From Hayek’s denationalization of money to algorithmic stablecoins
A review of Ampleforth, Terra, and Djed
Source: Bit2Me Academy
Have you ever thought of competing private currencies where individuals are able to choose the currency of their preference? In 1976, the Nobel Laureate Friedrich A. Hayek published Denationalisation of Money. In his book, Hayek envisioned a world where private token money could get into circulation taking away the state’s monopoly to produce money. In the chapter “Putting Private Token Money into Circulation”, he wrote about competing currencies that would certainly prove a more effective constraint, forcing the issuing institutions to keep the value of their currency constant […]. Moreover, he stated: The managers of the bank would learn that its business depended on the unshaken confidence that it would continue to regulate the issuance of private currencies so that their purchasing power remained approximately constant […] To achieve this aim, the amount would have to be promptly adapted to any change of demand, whether increase or decrease.
Almost 50 years later, Hayek’s private money in circulation has its closest configuration in the advent of blockchain-based algorithmic stablecoins, where the adjustment of the money supply to changes in demand is apolitical, decentralized, and fully automated by smart contracts.
Digging into algo-stablecoins
Cryptocurrencies are extremely volatile. Broadly speaking, cryptocurrencies do not accomplish the three characteristics to function as ‘money’, i.e., to be a good store of value — something that will retain its value in a reasonably predictable way; a unit of account — an easy way to price goods and services, and a medium of exchange — something that people hold because they can swap it for other things or currencies. Due to the high variability in value, cryptocurrencies cannot be used as units of account, therefore complicating their utilization as a medium of exchange. Imagine customers would need to know the exchange rates every time they want to buy something priced in BTC or ETH.
The volatility of cryptocurrencies is addressed by stablecoins that try to maintain the characteristics of the native cryptocurrencies in a blockchain such as decentralization, censorship resistance, trustlessness, while seeking to function as money. Stablecoins represent — in my view – the closest to Hayek’s private token money in circulation with the potential to transform legacy payments infrastructure such as SWIFT, interbank network handling global payments and others by ushering in a new era of frictionless global real-time value settlement.
Stablecoins are cryptocurrencies that keep its price pegged to a fiat currency that is widely accepted as ‘sufficiently stable’ such as the USdollar, Euro or gold. There are different types of stablecoins. One class is fiat-collateralized where the price is kept stable backed by the correpondent amount of fiat reserves, in this case we find examples such as Tether. Fiat collateralization has a few shortcomings, for instance, the need to use the traditional financial system to invest the reserves in money markets or bond markets which could transfer new risks to the stablecoin. Perhaps a major drawback is that these stablecoins are centralized demeriting the case of using a distributed ledger (i.e. the blockchain), first by trusting third parties in the management of reserves and secondly because the supply adjustment to changes in demand relies upon the operators of the stablecoin. A second type of stablecoins is crypto-collateralized which are backed with cryptocurrencies — for example DAI — posing a few other challenges, for instance the need to address the volatility of the cryptocurrency, higher risk of autoliquidation, or over-collateralization requirements. A third type is algorithmic stablecoins, where the adjustment of the money supply to changes in demand is encoded in the protocol, providing more capital efficiency and decentralization. Algo-stablecoins entail a few challenges as well, for instance, to keep stability and liquidity, and algo-stablecoins require continuous growth. Algorithmic stablecoins are the ‘youngest’ in the scene however there are some sound cases out there which will provably unseat centralized stablecoins and fiat currencies.
Three different stabilization mechanisms in algorithmic stablecoins
The mechanism by which an algorithmic stablecoin keeps its peg to the fiat currency is not unique. The following cases: Ampleforth, Djed and Terra utilize different mechanisms:
Ampleforth is an uncollateralized algorithmic coin for stable purchasing power. The coin, AMPL, is an ERC-20 token designed around the concept of denationalized money, functioning as base money.
Ampleforth has a maximum supply of 395,345,189 coins. As of writing, the circulating supply was 93.03m AMPL with a market cap of US 87.75 million. AMPL has a modest size and presence, for instance, it is ranked out of the top 100 cryptocurrencies. Nonetheless, AMPL is designed as a new primitive building block in DeFi just like BTC or ETH and its insertion to DeFi is growing for instance, AMPL is a lend or borrow asset in Aave alongside integrations to Avalanche, Binance Smart Chain, Uniswap, and Balancer which will foster adoption and growth.
Among the use cases, AMPL is considered to be a unit of account — i.e., can be used as a means of payment — . As a store of value, it is designed to keep purchasing power, and in DeFi, AMPL is suitable to denominate long-term on-chain stable contracts and crypto-backed derivatives. Moreover, it is functional money free from centralized collaterals, lenders of last resort and beyond the reach of politics.
