Mariano Conti of MakerDAO, head of smart contracts at MakerDAO, reads from his essay about his personal experience with cryptocurrency, particularly as an Argentine whose family has experienced periods of hyperinflation, and as a freelancer accepting payment from foreign countries. He describes the evolution of Maker from single-collateral Dai (now Sai) to multi-collateral Dai, and also gives his perspective on the events of Black Thursday and the subsequent debt auctions the protocol undertook to right itself.
Thank you to our sponsors!
Mariano Conti: https://twitter.com/nanexcool
Linda Xie’s essay about crypto memes on Unchained : https://unchainedpodcast.com/how-memes-can-help-crypto-go-mainstream/
El Auge de MakerDAO: Un Viaje Personal (Essay in Spanish)
My grandparents died without access to their life savings.
You see, in 2001 Argentina had one of its worst economic crises to date. The Argentine Peso, for years artificially pegged 1 to 1 to the US Dollar, could not hold on to its peg anymore. Exports stopped being competitive. There was no independent monetary policy. People were exchanging pesos for dollars, withdrawing them from banks, and often transferring them abroad, or holding them in cash “under the mattress”. There was fear of a bank run. So on December 1st, 2001, the Argentine government decided that people’s money wasn’t theirs to enjoy, and instituted the “Corralito”, a freeze on everyone’s bank accounts, allowing for only a small amount to be taken out per week. No withdrawals from US dollar-denominated accounts were permitted, unless you agreed to turn those dollars into pesos. Credit and debit cards still worked, but this lack of cash caused many problems in the economy.
From December 21st 2001 to January 1st 2002, the country had 5 presidents in 11 days. The situation was getting worse. The bank freeze was still ongoing, but now, most deposits were exchanged for peso-denominated bonds. To make matters worse, all dollar-denominated accounts were also turned into pesos, or “pesified”. Then they devalued the peso. First going from 1 to around 1.4 pesos per dollar, and then as they let it float, it reached around 4 pesos per dollar in a matter of months. This was around December 2002. Only a year later.
You can see where I’m going with this. Not your keys, not your coins. Money in the bank, or in the government’s hand, is not your money. I just wish I could have told my grandparents this mantra.
This was not the first time something like this has happened, of course. For years, Argentina has seen many devaluations, periods of hyper-inflation, bank runs, capital controls, asset freezing, among other things. It is one of the reasons people buy dollars and keep them under the mattress, why grandparents give their grandchildren 10 dollar bills on their birthdays, and why Argentina’s general population does not trust the government or the banks.
It is also why I believe Argentinians are sometimes more willing to take risks in other areas, and at times seem more finance savvy than in other places in the world. Here, most people actively work to not lose, let alone make a profit. If you forget a 100 peso bill in your jacket pocket and find it a year later, congratulations! You just found a 50 peso bill. It’s the reason why we research every alternative available, always searching for access to dollars, so many times denied to us. And in cryptocurrencies, it seems like we’ve found our answer.
My journey with cryptocurrencies started in 2014 out of necessity to control my own money. But let me take yet another step back. After a long period of living in Mexico City, (I moved there in ’86 at age six), I decided in late 2010 to move back to Buenos Aires, the city where I was born. I’d never felt fully Mexican or fully Argentinian, like I never belonged to either. And as I turned 30, I thought this was my opportunity to reconnect with my roots. So in April 2011, having broken up with my girlfriend, sold or gifted most of my possessions, I traveled to Buenos Aires with two suitcases, one filled with clothes, the other with a computer.
I could not have picked a worse time to relocate.
Say what you want about Mexico, but its closeness to the United States means it is a very stable economy, for the most part. Most of what I’ve mentioned about Argentina I had lived only through tales, anecdotes from relatives, or the news. Just six months after having moved back to my home country, the Government instituted new capital controls: regular people were mostly banned from purchasing dollars in banks and exchanges. Pair this with an economy that regularly sees double digit yearly inflation and the Argentine appetite for dollars, it was a recipe for disaster… and black markets. Dollars were bought and sold at a 50% premium of the “official” rate. This has gone on for years.
