Lombardy: A Solution to the MakerDAO Problem
The Problem At Hand
With the massive ‘black swan’ event that occurred in the cryptocurrency market the week of March 9–13, 2020, we came to find that the core mechanism supporting the MakerDAO system failed.
For those not familiar with MakerDAO, it is a smart contract system that lets people lock up their Ethereum in return for an amount of ‘DAI’ stablecoins that are worth $1 each. People can then return the DAI for their original Ethereum at any point in time.
To maintain the stability of the DAI token, which is backed by Ethereum and not the United States Dollar which it simulates, there has to be more Ether in the system than DAI. This is because Ethereum is a volatile asset, so if the total reserve value decreases 5% in a day, that fluctuation should not have an effect on the value of DAI. There should still be enough money in the system to cover the DAI when the user redeems their DAI for Ether.
This is good under normal circumstances, and this mechanism is how normal banks give out loans. If you have a car worth $100,000 (very nice car), the bank might give you a secured loan for $50,000 to be paid back in a year. This is because the car would depreciate over time. It also assumes the risk that you never come back and they have to liquidate the car at auction.
MakerDAO has a mechanism to liquidate your Ether if you lose your DAI, mismanage it, spend it, etc. This is to make sure the reserves always have the right amount of value to them. There is an auction mechanism that allows anyone to bid on tranches (chunks of Ether) for DAI, which would, in theory, return the DAI you lost to them, even if that DAI was currently in the hands of someone else.
But what happens if no one shows up to the auction? Well, then MakerDAO cannot offload the Ethereum to hedge against the fluctuating prices and the reserves cannot cover the amount of DAI in circulation. Thus, the system becomes insolvent. This is essentially what happened when the Ethereum price dropped a whopping 40% in one day.
No one was bidding on the tranches of Ethereum that MakerDAO needed to hedge against risk. Instead, savvy opportunists showed up to bid ridiculously low amount ($0) for tranches and were *winning.* So not only was MakerDAO unable to liquidate appropriately, it was liquidating its assets for nothing, causing it to become even more strained and closer to insolvency. Brilliant.
This is exactly the same type of situation that an overleveraged bank would find themselves into if they loaned out too much money, the economy tanked, and then they could not find any buyers of their debt because everyone wants to keep their nonvolatile cash. However, banks are not stupid enough to let any shmuck show up and bid $0 on a $100,000 car and give it away for free. A computer program is, as computers are stupid and do exactly what you tell them to do.
The fallout of this failure brings sharp criticism to the entire idea of Decentralized Finance or DeFi in general. To be frank, DeFi has been a fun experiment of automating financial structures on the blockchain, but has really been overhyped as some godsend that is going to disrupt humanity forever. Where is has failed is that people are too eager to adopt any solution that has ‘blockchain’ in it without passing it through rigorous criticism. People want blockchain to happen so badly that they are willing to put up with any solution, even if that solution is rotten or has core flaws. Try to sway a Ripple evangelist, and you’ll know what I mean.
So what is the solution? Well, for starters, we can stop being sycophants to blockchain technology and actually hold the developers accountable for their work. Any economist who read the MakerDAO whitepaper would have seen this coming a mile away, which is actually exactly what we did.
Our chief operating officer Dr. Alexander Butenko holds a PhD in Economics, so I figured that he would have an opinion or two about MakerDAO. I gave him the whitepaper and told him to look it over and tell me what he thought. He came back to me very excited to discuss what he had found. Alex told me that the MakerDAO system was flawed from the start and made absolutely no sense. In fact, he taught me more about the classical financial institutions that deal with this exact same issue on a daily basis.
The problem with MakerDAO is that it is putting too much faith in one of the most volatile assets on Earth: Ethereum / cryptocurrency in general. Without a heavy amount of extra reserves, the system was doomed to collapse from the beginning. The problem is that the amount of DAI issued for the amount of Ethereum posted is far too high, and if it were closer to the actual risk of MakerDAO holding the Ethereum as collateral, the amount of DAI issued would be substantially lower. In fact, it would be so low that users would be generally uninterested in using the platform altogether.
Finding a Solution
So what is the solution? Is DeFi a giant scam? No, it is a natural evolution in finance. That part is true. However, to not critically examine a system and just ‘ship it’ without consideration for real economic factors is naive and doomed to failure. Through these failures, the community learns how to make better systems. In fact, Alex and I spitballed on a whiteboard and came up with a system that would remove any counterparty risk and be able to replace MakerDAO entirely.
Rather than having a smart contract accept the risk, simply have the smart contract manage the transactions between private lenders (MakerDAO / the bank) and those seeking to post collateral for a loan (users). This way, you can return the human aspect of risk analysis to each transaction. You also are able to support pretty much any asset you want and make any sort of trade you want.
Think of it as a pawn shop. You can take your guitar to any pawn shop in town, shop rates, and get the best deal. The pawnshop owner is a private individual who can give you the deal that works best for them. They are not automatically obliged to accept your collateral, whereas MakerDAO is. They are not obliged to give you a stablecoin either. Maybe you want bitcoin? Maybe you want a tokenized stock? The possibilities are endless.
In this system, which we call ‘Lombardy’ after the original pawnbrokers of the Middle Ages, ‘merchants’ and ‘borrowers’ come together on the smart contract to post bids. A borrow would post a request and their collateral on the market. Merchants would then bid for that collateral by offering deals which would include factors such as:
- Amount of the loan.
- Asset type of the loan (cryptocurrency type, etc.)
- When the loan has to be paid back before the merchant can liquidate it.
- How much the borrow has to pay on top of the loan when they come back for their collateral.
- How much the collateral can depreciate before the merchant can liquidate it.
- How the merchant determines the price of the collateral, generally through an oracle, or potentially provided by the smart contract system via a trusted feed.
This eliminates black swan event risks, decentralizes the market (multiple merchants can compete with each other), and even allows the development of multiple stablecoin assets rather than a single DAI.
The smart contract can also have the mechanism for merchants to find buyers of their collateral. Hopefully, as this idea becomes more mature, we will be able to provide a higher fidelity specification of a system of this type.
Next Steps
So what is next? As it remains, this is an idea that we have fleshed out beyond simply a scope. This would be something very cool to launch on the Lamden platform as adoption increases and people begin to see why Lamden is superior to other blockchain solutions. For us, the development speed of such an application is about 10x that of what it would take on Ethereum. The depth of unit tests, integration tests, user interface development, and backend integration, is about just as fast to develop. Thus, Lombardy will continue to be fleshed out and will be released when the time is right. Hope to see you then!