AMPL token design
Ampleforth has an elastic supply that absorbs the volatility of demand by setting a target price (Pt) and a price threshold (δ) creating a price-band around the target. Within that interval, the price of AMPL is considered “stable” or in equilibrium. Whenever demand changes and the unit price gets out of that range, the supply expands or contracts proportionally until the price gets back to the equilibrium band (see the figure below). Expansion or contraction of supply modifies the quantity of AMPLs in holder’s wallet such that their net balance remains constant, calling this a rebase of balances. The user’s percent ownership remains fixed unless the user chooses to sell or buy AMPLs.
Suppose that Alice has 1 AMPL worth US$1.00. If demand increases such that 1 AMPL = US$2.00, the nominal exchange rate in AMPLs is higher than the target, thus a supply expansion occurs. In the next equilibrium, Alice will have 2 AMPLs worth US$1.00. The supply change is computed and executed no more than once every 24 hours based on the latest price difference and it is distributed uniformly over the course of k days. Changes in supply are timestamped automatically therefore it is possible to know when it will occur and in what measure.
Ampleforth is not strictly a stablecoin because it is subject to price fluctuations. Notwithstanding the foregoing, AMPL is capable of bootstrapping volatility by keeping the price fluctuation within an interval.
Ampleforth’s volatility fingerprint
Source: Ampleforth whitepaper.
In the figure of the left, an expansion of supply is represented by the semicircle in dotted lines above the price band whereas a contraction of supply is represented by the semicircles below the price band. During these dynamic periods where supply of AMPL is moving towards a new equilibrium, there is a time lapse where arbitrageurs can sell or buy AMPL, right after the supply increases or reduces but before the price correction takes place, producing patterns (step-like functions) in supply and market cap (middle and right figures), outlining its own volatility fingerprint. After the adjustment period, new equilibria in supply and market cap are reached, represented by the straight horizontal lines in the figures above.
The volatility fingerprint of supply and market cap is showing, foremostly, that there is no fixed trend in the long-run. The trend can alternate up and down depending on changes in demand, whereas the price fluctuates around band. In the figure below, one year series of price and stock shows that the trends match with the theoretical design.
Taken from Ampleforth.org
Moreover, AMPL is decentralized, meaning there is no control in the supply adjustment, the protocol sets the incentives by propagating the latest price information and the fast-actors (arbitrageurs benefiting for near term trades) propagate supply information back into price by selling or buying AMPL during the lapse of time they can make profits from an overpriced or underpriced AMPL. For AMPL holders, the dynamics between one equilibrium and the other goes unnoticed since their net balances are rebased.
The first thing you read in the website says: “Terra is a public blockchain protocol deploying a suite of algorithmic decentralized stablecoins which underpin a thriving ecosystem that brings DeFi to the masses”. In Terra’s philosophy, fulfilling the monetary needs of the average customer is the key to power price-stable global payments systems. In Hayek’s view, Terra is competing to attract billions of customers to use their stablecoins in the new Web 3.0 monetary order that is configured.
To do so, Terra uses a dual-coin mechanism where Luna (native token) absorbs the volatility of the various Terra stablecoins. In this case, an elastic monetary supply adjusts to keep the peg, contracting or expanding, according to changes in demand for money. Luna is also used as a governance token.
The dual mechanism works as follows: if the price of UST stablecoin (pegged to the US dollar) falls below the peg, Luna is auctioned in exchange for UST, that is, the system mints Luna and burns (destroys) UST. This process will contract the supply of UST, increasing its price until reaching the peg again. In this operation, there is a risk-free profit by arbitraging the difference in price since the system always exchanges at price of $1.00. Example: if UST falls to 0.90, by trading UST for 1 USD worth of Luna from the system, an arbitrageur makes a 0.10 risk-free profit compared to the 0.9 USD worth she could get in the open market. Conversely, when the price of UST is above the peg, the supply increases by minting UST and burning (destroying) Luna, reducing UST price until reaching the peg. Arbitrageurs also make risk-free profits for example, if UST increases to 1.10, arbitrageurs can trade 1 USD worth of Luna to the system to get 1.1 USD worth of UST. Originally, Luna burns were tokens deposited into Terra’s Treasury, using the funds to support new developments in the ecosystem and keep a growing demand for Terra stablecoins. In the recent upgrade, the protocol was modified to effectivley burn Luna tokens.
In lieu of the Treasury, earlier this year, Terra Money announced the commencement of Luna Foundation Guard, an organization that will support projects whenever they are open-source applications, algorithmic stablecoins, research and education and the building of community-driven ecosystems.
Terra stablecoin (UST) has gained traction becoming fourth among its type with 14.9 billion market cap as of writing. Likewise, Luna has climbed up — in less than one year — becoming part of the large market cap tokens with 32.25 billion.
Nowadays, Terra Ecosystem has approximately ninety projects, sixty are live platforms where Terra stablecoins are used in the ample functions of money, for instance, Anchor is a savings platform where UST can be stored with a not negligible 20% APY. Nexus Protocol is a yield farming optimizer, or Apollo Dao where capital investments take place. Chai is a payments platform where KRT (Terra stablecoin tracking the South Korean Won’s price) is used as a means of exchange, likewise Memepay does in Mongolia. There are also platforms using stablecoins in e-commerce, money markets, and automated market makers.