So in 2014, when I found myself working at a small digital agency in Buenos Aires, I negotiated a regular salary in pesos, and another “under the table” salary in dollars. This was standard practice for IT and tech professionals. The problem was getting paid those dollars. My employer could not deposit dollars into my dollar-denominated bank account in Argentina, they would be converted to pesos instantly at a rate 50% or lower than what I could sell them for in the black market (we call it the blue market, or blue dollar, don’t ask me why). At the time the official government quote was around 8 pesos to the dollar. In the blue market, a dollar could get me almost 14 pesos.
A deposit to my local bank account was no good. Flying to Miami and opening an account there was not an option. Paypal had long been banned. My employer owed me more than six months of salary. Until one day he mentioned Bitcoin.
The Bitcoin Era
Now, Bitcoin was already a thing in Argentina, it just wasn’t my thing. For a long time, Argentine freelancers had used Bitcoin as a means of getting paid by US and European companies, often negotiating salaries that were denominated in dollars, and with the freedom of getting paid in something that the Government couldn’t get their hands on. I just hadn’t seen its true potential till I personally had to find a way to keep my money in something stronger than the Peso.
So I did a bit of research, discovered that it was totally my thing, and proceeded to collect six months of salary, in the time it took to send some Bitcoin from one account to another. It felt like magic. I would later find out that I became part of a larger group, so many engineers in Argentina had gotten a head start using Bitcoin. Many because they saw the potential, but also many just out of necessity. Is this another avenue that leads to dollars and keeps inflation at bay? If so, then we will welcome it. It was a way to beat the system, a way to retain ownership of one’s work, a sidestepping of regular banking.
It was not without problems. Someone getting paid in Bitcoin had to be sure that clear terms were drawn out. Do you get paid at the beginning of the month or when a project is done? What is going to be the exchange rate? Is it when payment is due or when payment is finally done? Depending on the answers to these questions, you could either make a lot of money, or lose a lot of money. Oftentimes, these transactions did not end well, because of volatility. I loved the freedom that Bitcoin provided, but there had to be more, right?
I discovered Ethereum in late 2015, a few months after Mainnet release, and instantly fell in love. Here was a blockchain and cryptocurrency that had all the benefits of Bitcoin but it was programmable. To my Engineer brain, there was nothing better. The possibilities were endless. It was like finding a new world all over again. I put most of my life savings into Ethereum, knowing that this was where I needed to be, where my focus should be. There wasn’t much there yet, but the thought of what could be was alluring.
And then The DAO hack happened. It didn’t shatter my hopes in Ethereum, but it did remind me of the risks involved with any new technology. I said to myself I would be fine either way, hard fork or no hard fork. I decided such risks were very much acceptable, the alternative being using the traditional Argentine banking system.
But there was still something missing from Ethereum: stability.
I joined Maker in 2016, out of a series of coincidences. The digital agency where I was still employed contracted for Maker and I joined as a junior developer, at age 35. Slowly, I began to learn about the premise of Maker, of decentralized exchanges, of open finance. I designed the first version of the Maker Oracles, still used to this day. My main motivation was the promise of a stablecoin, an unbiased world currency. A medium of exchange pegged to the US dollar, which is a known and universal unit of accounting, but with all the benefits of a cryptocurrency. I was in, this was my new mission.
I had not been part of the initial team that wrote the white paper or worked on the first prototypes, but I still consider myself a Maker OG. During those days, we were still on the design and early experimentation stage. There was no Open Zeppelin, no Truffle, or if there were, they were still also in their very beginnings. Everything we did was new. If we needed a tool to solve a problem, we would build the tool.
Joining Maker also meant that I stopped getting paid in Bitcoin, and I started getting paid in Ether. This turned out to be a good thing for me, as the price of Ether (and really, almost every cryptocurrency) was going up and up. But trading some of that Ether into pesos to pay rent and groceries was not so easy. Bitcoin was still the most traded crypto in the market, so I had to do a jump from Ether to Bitcoin before I could enter the fiat world. Still, it was worth it.
