# 27 | Meet Abracadabra's Money Team, It's a Small One Anyway
Give Me a Lever Long Enough and a Fulcrum on Which to Place It, and I Shall Move the World
You can’t always get what you want But if you try sometimes You’ll find you get what you need
Nature listens to the Rolling Stones. The DeFi we have today is the DeFi we deserve. But everything changes, species adapt, and what used to work in a certain ecosystem won’t necessarily thrive in the next. Dinosaurs dominated the Earth for 150 million years until the Cretaceous-Paleogene extinction event almost wiped them all out. Interestingly modern birds, small and light, are among the closest relatives of the gigantic creatures of our collective dreams. Nature finds a way. It is my belief that most of the protocols that have recently surged in value (being it locked, staked, or recognised through governance token valuations) won’t be around for much longer as investors’ needs change. Some will successfully pivot into something else the same way IBM moved from time recorders to hardware and ultimately to consulting, while others will succumb.
Who, among today’s most relevant projects, will survive the DeFi version of the Cretaceous disaster? The updated list of the top 15 protocols by Total Value Locked (source Defi Llama) is dominated by AMMs - 5/15, and includes only 3/5 stablecoin-based leverage machines: MakerDAO, Anchor, and Abracadabra. Each one couldn’t be more different from the other.
Who will be the winner in the ruthless race of evolution by natural selection? MakerDAO, the old and conservative grandfather? Or Anchor, the shiny and profitable gate to the lands of Terra Money? Or Abracadabra, the ludicrously successful enfant prodige? Yet again, we might be surprised. The only thing we can do is to look at what we have in front of our eyes, and account for a fairly thick buffer to deal with the unknown unknowns.
What Is Abracadabra?
Giving degens the ability to leverage is hardly a distinctive factor in today’s DeFi ecosystem. How, then, did Abracadabra manage to reach > USD 5b of TVL (you can surf the updated dashboard here) in the span of few months? The question is relevant, and requires a multi-faceted answer.
In its simplest description Abracadabra is a wallet-agnostic | flexible | proprietary | yield-focused | margin lending | amusement park. Let’s unpack.
Wallet-agnostic → Money supermarkets like Compound and AAVE have a wallet-holistic view of credit risk. In other words, users can pledge several white-listed crypto-assets as collateral, the value of those assets is translated into collateral value (multiplying it by a factor) and aggregated on a wallet basis. The aggregated collateral constitutes a guarantee against which wallets borrow other assets. If there’s some sort of diversification across the collaterals - hint there isn’t much, that should benefit creditors and borrowers. In the case of Abracadabra instead, Sushi’s Kashi Lending Tech is used to provide isolated lending markets that allow users to isolate their exposure for each single underlying collateral. Although it doesn’t make necessarily leverage more robust, it makes it more digestible.
Flexible → Within single pools users can choose many things: (i) how much of such asset to provide as collateral, (ii) the borrowing amount - i.e. LTV, (iii) the leverage amount - i.e. the turns of leverage if users decide to lever their yield up - more on this later.
Proprietary → Differently from Compound of AAVE, that operate a proper money supermarket model where users can borrow and lend certain whitelisted assets, de facto tweaking their long and short relative exposure to those, Abracadabra lends uniquely their own proprietary token, $MIM - or Magic Internet Money. $MIM is a promissory stablecoin that aims at keeping a 1:1 peg to the USD - more on this later.
Yield-focused → Most of Abracadabra’s lending pools (or at least those that constituted its original focus) are accepting yield-generating strategies - wrapped in a fungible token. $yvcrvSTETH is a good example. The token represents a Yearn Finance strategy where the protocol supplies $crvSTETH to Convex Finance to earn $CRV and $CVX; those tokens are harvested, sold for more $crvSTETH which is deposited back into the strategy. $crvSTETH is a liquidity pool deposit token that is obtained when $stETH (a receipt of $ETH deposited in liquid staking protocol Lido) and $ETH is deposited in a liquidity pool on Curve. If your head is spinning you are not alone. Hopefully the chart below helps you following the flow.
