4 years ago at Devcon, I presented the possibility of deleveraging spirals happening in stablecoins.
5 months later, one happened in Dai. This was the first major test of Dai and triggered Dai’s evolution to incorporate exchangeability with USDC and RWA backing.
🧵👇
The Terra UST stablecoin could collapse in a bank run effect. UST is backed by an endogenous collateral Luna. Current Luna marketcap has fallen to arguably < outstanding UST. We are now in a dangerous spiral: as users panic out of UST, this reinforces the Luna crash further.
Iron, an under-collateralized-by-design stablecoin, is experiencing a bank run on its algorithmic/Titan portion, similarly to PegsUSD and Dollar Protocol before it. And similar to how Terra had a scare last month. And also a bit similar to Malt just last week
In our research, we've consistently warned about risks in endogenous collateral, like what is materializing now. It's why I'm highly skeptical of seigniorage shares designs for a primary mechanism.
Read more in our paper here
👉
The only truly innovative stablecoin protocols that are trying new designs (not
@LiquityProtocol
for LST) that I found so far are :
-
@0xDYAD
-
@GyroStable
-
@InverseFinance
Let me know if you know other protocols that are worth adding to this list.
Check out my work on modeling noncustodial
#stablecoins
"(In)Stability for the Blockchain: Deleveraging Spirals and Stablecoin Attacks". A foundational step in a scientific approach to stablecoin design. I will present at
#CryptoValley
#CVC19
June 25.
The Terra design is based on endogenous collateral, or seigniorage shares: the value of Luna derives in a ~self-fulfilling way from the anticipated usage of UST. While this brought benefits on the upside, it is now materializing in dangerous spirals on the downside.
Proud to unveil Gyro Proto, the first live version of Gyroscope.
Gyro Proto puts into production much of what we've been building over the last two years, including several innovations that we expect to persist as new staples in the space.
Test it out now on Polygon 👇
Gyro Proto, a guarded version of the Gyroscope stablecoin, is now live for testing on Polygon.
It showcases groundbreaking innovations in redemption mechanisms, liquidity concentration via
@Balancer
, and risk control.
Test it out:
This is exactly why stablecoins need a risk control policy like
@GyroStable
's DSM that manages limited liquidity when reserves are illiquid.
DAI's exposure to RWAs subjects Maker to similar potential risks. This is why Maker has been interested in Gyroscope's DSM.
*Weekend reading*: how to understand the 🔥 space of algo stablecoins? 🤖 What is going to work and what won't (and what already hasn't)? ⚠️
👉 We got you covered:
👀 Gyroscope’s E-CLP was the most capital efficient wstETH/wETH pool on Ethereum the past 24h, with volume/TVL ~40%.
This E-CLP also enjoys increased LST yield exposure (currently 20% above Stableswap) due to the E-CLP's asymmetric liquidity, which minimizes idle ETH.
Gyroscope is growing. And in a time when many other DeFi protocols are evaporating.
Currently driven by the E-CLP, which is one of the components that we designed to unlock Gyroscope's resilient stablecoin.
Watch out for the complete debut on Ethereum soon 👀
It’s been time for something new for a while.
Gyroscope provides the new tools needed for DeFi projects to work together in building a more resilient financial system.
Gyroscope is a stablecoin but also a risk control layer that tackles a problem that applies across DeFi.
The Gyroscope stablecoin GYD is now live.
3+ years in the making, GYD introduces fundamental stablecoin innovations that put risk control on autopilot.
As I move on from my PhD this final semester, I'm very excited to announce our capital raise for developing Gyroscope.
Gyroscope is about building the vision of a decentralized stablecoin system that is as resilient as possible to the full spectrum of risks
Most decentralized stablecoins today (Dai, Rai, LUSD) have a weak link in using some form of trusted oracle.
Gyro Proto is the only live stablecoin that substantively addresses this imo
For
@GyroStable
, we built the Consolidated Price Feed (CPF) to handle this problem 🧵👇
Gyroscope's E-CLP is performing at its peak today despite the wider DeFi turmoil.
