The Rise and Fall (and Rise and Fall) of Ampleforth — Part II
By Derek Schloss and Stephen McKeon
Ampleforth has received a lot of attention lately from both market participants as well as new experimental projects that are incorporating similar features into their own protocols. It’s a project that we’ve been following for a long time so we wrote this two-part article as an explainer and a framework for thinking about how it may evolve. Part I covers Ampleforth’s design and pricing mechanics. Part II (below) covers Ampleforth’s psychology and demand drivers.
The Psychology of Ampleforth
The network value of Ampleforth spiked in July and subsequently retraced about 65% from the high point. What caused this spike in demand and, more importantly, what are the longer term drivers of demand? We can point to three factors:
- (1) The Power of Incentives
- (2) Short-Term Speculation
- (3) AMPL’s Rising Economic Utility
The Power of Incentives
The initial spike in demand for AMPL illustrated the power of incentives. As we’ve been watching Ampleforth over the past year, one question we’ve debated is what the catalyst might be for a greater number of people to start participating? If AMPL units are going to be used as a commodity-money in the Ethereum ecosystem and beyond, the Ampleforth system needs more players.
Enter yield farming.
On June 23, 2020 the Ampleforth team launched ‘Geyser’ — an incentive program that distributed rewards (in AMPL) for users who supplied AMPL-ETH liquidity on Uniswap, a non-custodial decentralized trading platform. Liquidity begets liquidity, and this was the demand catalyst.
In addition to earning Liquidity Provider (LP) fees on Uniswap, users who provided AMPL-ETH liquidity could take their LP tokens and stake it on the Geyser platform (for background reading, see Placeholder’s Proof of Liquidity). Users who provided more liquidity on Uniswap and staked on Geyser would earn a greater share of AMPL rewards through the program.
It’s important to note here that blockchains aren’t just useful for tracking asset balances — they’re also pretty killer at tracking the length of time those assets are held. How does this play into Ampleforth’s incentive program? In addition to the amount of liquidity supplied in Uniswap, Geyser also rewarded users for the length of time they staked in Geyser — up to a 3x reward multiplier over a 90 day period of staking.
As a result, one could say that part of Geyser’s success thus far isn’t just predicated on incentivizing liquidity. By adding a temporal component, the program also incentivizes not withdrawing liquidity deposits. This has led to a mind-boggling ~70% retention rate since Geyser launched.
Geyser’s three stacked incentives — (1) income generated as a LP on Uniswap, (2) rewards generated based on deposit amount; and (3) rewards generated based on deposit length — has created a flood of new interest in Ampleforth with a few interesting second order effects on Ampleforth’s system.
One effect of Geyser has been the much improved distribution of AMPL ownership. In June 2018, Ampleforth was still a private project with ownership over the network tightly controlled by the team, their advisors, and a small group of investors. In June 2019, Ampleforth made headlines when they raised $5M in just eleven seconds in a public Bitfinex IEO — an important first step in the progressive distribution of AMPL. By June 2020, roughly 4,000 Ethereum wallets held AMPL.
Then came Geyser.
The number of wallets holding AMPL today has reached nearly 20,000 as of this writing.
Another effect of Geyser on Ampleforth has been the massively improved liquidity for AMPL. The AMPL-ETH Uniswap pool has been the largest pool on Uniswap for the last 30 days (reaching as high as $45M) — allowing a fairly deep market to form around the trading pair. At one point, the AMPL-ETH pool accounted for over 50% of total daily volumes, and about one third of the total liquidity on Uniswap. During the month of July, AMPL-ETH represented over $500M of Uniswap’s volume.
Ampleforth is optimizing around a commodity-money use case, and reliable money requires reliable liquidity. The incentives they created to establish a deep pool of liquidity catalyzed demand. As a result of Geyser’s success, the Ampleforth Foundation recently announced that 23.5% of the network would be earmarked to support Geyser or Geyser-like programs over the next 10 years, with a focus on decentralization, liquidity, network health, and broad rewards.
