidence this is happening at the moment.

  • Introduction of increasingly more illiquid/securitised assets. Real estate, bonds and much more. All in the name of "diversification". Liquidity is king and not many assets have good on-chain liquidity. Don't forget this.

  • Lack of an aggressive innovative roadmap. Compared to other projects which are hungry to continually innovate, Maker simply does not posses this trait. Execution is continually overlooked in crypto, I think it is here as well.

  • Tokenomics that really don't make sense. It took MakerDAO two years to burn a tiny percentage of MKR. The upcoming MKR auction will reset that to 0. Considering MKR appreciates only if ETH appreciates (or looks like it will appreciate), you should probably hold ETH.

If I was to sum all of this up, it really comes down to what I've mentioned in previous DeFi Weekly posts: when you're trying to be half-decentralised and half-centralised, you die.

Maker very much seems like it is on this trajectory.

Agree, disagree? Let me know via Twitter or replying to this email.