The stMATIC E-CLP here holding strong at:
- 156% utilization
- Almost 2x volume of the much larger boosted pool
- 29x volume/TVL of the larger boosted pool
Let me rephrase that for you: “After getting caught with pants down on Terra, Frax decides to adopt Gyroscope.”
After all,
@GyroStable
has always been the all-weather stablecoin design. Even before endogenous collateral developed into a household name. 🙃
Keep your eyes on Gyroscope. It's only the beginning and GYD is nearly here.
72% utilization of 31m TVL now, with pools as high as 450%.
Here's why the E-CLP outperforms in one graphic 👇
Look forward to GYD outperformance next 👀
Showing the power of E-CLPs: Gyroscope's TVL and volume at all time highs today.
- 69% total utilization of the 31.8m TVL
- 172% utilization for cbETH E-CLP
- 184% utilization for wstETH E-CLP on Optimism
- 415% utilization for USDC-USDT E-CLP on Arbitrum
And GYD imminent 👀
New paper, new tools for modeling
#stablecoins
and understanding and preventing crises. Probabilistic guarantees on stability and characterization of instability. To mitigate crises: USDC is one solution, we present another
Read our piece on governance extractable value (GEV) in DeFi. Rug pulls, governance attacks, and short-termism. And how network governance can be designed safely, including timelocks and optimistic approval.
w/
@Lsquaredleland
in this week's
@OurNetwork__
A good breakdown of the risks of
@ethena_labs
upcoming stablecoin.
All stablecoins have risks. And new stablecoins like Ethena feature new risks.
This is where we see
@GyroStable
as the missing piece: the infrastructure that manages the different risks of a diverse asset base.
With apologies to my friend
@nic__carter
and
@CryptoHayes
, this is not a great design for long-term stability.
So first, let's talk about the mechanics of what is going on with Ethena.
In a traditional crypto-backed stablecoin, what you have is a pile of collateral that is…
Preempting DAI's USDC problems is why we spent the last year building a more resilient "PSM 2.0".
The primary market (P-AMM) we built for
@GyroStable
is designed from first principles to autonomously govern risk control and stablecoin monetary policy.
🧵👇
The R/sDAI E-CLP is hitting its stride, showcasing the power of asymmetric concentrated liquidity.
Now doing 11x the volume of the larger stableswap pool while earning sDAI yield for LPs.
Check the date and content for this paper.
Maker is awesome, but there are many ways it can be improved. And you need to have done the research to get there.
That's how we've spent the last several years leading up to
@GyroStable
@cburniske
@MakerDAO
@GyroStable
Imo this does a pretty big disservice to maker, comparing them to a project that’s like 10 minutes old. Yes I know they’ve been on testnet for a year or something but seriously that’s not even close to the lindy of DAI.
Gyroscope's stMATIC E-CLP today is doing > 100% utilization (volume/TVL).
It's now handling more volume than the largest stMATIC pools, despite having just 4% of the TVL and earning 2x the fee rate for LPs.
It is becoming the go-to source for stMATIC liquidity.
This E-CLP on zkEVM showcases just how much a customizable asymmetric liquidity profile can increase capital efficiency in the specific case of rETH.
Coming in here at 74% improvement over a stableswap pool.
The E-CLP's customizability is a unique edge for rETH liquidity.
rETH can trade significantly above peg due to the bonding requirement of RPL. This makes high amplification stableswap pools unsuitable.
The E-CLP enables a unique combination of concentration and tail liquidity.
One reason why DeFi protocols really should use something like
@GyroStable
's Consolidated Price Feed, which validates consistency of pricing on-chain.
While Chainlink is usually dependable, issues happen, and mechanisms should be as resilient as possible to such corner cases.
Earlier today, Chainlink wstETH/ETH price feed on Arbitrum reported an inaccurate value, resulting in the liquidation of 5 positions.
Our friends at Chainlink are on it.
No bad debt accrued, Silo is performing as usual ✅
🧵
We were an early adopter of this idea 2 years ago using Coinbase OAuth verifications for
@GyroStable
's Sybil challenge game.
Was an important addition to making Gyroscope's Sybil-resistant founding member NFTs.
Exciting to see this made much more accessible now!