Short-Term Speculation
The second source of demand is short-term speculation, also known as FOMO, hot potato, trend trading, etc. The idea here is demand begets demand. As the asset gains momentum, more speculators pile in hoping to make a quick buck. Let’s be clear that this definitely drove part of the run up in AMPL’s network value and also that it’s not long run sustainable. But, we hasten to note that this alone is not a reason to dismiss the project. We’ve witnessed the same demand pattern in BTC and ETH. The popular press rushes to declare the death of Bitcoin after each of these events, and we’re loath to jump to the same conclusion on AMPL.
Repeated episodes of rise and fall are characteristic of cryptocurrencies as they go through the maturation process. BTC saw demand spikes in 2011, 2013, and 2017. Each event led to a huge run up followed by a substantial decline, but leveling out at a new higher price. For example, at the beginning of 2013Q4 BTC had a network value around $1.5B. It spiked up to over $13B that quarter, only to drop 50% relative to the ATH by 2014Q1 and then half again by 2015. It never touched $1.5B again, leveling out between $3–4B for most of 2015. Fast forward to 2017. BTC started the year around $15B, spiked to over $300B during Q4 and declined substantially during crypto winter, leveling off around $60–70B in late 2018 and early 2019. It never touched anything close to $15B again. A key indicator for AMPL will be leveling out at a higher network value following demand spikes and subsequent declines.
Only time will tell how resilient Ampleforth proves to be. The long run success or failure will be a function of the usefulness of the protocol and strength of the community. Speculation may provide a window into the market’s views on that usefulness, but the short term volatility in the network value is also a bit of a distraction.
Gauntlet Network’s recent August 2020 research report on Ampleforth analyzed the trading period of Ampleforth between January 3, 2020 — June 22, 2020, while also running a simulation that measured the effect of different short-term trading strategies on the Ampleforth system. The simulation employed a hypothetical distribution of various market participants such as arbitrage traders, equilibrium traders, mean reversion traders and momentum traders. The headline finding was that the Rebase Arbitrage Trader’s strategy appeared to be the only modeled trader type that demonstrated significant profitability. It will be interesting to reestimate the simulation on data from the second half of 2020.
AMPL’s Rising Economic Utility
Neither short-term incentives nor short-term speculation is sustainable in and of itself if the asset offers no economic utility. To have long term sustainability, the Ampleforth protocol (and AMPL unit) must be a compelling economic product, and the compelling-ness of the product must result in a sticky community of users.
In the years since Bitcoin launched, the network’s open, immutable, transparent ledger — working in concert with provable scarcity — has reshaped how individuals and institutions across the world think about money and digital value.
However, as it relates to fiat money (e.g. USD in 2020), paper claims on commodity-money (e.g. USD’s “gold standard” pre-1971), or commodity-monies themselves (e.g. gold coins), sovereign digital money is still a fairly young monetary experiment.
A number of other cryptoasset networks are also optimizing for the commodity-money use case. By leveraging some of the economic design features of Bitcoin (e.g. fixed supply) while iterating on other features (e.g. privacy, governance, consensus), we’ve seen a number of differentiated experiments create new forms of cypher-money. Supply is simply another dimension of the design space and one that hasn’t been the subject of nearly as much experimentation. This is in part because supply increases negatively impact asset holders if they don’t accrue pro-rata to holders.
When an asset has a fixed supply, the volatility in demand to hold that asset manifests entirely in price. In supply-elastic assets like AMPL, supply expands and contracts to absorb positive and negative demand shocks, allowing price to remain comparably more stable. Importantly, a holder’s wealth (units*price) will still respond to demand shocks, but the volatility is driven into units rather than price. This comes back to the old adage that financial risk is rarely eliminated and more often just shifted somewhere else.
Is it useful to shift volatility from price to units? Perhaps. Consider the three functions of reliable money:
Unit of Account (UoA). “This candy bar costs 1.00 AMPL”
Medium of Exchange (MoE). “Yes, I will gladly accept AMPL for my candy bar”
Store of Value (SoV). “I will put this one AMPL in cold storage until I want to buy a candy bar in ten years”
We note that prior to 1971, the US used gold to regulate the supply of base money dollars, but gold was not very useful as a medium of exchange, and the unit of account was the dollar on top of the gold rather than the gold itself. Furthermore, there were switching costs to move between dollars and gold. As crypto wallet and exchange infrastructure continues to evolve and merge, transitioning between assets is becoming increasingly frictionless. Switching costs are trending towards zero — eventually, it won’t even be visible to the user.