Djed is a newborn stablecoin in the Cardano ecosystem. The whitepaper was released in August 17, 2021. Djed is a hybrid stablecoin because it is crypto-backed and algorithmic. In Djed, an autonomous bank — powered by smart contracts — buys and sells the stablecoin for a price that is pegged to a target, keeping the volatile currency as a reserve. While doing so, it charges fees and accumulates them as part of the reserves, benefiting reserve holders who ultimately contribute to the reserve funding and take the risks of price fluctuations.
Djed has gone through a long verification process including a robust mathematical research checked on paper and through Isabelle, an automated tool for theorem proof after which Djed was tested.
Djed’s token design
Djed has a more complex design than the previous two cases. The protocol starts with a simplified version called minimal Djed that encodes the stability properties under which the automated bank operates.
In minimal Djed users buy stablecoins from the contract and deposit the equivalent value of the pegged fiat currency in a base coin (e.g., ADA, ERG, etc) and the contract mints Djed. Likewise the user can sell stablecoin and get back the equivalent dollars in base coins. The protocol establishes a lower and upper bound where the smart contract is always willing to buy stablecoins starting 0.99 cents and sell at 1.01. If reserves were too low due to fluctuations in the price of base coins, holders of stablecoins will not receive 1 dollar for each Djed, instead they will be receiving the proportional amount of reserves in relation to the total number of Djed in circulation (R / Nsc). The protocol is gauged to resist a market crash up to 67% where Equity never goes zero, relieving the protocol from an insolvency problem and bankruptcy. However, to improve the situation in which holders of stablecoin receive less than 1 dollar for each Djed, the protocol buys and sells reserve coins (RC) to provide more equity to the contract to have enough reserves.
Source: Input / Output
The protocol sets limits to the buy / sell operations of reserve coins by establishing the reserve ratio (r ) equal to Reserves divided by Liabilities (i.e., the equivalent in US dollar circulating in Djed) and the parameters that do not allow the operations. If too many RC were sold the ratio can be too low and the contract will not be able to face liabilities. Likewise if the demand for Djed increases. In the opposite case, one might think that a verey big (r) is desirable, however there is also a maximum to avoid diluting the fees per reserve coin that is the incentive for holders of reserve coins.
During the testing period, it was detected that minimal Djed required a few adjustments, to mention a few, reserve draining was possible if a malicious actor could estimate the future exchange rate; the fees in the minimal version were rigid and could not serve as a lever to encourage or discourage operations (to buy / sell reserve coin) whenever it was required, more importantly, if the peg was lost, stablecoin holders could suffer financial losses and as the reserve ratio approached to zero, holders of reserve coins the possibility of a bank run was there.
An extended Djed protocol overcomes the issues detected in the minimal Djed, moreover, it considers dynamic fees that are increased for operations that shift the reserve ratio away from an optimum reserve parameter and decreased fees for operations that bring it closer to the optimum. By doing so, holders of Djed have robust stability conditions while reserve coin holders have attractive rewards.
Future developments can be expected in Djed, for instance, the possibility to stake the Reserves and provide better incentives to reserve coin holders, introducing governance to the protocol — for instance allow the governance community to adjust key parameters in the protocol — , KYC and AML functionalities, debt financing for the autonomous bank, pegging the stablecoin to a basket of assets, and so forth.
Djed has quite relevant use cases besides being the stablecoin for the DeFi ecosystem in Cardano and Ergo blockchains:
One of the most outstanding use case is that Djed is meant to be a decentralized stablecoin used for payments. Issued by Cardano’s partner COTI, the stablecoin will be instantaneously available in AdaPay — a payments gateway already supporting over 30 fiat currencies. Secondly, Djed is the stablecoin to denominate all the Cardano’s network transaction fees, introducing the novel concept of stable fees, which will be a major advantage and game changer for users and operators of DeFi platforms that are currently bearing the headaches of highly volatile gas fees. Last but not least, Djed is a powerful diversified stablecoin as the autonomous bank can hold reserves in multiple cryptocurrencies including stablecoins (e.g., ADA, BTC, ETH, USDC, etc.) and the target price can be referenced to fiat currencies or other assets (e.g. dollar, euro or gold).
If Hayek could see what algorithmic stablecoins are capable of doing, I think he certainly had written a few of them as vivid examples of private token money in circulation. Perhaps algorithmic stablecoins are the killer app for massive adoption in payments, savings, trading, long-term contracts, etc., as they are proving the possibility to create monetary systems free of political controls and a variety of ther such that customers can choose the most competitive stablecoin that fulfills their needs. Finally, for their capital efficiency — seen in decentralized monetary adjustments — algorithmic stablecoins might be unseating the collateralized ones and perhaps fiat currencies too.
From Hayek’s denationalization of money to algorithmic stablecoins was originally published in Coinmonks on Medium, where people are continuing the conversation by highlighting and responding to this story.