Introducing Coinbase verifications – Users can attest to account and country credentials onchain, all verified by Coinbase, now live for our 100M+ users at
Gyroscope's E-CLPs on
@Balancer
boost efficiency on several fronts.
In addition to asymmetric concentrated liquidity, E-CLPs build in rate providers.
Rate providers make yield-bearing asset pools more efficient by:
- automating liquidity management
- mitigating LVR
.
@GyroStable
marks the first innovation in scalable reserve backed stablecoin designs since
@feiprotocol
in 2021.
Fei was, unfortunately, a disaster. It was rushed to market without proper design or testing and with problems that we predicted.
Gyroscope is a stark contrast.
The issues with Fei were predictable, and we wrote about them ahead-of-time. While there are interesting parts that could work around reserve-backed designs, getting the design wrong can mean the difference between Venezuelan bolivar and USD.
Another casualty of today is on-chain stablecoin liquidity
The current market structure is not designed to be resilient to a major stablecoin depeg. Even for relatively unaffected stablecoins.
🧵👇
Finally caught up on crypto twitter 😅 A few thoughts on this AMM debate:
1) The math itself seems plausible based on volatility harvesting results (though I haven't checked the details). The interpretation is a little optimistic though imo. 1/7
This is mathematically wrong but in a pretty subtle way. It's not really going to fit into a tweet but
@danrobinson
I'd be happy to have a call/chat/podcast/debate/etc. with you on it :)
Today's stablecoin space is evolving into one of the most bullish scenarios for
@GyroStable
.
The more diverse the stablecoin asset space becomes, the stronger Gyroscope grows.
Gyroscope = the ultimate stablecoin protocol that balances the risks of other assets.
A good breakdown of the risks of
@ethena_labs
upcoming stablecoin.
All stablecoins have risks. And new stablecoins like Ethena feature new risks.
This is where we see
@GyroStable
as the missing piece: the infrastructure that manages the different risks of a diverse asset base.
The Next Generation: Crosschain Yield Optimization with Beefy x LayerZero x Aura
Every once in a while, an innovation comes along that changes the way the game is played. No, we're not talking world-ending or life-changing... that may be a tad dramatic! But we are talking about…
Learn more about the Gyroscope PAMM in our recent talk, or join our upcoming presentation at SBC.
A full paper and the implementation will also be released soon with the Gyroscope launch.
Gyroscope's stMATIC E-CLP today is doing > 100% utilization (volume/TVL).
It's now handling more volume than the largest stMATIC pools, despite having just 4% of the TVL and earning 2x the fee rate for LPs.
It is becoming the go-to source for stMATIC liquidity.
1/ Without worrying about aggregator integrations, liquidity swap routes, or new incentive structures,
@GyroStable
has launched a custom Elliptical concentrated liquidity AMM that has facilitated nearly $100 million in swap volume.
Here's everything you need to know.
A 🧵
15/15 How to fix the issues that arose this weekend and build more all-around resilient mechanisms?
For some possible routes, check out the mechanisms that we designed for Gyroscope
Gyroscope was designed for a day like today
- Built-in diversification of reserve assets
- 'PSM 2.0' (DSM) adapts autonomously to maintain sustainability
- CPF oracle handles corner cases of pricing assets, incl USDC
- Resilient trading network for stablecoin liquidity
🧵👇
Today Gyroscope’s stMATIC E-CLP is doing 9x volume compared to the much larger Balancer boosted pool.
It's also doing 222x volume/TVL, a better measure of pool performance as it shows how well the liquidity is being used (and so generating fees for LPs).
Gyroscope’s technology unlocks a new kind of concentrated liquidity.
A previously deployed pool on Polygon shows that Gyroscope’s concentrated liquidity pools just *work*👇
ICYMI our talk on
#Stablecoins
2.0 from ACM AFT is online. In ~15mins,
@ljfgudgeon
and I give the general flavor of our framework for understanding design choices and trade-offs in stablecoins, including models for security and stability.
Is the system underwater? It's hard to say precisely. On paper, Luna FDNV is still ~2x the 2b UST supply, but measures of circulating marketcap are only ~0.8x. Even the optimistic 2x number would be very unhealthy though given the spiral effect on endogenous collateral.