What this allows is the unbundling of the properties of reliable money. A supply-elastic asset may be more useful as a unit of account relative to fixed supply assets, and an asset with the deepest liquidity pool may be best as a medium of exchange. If AMPL can reach a steady state around the target then it would represent the optimal UoA since the target rises with fiat price inflation.
Of the three functions of reliable money, the store of value property is the hardest to predict over long horizons for young experimental assets. Since volatility is driven into units rather than price, the candy bar might still cost 1.00 AMPL in 10 years, but you could open up your wallet to find that you either have 0.01 AMPL or 500 AMPL. Instead of putting your one unit of AMPL into cold storage, you would buy a forward contract on AMPL, shifting the risk onto an investor that wishes to bear the market cap volatility.
To prove out the SoV property, AMPL needs to reach a steady state where it reliably trades within the equilibrium band over long horizons, perhaps only periodically bumping outside in response to shifts in genuine usage demand (not speculation). For example, say that a popular app adds AMPL as an accepted payment and it drives demand for AMPL. Ideally, the price would rise outside the band in response to the demand, generate the new units to satisfy the demand, and then return to equilibrium.
This steady state is not likely to materialize in the presence of substantial speculative trading activity and will instead be driven by integrations of AMPL that enable non-speculative usage. Additionally, it’s a psychological phenomenon — AMPL reaches steady state only when the market believes it has reached steady state. Steady state will be characterized by consistent buying pressure just below the target and selling pressure just above the target, within the equilibrium band. Investors that are holding AMPL long term are betting that the steady state will eventually materialize, and that it will materialize at a high network value.
While it’s still early days, AMPL appears to be generally working as intended:
- Supply Elastic. The protocol has been expanding and contracting AMPL supply for the last year without pause.
- Low Price Volatility. Outside of a period of exceptionally low demand in Sept-Oct 2019 and a period of exceptionally high demand in July 2020, AMPL has continued to oscillate between $0.50 and $1.50. This is still a bit outside the equilibrium band, but it’s proven to be comparatively price stable relative to fixed supply assets.
- Uncorrelated. Gauntlet Network’s August 2020 report concluded that on a market capitalization basis AMPL’s “historical returns have been uncorrelated from both BTC and ETH across various timescales.” We’ll be watching this closely to see if it maintains a low correlation with other cryptoassets and the Ampleforth network value grows.
Importantly, AMPL (the unit) reaching steady state as a base money must also coincide with Ampleforth (the system) reaching a point of credible neutrality. Over the coming years, its decentralization will be graded by the market across many qualities — for example, protocol changes, oracle reporting, and ongoing development around the project.
Some projects striving toward progressive decentralization are solving these challenges through the construction of community-powered governance processes. For other projects, the protocol’s reliability over time may better align with removing human coordination wholesale. Last week the Ampleforth team announced that the protocol’s “emergency pause” abilities — setRebasePaused and setTokenPaused — had been removed from the protocol contracts entirely. Early indications are that the team appears committed to a minimized governance, rules-based approach as the system evolves.
As we’ve learned with all types of money — shells, stones, gold, fiat — durability is predicated on market psychology. The question is not only whether Ampleforth has created a compelling economic product, but whether the compelling-ness of AMPL as a commodity-money is perceived and used as such in an economy over time.
To complicate the money narrative even further — it’s possible that our legacy construct for reliable money (MoE, UoA, SoV) doesn’t end up mapping as cleanly on-chain as it has off. Today, we find ourselves exploring a new world where any asset can be digitally wrapped and programmed to taste, giving us the power to exchange, price, and store value with minimal friction and switching costs. The functions of money we’ve come to rely on may end up becoming unbundled and rebundled in alien forms difficult for us to imagine today.
We welcome this monetary experimentation.
A special thanks to Evan Kuo, Spencer Noon, Paul Veradittakit, Curtis Spencer, and the Collaborative Fund team for their feedback during the construction of this piece.
Disclosure: At the time of publication, Collab+Currency has exposure to some of the assets described in this piece, including BTC, ETH, AMPL, and USD. We do not own any gold or cows.
Stephen McKeon (@sbmckeon)
Derek Schloss (@derekedws)
Collab+Currency (@Collab_Currency) // collabcurrency.com
