Refund is a word rarely heard in DeFi. So is redemption.
Malt protocol is an algorithmic stablecoin that launched two weeks ago on the Polygon network. After an explosive launch, bugs in the smart contracts caused the MALT stablecoin to lose its peg and left many users unable to withdraw their funds. The protocol was essentially shut down and the team immediately promised to compensate users after analyzing the situation. While many in the project’s Discord remained unconvinced, the team did a great job managing expectations with daily updates.
A week later, the team released a compensation plan outlining that users with stuck funds and those who participated in failed arbitrage auctions would be reimbursed 74 percent of their initial investments.
Cloak of Anonymity
Yesterday, just 12 days after the fatal bug was discovered, more than 1,900 users were airdropped their refunds totalling 13.4M DAI.
Many project teams use the cloak of anonymity to shield themselves from legal liability; that also makes it easier to walk away with minimal repercussions when something goes wrong. Users who venture into degen territory have gotten used to seeing compensation plans and IOU tokens with little hope of actual repayment.
In such an environment, it’s refreshing to see a team step up and take ownership of mis-steps. The community was understandably elated at receiving their money back.
We caught up with 0xScotch, the founder of Malt protocol. Here’s what he had to say:
The Defiant: The last two weeks must have been difficult to say the least. How did the team handle it?
0xScotch: The past few weeks have been crazy. Malt was just my passion side project two weeks ago. Basically the only people that knew about it were people I knew IRL and a handful of people who had heard about it from friends. I was not prepared for how quickly it would blow up. It all came as such a surprise and I had to learn a lot about what it means to run a DeFi project. Then the bugs started and I found myself having to manage a crisis and navigate keeping the community informed, while also trying to make good decisions about what the correct course of action is. I was mostly on autopilot the whole time.
When we lost peg I jumped on a Discord call with over 100 members of the community to answer their questions, and I think that went a long ways toward building the strength of community that we see now. The community that has developed over the last few weeks is nothing short of extraordinary, despite the circumstances. There is clearly an immense appetite for innovation in difficult areas. It’s amazing to see people coming together in pursuit of something cool.
The Defiant: The majority of the reimbursements came from the project Treasury that was earmarked for development. You could have walked away with $10M+. Why stick around?
0xScotch: What made me stay is that staying is obviously the right thing to do. I did nothing to deserve $13M, I wrote the bug that caused it, so it’s on me to do my best to make it right.
I think we have a strong concept for an algo stable that could actually work. But algo stables are hard and the journey will never be smooth. There will be bugs, there will be protocol mechanics that could be better. For us to have the time to be able to continue building and improving on this mechanism we need a community to support us, so perhaps somewhat selfishly the best way to do that is to keep the community happy. So paying as much back as we could was morally the right thing to do but it is also the best decision to further our own goals of being able to keep improving the protocol.
The Defiant: What can you tell us about MALT V2?
0xScotch: In terms of V2 we can’t say too much yet because we haven’t finalized any details but we have some exciting ideas. The above peg mechanism really worked well. The auction kind of worked but definitely has a lot of room for improvement. Auctions got aggressively frontrun which extracted value from the liquidity extension (LE). We need to improve that. We definitely need a higher % of LE as well. At least initially. Dynamic LE is an idea we are throwing around. The auction time is also something we need to work on. It was 30 minutes but I changed it to 10 minutes, which definitely improved the situation.
Overall we are really happy with how V1 core mechanics performed. It was just a bug that stopped us this time. We learned a lot and have a lot of data to use to improve V2. In many ways failing fast was a blessing in disguise because we get to start over and learn from the data rather than hobbling along with a subpar V1.