This is why we need a decentralized stablecoin that controls for existential risks. Like geographic, regulatory, custody risks of centralized tokens
Perfect time to check out
@GyroStable
Proto, now live
New York regulators directed a crypto company to stop issuing one of the largest dollar-pegged cryptocurrencies as a government clampdown on the sector widens
And that's why it's critical to diversify across even stablecoins like USDC.
And to build in contingency mechanisms like the DSM
And to build liquidity structure that can survive an individual stablecoin depeg
All things we factored into designing
@GyroStable
Silicon Valley Bank is one of six banking partners Circle uses for managing the ~25% portion of USDC reserves held in cash. While we await clarity on how the FDIC receivership of SVB will impact its depositors, Circle & USDC continue to operate normally.
The Gyro design has now undergone extensive audits. We recently talked about how it works at
@IFCA_Conference
DeFi Workshop.
Stay tuned for the recorded talk and full paper soon, with the full Gyroscope system being deployed this summer 👀
#CropRotation
: Since the new governance change, the most profitable $COMP farming is now to borrow and re-deposit
#Dai
. With multiple layers of borrow-redeposit, you can get >40% yield still. The $BAT positions have massively unwound (w/o incident!) and replaced by Dai
GYD E-CLPs have expanded to 6.9m TVL with the pools now live on Aura ✨
E-CLPs provide deep liquidity for GYD by design and set up the system to accommodate rapid growth with new campaigns coming in January 👀
Boosted incentives of up to 65% APR from
@Balancer
and
@AuraFinance
are now live on GYD E-CLPs ✨
It's simple to LP:
1. Swap any token -> GYD on 1inch
2. Add liquidity to one of the GYD pools at . You need both tokens.
3. Stake in the Aura gauge.
Actually the PSM *isn't* working smoothly, though it is working as expected
The 1:1 nature of the PSM means that there was a run *into* the USDC PSM and a run *out* of the USDP PSM. At non-market prices to the detriment of Maker.
PSM working smoothly as expected with reserves standing at $3.15 billion backing 3.15 billion DAI at 100% collateralization.
PSM's code is law for DAI stability.
Just check Maker Protocol's data, transparently, in real-time.
→
This is how we've built
@GyroStable
, to raise the bar for sound stablecoin design.
Read more in the Gyroscope docs and join us as we roll out the Gyro system this summer
Decentralized economies deserve decentralized money -- but we as the crypto community need to raise the bar.
It's possible to make decentralized stablecoins that are both truly decentralized and economically sound.
Another flawed design for a stablecoin has failed and I’d like to explain why. This one is essentially the next iteration of basis/esd/esdv2/nubits so really this isn’t too surprising 🙃
Two under the radar but steady performing
@GyroStable
pools to highlight today
1) The 3-CLP (BUSD/USDC/USDT) casually doing 113% utilization today.
The 3-CLP is a unique pool type that concentrates liquidity among three assets, improving capital efficiency 50% vs Uni v3 pools
Tightness around peg for a stablecoin is the wrong metric for tail risk. Sustainability of that tightness is more important, which relies on whatever is (explicitly or otherwise) backing it. It's possible to maintain a peg up until there's no value left to maintain it with
Gradually, then suddenly.
This is a hidden problem, understood by few, that will someday come to bite: how should a stablecoin (e.g. Dai) react in the contingency that something happens to PSM assets?
This is why we built the Dynamic Stability Mechanism (DSM) in
@GyroStable
Opened a discussion about adding a redemption curve to the PSMs to improve the resiliency of Maker.
This will help to establish a market price for DAI in the event of a big loss. An important step now that we have multiple large RWA vaults.
When your paper on decentralized finance for a decentralized finance workshop is rejected for "assum[ing] too much financial knowledge"... much of DeFi in a nutshell, unfortunately 🤣
E-CLP for MATICX doing over 100% volume/TVL today on
@GyroStable
And key E-CLPs for wstETH and stMATIC also handling high volumes well, serving as go-to routes for 1inch order routing on Optimism and Polygon
Hopefully by now, this possibility shouldn't come as too surprising. Read more about the mechanics in the bit on Frax, which can face similar issues, and seigniorage shares here 👇
And more details in our papers
👉
Our paper detailing economic security issues like this should be required reading at this point.