Enter June 2017. After years of work, Maker released a private version of Single-Collateral Dai (which we now call Sai). The sole purpose of this first version was to be deployed to Mainnet, where the Foundation would open a CDP with Ether as collateral, and create enough Sai to pay for the salaries of a few volunteers inside Maker that wanted to dog food what we were building. After creating the Sai and sending it to employees, the system would be immediately shut down, and those of us with Sai would be able to claim a portion of the Ether collateral.
It worked perfectly. And it showed us that the system was getting ready for prime time.
A few weeks after and with some bugs fixed, a second version was deployed to Mainnet. This is the version we now refer to as Proto-Sai. Only the Maker Foundation and a few partners were invited to test it. There was not much of an economy around it. But still, this was the first asset-backed stablecoin on the Ethereum blockchain, soft pegged 1 to 1 to the US dollar, backed by Ether. And again, in the ultimate show of dogfooding, many employees and contractors, myself included of course, decided our next paycheck would be 100% Sai.
It was a historic moment. Seeing that number on an Ethereum transaction, knowing that as long as the system worked it would remain stable, was eye opening. I got flashbacks to the DAO, then those flashbacks turned into memories of tales of the Argentine banking system, and I felt safer. If I’m going to trust someone with my money, let it be the nerds and the innovators, not the politicians, the banks or the government.
There wasn’t much I could do with my newly minted Sai. This was still a private beta and there was no Dai Savings Rate or Uniswap or Compound. Other than sending it to a couple of people for novelty, the idea was that it would be converted once more into Ether. This happened a few times during development, but it gave us a glimpse of what was possible, of the future.
And so we arrive in December 2017, the month when Single-Collateral Dai launched. This was the proper release, the one the ecosystem was waiting for for so long. And it did not disappoint. It’s a wonder seeing a whole financial system grow before your very eyes, stemming from the simple promise of a stable asset. We saw the birth of Decentralized Finance, or Open Finance, whatever you want to call it.
To put things into perspective, projects like Uniswap or Compound Finance, so ubiquitous now, would not launch until almost a year later. I like to think, and I hope you agree with me, that having Single Collateral Dai around kickstarted so many of the projects that we see today, because of the promise of a stable medium of exchange. Companies like Instadapp started off as 48-hour hackathon projects, giving us different (and sometimes better) tools to interact with the Maker protocol. And they didn’t need to ask permission. There’s nobody from whom to ask permission in a decentralized, permissionless protocol.
But who makes the decisions? Who says how much Sai one can borrow for an amount of Ether? Who determines how much Stability Fee one has to pay for their generated Sai? That’s the role of Maker Governance and it’s what makes MakerDAO a Decentralized Autonomous Organization.
The Role of Maker Governance
In Single-Collateral Dai, there were not many levers you could use. It mostly consisted of two things: Sai is trading above a dollar, lower the Stability Fee. Sai is trading below a dollar, raise the Stability Fee. They were simpler times. Still, the Maker Community gathered every week, as it had done since 2015 to discuss how the system was operating and what to do about it. Fun fact, Governance calls used to be held on Sundays, in my case, at lunch time. I cooked many asados with headphones on, listening in on calls, much to the chagrin of my family.
But of course, as per the whitepaper, Single-Collateral Dai was not the end goal. Being backed only by Ether, Sai is still very vulnerable to black swan events. Say the price of Ether drops drastically in a matter of minutes or hours. It would be very hard for the system to recover from such an event. And even though Sai went through the greatest bear market crypto has ever seen, it maintained its peg and remained a stablecoin. Ether’s price went from $800 around the time Sai was released, up to $1,400 and down to $80. It did so over the course of several months though.