There is a lot of stuff going on - and a lot of sources of risk added on top of each other. Luckily, at least for what concerns this specific strategy, it doesn’t seem that a lot of leverage has been added to the picture in each step: the yield seems composed by a bit of structural farming - in this case the participation to Ethereum's PoS consensus mechanism, some liquidity pooling (providing the liquidity for $stETH to be swapped back to $ETH needed until the Beacon Chain allows us to natively unstake) and some marginally dilutive farming yield that large protocols like Lido, Curve, and Convex are offering. The sustainability of those yield strategies, however, will require a set of dedicated posts and goes way beyond the scope of this one. The market is deep and liquid and the participation incentives are not as crucial as before to attract users; as a consequence, this specific combined Yearn strategy ($yvcrvSTETH) delivers c. 3.4% on top of Curve’s fees and Ethereum’s staking incentives - I believe the total yield is somewhere around 12-13%.
$MIM is borrowed against the appreciating token and this offers some sort of buffer against token volatility and hence liquidation that, in the case of borrowing against stables, might be sufficient to sustain very high levels of leverage. The conditional is important. Currently, however, 65.5% of total borrowing is constituted by non-stablecoin vaults.
Interestingly, $UST currently represents the largest pool (USD 550m TVL) on Abracadabra. The native yield on $UST achievable through depositing on Terra’s Anchor protocol is so high (c. 19.5% currently) that no yield transformation through smart farming protocols like Yearn or Convex is needed before pledging it to Abracadabra to spice that yield up a bit.
The natural next step from providing over-collateralised leverage on yield strategies is to offer levered versions of such yielding strategies directly. The flow chart below exemplifies how this works for Abracadabra. Individuals access a (whitelisted) yielding strategy (e.g. a $USDT farm on Yearn), they provide the receipt token (e.g. $yvUSDT) to Abracadabra’s pool choosing how many turns of leverage they want to get. The pool automatises the following steps, i.e. getting $MIM, swapping those $MIMs for $USDT in a Curve pool, and re-depositing those $USDT into the Yearn pool and so on until the desired leverage is achieved. Expected levered yield, and leverage, can go up by a significant quantum (e.g. 5-10x) but the tolerance buffer gets proportionally slimmer.
Margin lending → Although it is mainly against yield-generating strategies/ assets, Abracadabra ultimately still offers margin lending to customers. This means that a borrow fee and an interest is charged (currently 2.5% on $UST) and that the collateral is continuously marked-to-market and at risk of liquidation - such risk is commensurate to the level of collateralisation chosen by the user. In order to incentivise liquidators to do their work, those liquidators have the ability to purchase at discount (5% currently for $UST) any collateral flagged for liquidation. Such liquidation discount is tailored parametrically by governance in order to attract liquidators on the platform. But there is an interesting question I am asking myself: in the case of stables like $UST, that are structurally constructed to keep a peg vs. a fiat currency, a (sustained) pegging diversion as big as 5% could be absolutely destructive for the whole construct, and in that case what would be the incentive for a liquidator to sweep the position - and protect the $MIM? I guess I will have to wait for one of those situations to happen.
Amusement park → Let’s face it, Abracadabra is fun. You don’t borrow dollars, you borrow Magic Internet Money. And you don’t actually borrow $MIM but rather you summon it. Staking receipt tokens of subsidised strategies (i.e. Sushi’s $ETH-$SPELL liquidity pool or Curve's $MIM + 3Crv liquidity pool) provides a governance token called a Spell. Abracadabra’s UX as well, is definitely not intimidating. Do you have in mind the several hundred of font 1 lines you have to read and undersign when you get a credit card with 500 USD monthly limit? Forget it, here you can borrow USD 10m at 10x over an already levered yield strategy and it will be all pixelated coins and cute magicians and warm blues. Today’s average loan on Abracadabra is USD 100k. It’s not Wonga. We cannot complain if the grey-haired regulators don’t like it a single bit.