👉
DeFi projects can be designed with these problems in mind.
This attack was qualitatively different than others because the hacker risked a lot of their own capital over several blocks with a few different failure modes.
They made a cold, rational calculation they could break the economic security of the protocol. And they were right.
.
@raft_fi
R/DAI is yet again the highest 24h volume pool on
@Balancer
Not for long though...
the new R/sDAI (!) Gyroscope pools with concentrated liquidity will launch soon
I hope this is the end of 'FOMO dropping' of 'fair launch' projects. The first few may have been innovative, but at this point it's only a hype strategy and not really 'fair'. If long-term communities do develop, it is *despite* this strategy.
What does Gyroscope do differently?
1. Reserve structure limits exposure to different existential risks ahead-of-time
2. The 'PSM 2.0' (DSM) adapts autonomously to maintain sustainable short-term liquidity
This buys time to make gradual changes over time as risks change
TLDR DeFi research program, supported by the Uniswap Foundation, was announced last week and is open for grant proposals until Aug 15.
TLDR = The Latest in DeFi Research
Check out if you're interested in governance design, MEV, and DeFi mechanism design research.
In our paper on Decentralized Governance of Stablecoins, we modeled the principal-agent problem between governance layers and users of DeFi protocols.
👉
This work was presented by
@Mads_Rude_Wind
and Frederik Münter the other week at MARBLE.
🧵👇
Nice read on exploring stablecoins using the lens of banking. While stablecoins aren't exactly "banks", there is a lot of knowledge to draw on.
We shared many similar thoughts in designing
@GyroStable
. It's nice to see our work on redemption curve design featured in the article.
Our team contributes to
@MakerDAO
, implementing >$2bn of RWA generating >$150M of annual revenues. We also advise
@VenusProtocol
@AngleProtocol
We know a thing or two about stablecoins
Today, we release our operating manual for decentralized stablecoins
Our concept of Governance Extractable Value (GEV) is a next frontier in DeFi exploits. To dive deeper:
Our intro article
👉
Our DeFi SoK
👉
First steps in modeling GEV in our Stablecoins 2.0 paper
👉
🎙️"There are many DAOs out there with billion-dollar treasuries. I think we're going to start to see sneaky attacks on governance that really could put some of these systems in jeopardy. -
@danrobinson
After MEV, is Governance Extractable Value next?
A neat thing about GYD E-CLPs is that they're designed for dual functionality:
1) Provide deep GYD liquidity
2) Connect volume for other token pairs through GYD as a central asset
Both already work. See this chart for cross-volume being routed via GYD over the first week
We built
@GyroStable
’s E-CLP (elliptic concentrated liquidity pool) as a new pool type on Balancer.
The E-CLP was designed to optimize liquidity for the Gyroscope stablecoin (launching soon) and is already outperforming liquidity in LST markets
10/ So, what's been built so far?
Harnessing this streamlined tech stack,
@GyroStable
built a truly unique concentrated liquidity model tailored for Liquid Staked Tokens.
With an instantaneous plug into vault liquidity, the Pool facilitated $4m+ in volume in just one month!
There are many directions and challenges worth exploring in these stablecoins.
One such direction we are building in
@GyroStable
is a new mint/redeem mechanism (the stablecoin 'primary market'), which is designed to gracefully and transparently unwind in the face of lower demand
$COMP farming with $BAT looking pretty curious. Let's consider how it got here: likely some large
#Compound
depositors realized that they can skew the COMP distribution toward themselves by over-utilizing a smaller market than $USDT. And perhaps similar happening to $ZRX
Kinda funny seeing responses saying it won't happen.
Of course a stETH depeg will occur at some point with high likelihood. This is finance, and extreme events happen more often than people think.
Ultrashort duration ETFs 'depegged' against NAV and liquidity dried up in US…
And now $3,000,000,000 of leveraged stETH on Aave that was "very liquid" is relying on $300mn of Curve liquidity; there is still virtually no off-chain liquidity.