And yet, Sai kept up its promise of 1 Sai equals 1 dollar. And to an Argentine, this meant yet another way of getting access to dollars. When I said that it used to take me some hoops to go from Ether to Bitcoin to pesos, after a few months of Sai being on the market, those hoops were gone. Traders and over-the-counter trading desks started accepting Sai as much as they were accepting Bitcoin and Ether. Exchanges made it easy to buy and sell Sai, and have pesos deposited directly into your bank account. Debit cards loaded with Sai became the norm. And gone were the days of discussing whether your invoice denominated in Bitcoin would be honored at the price when it was issued or when it was paid. It just worked.
Stability brings prosperity, we all know that. And in an economy so unstable like mine, you pair stability with decentralization and transparency, and you get an unstoppable economy.
Here, permit me to jump ahead almost two years to November of 2019, and the release of Multi-Collateral Dai.
Multi-Collateral Dai (or MCD) represents the culmination of years of work by hundreds of people. It takes lessons learned from two years of running SCD, iterating and fixing bugs and assumptions. Here, Maker Governance has a much tougher job to do. It has to manage not one, but two different assets used as collateral for generating Dai: Ether and Basic Attention Token (BAT). Sai is also a collateral but it is a special case since it is only used for migrating from the old Single Collateral system into the new one. And in recent weeks, USDC has been added as the third collateral. Governance has to discuss weekly the state of the peg, the relation between price and Stability Fees and also one of MCD’s greatest achievements: the Dai Savings Rate. And as time passes, the community has not only grown, but it has become better at organizing itself, raising its voice, proposing changes to the system, and overall reacting quickly to market conditions.
The Markets Collapse
As I was writing this essay at the beginning of March 2020, there was a perfect storm brewing. An oil price war, a virus generating uncertainty in the markets. Late February brought unease to everyone. Eventually we saw an outright collapse of both traditional and crypto markets on March 12, 2020. Around 50% of Ether’s value (and most cryptocurrencies in general) was erased very quickly. And Multi-Collateral Dai, being backed mostly by Ether, was impacted as well.
The Ethereum blockchain became congested, with gas prices going up well into the 300s of gwei per unit of gas, something practically unheard of. Not only were transactions insanely costly, there were so many of them going on at once that most remained in the mempool, waiting for their chance to be mined. Every actor in the ecosystem was affected, be it liquidity pools, Maker Vault owners, Dai holders, or keepers. Keepers are (usually) automated bots that scan the Ethereum blockchain looking for arbitrage opportunities. In the case of the Maker Protocol, Keepers participate in collateral auctions that happen when positions become undercollateralized and the system auctions off this collateral to recover Dai and remain solvent.
Oracles as well, were affected. With the increasing gas bidding war, and so many transactions in the queue, the Oracles which are charged with updating the system with real world prices, were slow to update. In retrospect, this allowed some Vault owners to add more collateral to their Vaults or pay off some Dai, because they had some extra time to do so before the Oracle would update with a lower price, triggering their liquidation. But even then, there just wasn’t enough room in the Ethereum blocks to fit everyone’s transactions.
As a result of all the above, a large number of collateral auctions were triggered, and a subset of these auctions were won by bidders who submitted bids just above zero.
This was truly the perfect storm. Vault owners struggling with high gas prices and network delays, trying to add Dai to their Vaults. Exchanges were hit as well. Coinbase effectively froze withdrawals for a while because of how much gas it was costing them. And with the increasing demand for Dai and the liquidity crunch, the price of Dai started going up, reaching almost 1.14 Dai per dollar on March 12.
Keepers struggled too. While the Maker Foundation runs a keeper to participate in collateral auctions, it too felt the stress of high gas prices and delayed transactions, and the lack of liquidity. With so many auctions happening at once, there was not enough time to recycle collateral back into Dai to participate in further auctions. Out of four known Keeper bots that participated in these auctions, two started submitting “zero bids”, (in fact, bidders submitting bids decimal points above zero), and ended up winning more than four millon dollars’ worth of Ether for basically free.
After a while, Keepers managed to regain liquidity and started bidding again, restoring competitive auction space. In the end, around 1,200 Vaults were liquidated, resulting in 4,447 auctions. Because of the collateral auction shortfall, the Maker System would need to recover approximately 5.4 million Dai.