U.S. Sen. Sherrod Brown (D-OH), Chairman of the U.S. Senate Committee on Banking, Housing, and Urban Affairs. December 14, 2021.
In other words, stablecoins are a particular type of cryptocurrency whose value is managed by a single company. These include Tether, Circle, and Abracadabra – a fast-growing scheme that makes “Magic Internet Money.” Their words, not mine – what could possibly go wrong with something that claims to be “magic money”?
In case you are curious, here’s the video.
Why So Famous?
Based on what we have said so far it is not difficult to understand why Abracadabra is so popular among crypto-native investors. It is fun and easy to use and, at the bottom of it, it allows them to take huge amounts of leverage on top of their yielding strategies. It is possible, for example, to take Anchor’s deposit yield of c. 20% and lever it up through Abracadabra to obtain 100%+ APY. What there’s not to like in getting 100% yield on a strategy that is based on a stable yield of 20% over a currency that is supposed to remain absolutely stable thanks to a promised peg between $UST, $MIM, and the USD? There is a lot to be concerned with in my opinion (e.g. Anchor’s peg sustainability, frictions in the AMMs that provide swap ability between $MIM and other stables, market inefficiencies in general, etc.) but I might be just yet another old dude pretending to understand the new YOLO echoes.
Let’s put those concerns on the side and embrace the degen’s viewpoint for the time being. Abracadabra has been successful, so far, because it focused on solving users’ needs rather than trying to provide a solution for a problem that doesn’t exist yet. DeFi users currently want yield, not risk management, and Abracadabra provides yield by squeezing it from all the pockets it can find. And when I mean users, I point to every user, being a degen or an institutional investor. It won’t be like this for ever (that’s my bet) but currently everyone who engages with DeFi is risk on! Investors either deploy a small portion of their assets (that they are ok with losing) in DeFi, or simply they own so little that for them such a risk is a fair price to pay for the opportunity of becoming rich. Both have their merits, and DeFi has evolved to solve them with a plethora of solutions.
Today’s DeFi is solving for its investors’ needs → I firmly believe that by participating to, rather than invest in, the DeFi ecosystem investors would be able to sustainable benefit from 20%+ yields. DeFi participants, through effort and capital, are substituting financial intermediaries and should demand a compensation that is commensurate to those intermediaries’ profit margin - rather than to their RoE. For context, JP Morgan’s profit margin has been 48.66% for the most recent quarter. But let’s try to make this 20-40% YoY the target return of a DeFi-focused fund raising effort and my gut feeling tells me we would encounter a lot of resistance. LPs (meaning Limited Partners) do not want a great risk-adjusted 20% from DeFi, but the opportunity to make 2-10x per year. They have already accepted the obliteration scenario when they decided to approach. Hence today’s DeFi provides them with the tools to get what they want.
In doing so, Abracadabra has generated hefty fees - USD 39m this year and counting. Most of those fees (interest, borrow fee and 10% of the liquidation fee for certain markets) are deposited in the $SPELL fee pool. Those fees are used for open market operations around the $SPELL tokens - some of those (buybacks) are value accretive, whereas others (farming and similar incentive programs) are dilutive. Staked $SPELL ($sSPELL) tokens get exposure to those forces. Interestingly, only < 15% of those tokens are currently staked. Abracadabra is another protocol that has sought help from Olympus to market their governance tokens and benefit from the liquidity fees - the impact has been so far very limited.
Why So Nervous?
It depends on who you are. Are you a $MIM borrower, a $SPELL holder, a $MIM user? Although all the stakeholders of a project are connected, each one has his or her own type of exposure to the protocol.