Especially with volatility from ETF news, if any large stETH holder gets impatient and sells, things could get ugly
📣 DeFi researchers: submit your work to the 2nd DeFi Workshop
@IFCA_Conference
by Dec 5!
Last year's workshop was a huge success with great speakers and papers all around. DeFi research has grown a lot since! This year should be even better in person.
I'll be presenting our work exploring alternative oracle designs today at MARBLE.
This paper explores using alternative on-chain state data to sense check the consistency of oracle-reported prices.
Can anyone help us get in touch with
@SnapshotLabs
? The level of engagement (>5k) on our
@GyroStable
proposal seems to have just broken their vote counting. Seems Gyroscope is really stress testing DeFi infrastructure 😂
Extremely high engagement on Gyro Level 3 (!) is causing trouble with , and votes are no longer able to be counted. We are trying to get in touch with the
@SnapshotLabs
team to get this fixed ASAP
Stablecoins have come a long way since I started covering them.
There have been a number of recent innovations that expands the quality of stablecoin offerings today.
But a gap has remained in building resilient reserve-backed mechanisms, which
@GyroStable
is set to fill.
The stablecoin space is finally evolving ✨
Early 2020: USDT and USDC were the only real options, and DAI was a minor player.
2021-22: new entrants were mainly ponzi schemes that collapsed.
Fast forward to 2023: there's a wave of new innovations and safer stablecoins.
🧵👇
My work on stablecoin deleveraging spirals ---which materialized in the Black Thursday crisis in Dai--- received Honorable Mention for Best Finance Student Paper
@INFORMs2021
Watch the presentation here
We continued this work into what is probably the first DeFi paper published in a top journal like Mathematical Finance.
The expanded work formally characterized deleveraging spirals as a region of higher volatility with a submartingale ( = expected to increase ) price process.
4 years ago at Devcon, I presented the possibility of deleveraging spirals happening in stablecoins.
5 months later, one happened in Dai. This was the first major test of Dai and triggered Dai’s evolution to incorporate exchangeability with USDC and RWA backing.
🧵👇
There is room for innovation in building autonomous asset-backed stablecoins, which more resemble non-custodial forms of the commercial bank monetary system.
We discuss this in our Stablecoins 2.0 paper in relation to Dai and similar stablecoins.
👉
Still highly skeptical of seigniorage shares stablecoins. It's like if for every $1 you deposited at your bank, $0.15 was skimmed off and paid out to bank shareholders as buybacks. There's no good reason to do that. That capital can be better used to maintain stability
Purely in the interest of being fair:
- Gyro Proto has been live since Jan 2023
- Gyroscope was devised before Angle (or Fei!) and I'm pretty sure was the motivation for Angle in the first place.
Not that it really matters. Angle's cool too and let's build stuff together now!
I like
@GyroStable
's design but I can't agree with this one
@aklamun
@AngleProtocol
Transmuter system is the first reserve backed stablecoin system that guarantees a claim to all participants on all the collateral backing the stablecoin.
Launched in July 23😉
In contrast, we designed Gyroscope coming from experience in evaluating stablecoin design risks in DAI, Terra, and more.
We describe the issues that plagued Fei in this thread, as well as how Gyroscope was designed with foresight on these topics 👇
Amazing how prescient this was. Let's dig in. Fei's liquidity has been so bad b/c the supply needed to contract to reach equilibrium with demand. But this led to depegging b/c the supply literally can't contract in the Fei design
Gyroscope's resilient oracle system (the Consolidated Price Feed) has obvious utility once you know how it works. It's something sorely missing in DeFi security.
It seems inevitable that future DeFi protocols will be shaped around using the new information it provides.
At
@EthereumZurich
?
Gyroscope dev team member
@sschuldenzucker
will speak abt building resilient oracles at the example of Gyroscope’s Consolidated Price Feed.
⏰Time: Apr 16, 9:30AM
📍Location: Execution Stage
🗒️Details:
Registration is still open for the 2021 IC3 Blockchain Camp (virtual, July 25-31)!
Featuring talks by
@VitalikButerin
,
@el33th4xor
,
@stonecoldpat0
,
@SergeyNazarov
, and many more (open to all) and a hackathon (IC3 members)
Registration and details at