Here is where a decentralized and transparent Maker Governance and Community becomes so important, and it’s also here where it shined the most. In Maker Forums and Chat, the community started discussing the events and planning how to best react to them. There was a vote to change some risk parameters in the system, giving Keepers more time to bid on auctions. Also important was to bring the peg back down to 1 dollar. Stability Fees were lowered and the Dai Savings Rate was brought down to 0 for the first time since it was created.
Also, a new collateral uncorrelated with the crypto market was added: USDC, the fiat-backed stablecoin. This would provide diversification and be an additional source of liquidity should it be needed. This was not without controversy, and I encourage anyone to watch or read the transcripts of the Governance calls from March 12 through March 20 to understand why the community made this decision.
The system also needed to recover 5.4 million Dai. The Maker Protocol is designed in such a way that MKR holders cover this shortfall by way of MKR dilution. If a Vault is liquidated and the collateral auction does not raise enough Dai to replace the Vault’s outstanding Dai, after a waiting period, a debt auction is created that seeks to raise this Dai by minting of MKR, and thus diluting MKR value. It is a risk that MKR holders know about and urges them to act responsibly, be informed, participate and vote.
As I write this, the Maker Protocol created and successfully completed 106 debt auctions. Each one raised 50,000 Dai and offered decreasing amounts of MKR. The total MKR minted was 20,980.
The Maker Community is discussing the circumstances around the zero bids and liquidations, and it is they who will ultimately determine how to address this issue. I encourage you to go to the Maker Forums and participate in this discussion.
There is much to be learned from this experience and I expect everyone to come out stronger in the end. There is increased participation in governance calls and voting. New tools are being built to deal with auctions, both by the Maker Foundation and the greater DeFi community. MKR holders understand now that it is a great responsibility to hold MKR, because their swift action or inaction can have good or bad consequences.
As for me, I fully believe in the system, the Foundation and the community. My paycheck remains 100% Dai. And I will continue to work on making the Maker Protocol more resilient for as long as I can.
Did you notice something? At some point during my tale, I stopped discussing Argentina’s economy and its inflation. This is because Dai has given me the privilege and opportunity of earning a stable asset that doesn’t require so much effort and planning. It doesn’t make me think “how much am I going to lose this month” or “do I put my money in savings or try to buy something tangible like a motorcycle, that more or less follows inflation”. I just work and every month, I earn a salary in a strong currency. It’s as simple and beautiful as that. What so many countries take for granted, we cherish.
Linda Xie mentions in her own essay released on Unchained “How Memes Can Help Crypto Go Mainstream” that she asked for perspectives of different kinds of crypto users. And she says and I quote: “I received by far the most responses from Argentinians”. I am not surprised by this.
The journey to control one’s money in this difficult economy has led many regular Argentinians to discover cryptocurrencies, but in a different way than I or a technical person would have done back in 2012, 2013 or 2014. With the rise of stablecoins, the idea of money that cannot be censored or seized, yet remains predictable in its value, has gotten people of all aspects of life interested.
And it is here where I make the distinction that gives us South Americans an edge: an insanely high tolerance for risk. When you believe and know in your heart that nothing is riskier than your government or a bank, any alternative becomes much more enticing.
I’ve seen a new economy form before my very eyes, on the Internet, in Telegram and WhatsApp groups, via weekly meetups, that I never saw possible with volatile cryptos.
If you want to see the real power that cryptocurrencies have on regular people, I urge you to look South. You will find a large number of savvy users, who’ve been finding ways of protecting their money for years, and always looking for the next big thing. I believe decentralized stablecoins like Dai are the next big thing they’ve been waiting for.
These aren’t people looking to go long, or margin trade, or leverage. They are Argentines, Colombians, Venezuelans, and others that want to save and transact. They want to stop fighting to just break even, always lagging with inflation, and take back control of their money. They want to look at a banking system that has year after year, decade after decade, ruthlessly taken from them and say: ¡¡bastaI! ENOUGH!
Abuelo, I wish you could have seen this.