Let’s start with $MIM borrowers → Borrowers are holding $MIM-denominated liabilities. What this means is that borrowers are short $MIM - if $MIM devalues in asset terms it takes less for them to pay back. It works the same for a bank’s borrower although we rarely think about it that way. A nominal depreciation of the borrowed asset is good for a borrower: inflation is good, de-peg is also good - usually de-pegging forces tend to smash a currency rather than lift it up. But that’s true only partially, as many of those borrowers are exposed to $MIM on the asset side through some of their investment strategies. DeFi protocols incentivise cross-usage even more than traditional banks. To be fair, Abracadabra-based strategies that are levered through Abracadabra itself are not many: $sSPELL and $SPELL for a total of c. USD 50m out of c. USD 5b. USD/ $MIM 1.3b of $MIM - i.e. a large portion, however, is sitting in a c. USD 3b Curve pool that provides liquidity and arb opportunity for pegging purposes - a pool that is subsidised through token farming. In simple terms, instability in the value of $MIM or $SPELL could destabilise the peg further in a downward spiral hitting many, it’s the nature of algorithmic stables.
In addition, borrowers are those who pay the bills around Abracadabra. Although a $MIM crash might not hurt them too much, it would definitely dry new usage and ultimately put the whole protocol at risk. USD 1.6b have already been liquidated so far by the protocol - since September. USD 1b in the first week of December only. That’s a lot, although not enough to encourage users to calm down and bring their money somewhere else. My guess is that a destructive liquidation event would happen only with a de-pegging event for some of the stables in use. Including the infamous $USDT. There is a lot of volatility within stables these days. Although you don’t see it in the price movements, you see it in the borrowing rate/ short interest. Borrow rates on $USDT on Compound have reached 26.5% at the end of October, and have fluctuated massively for a long while before stabilising around 4% - which is the value Compound incentivises through their bonding curve. The spread between $USDT and $USDC/ $DAI is a good indication of the peg risk, and has ranged around 10% for a while - it’s a lot.
Centralisation → We cannot talk about a web3 business without talking about the degree of centralisation it has. Currently, $MIM price peg mechanism is supported by token minting through a 6/10 multisign composed of:
Poolpi (Yearn)
Leo Cheng (Cream)
Michael (Curve)
Squirrel (Popsicle Finance, Abracadabra)
Danielesesta (Popsicle Finance, Abracadabra, WondelandDAO)
0xMerlin (Abracadabra)
Julien (Stakedao)
C2tp (Convex)
Georgiyxo (Popsicle, Abracadabra, WonderlandDAO)
Sifu (WonderlandDAO)
If we consider WonderlandDAO, Popsicle, and Abracadabra as related parties, this means that 5 of the 6 required votes are already centralised in a controlling team. Make your bets.
What about $SPELL holders → Equity holders are more directly exposed to fluctuations in the value of assets. As a reminder, the fair value of a bank’s equity corresponds to the difference between the fair value of its assets and that of the liabilities. A good measure of what is the risk profile of $SPELL should start by re-marking Abracadabra’s balance sheet. Unfortunately, information, price discovery, market depth, past experiences, are all limited, hence we might have to apply some uneducated guessing.
Based on market values and the table above, and considering outstanding $MIM of c. USD 2.77b, we estimate the following:
Book value of $SPELL: c. USD 2,180M
Fair value of $SPELL: c. USD 1,110M | price-to-book of 51%
Market value of $SPELL: c. USD 1,142M | price-to-book of 53%
I am happy to see that I think somehow similarly to the market. A price-to-book ratio of 53% is almost exactly that of UniCredit, one of Italy’s leading banks. Considering that Abracadabra’s growth prospects are not UniCredit’s, this means that there is quite a lot of risk and dilution embedded in the business.
If $SPELL holders are getting a fair deal, the question that remains is whether the same can be said for $MIM’s. Most of $MIMs sit in liquidity pools to support the exchange of $MIM into other stables. The base rate in Curve’s MIM + 3Crv pool is quasi 0%, although Curve is boosting it to 5-13%. Is that enough? In my mind, it’s not.
