Philip Forte
@0xPhilip
Followers
21,299
Following
113
Media
22
Statuses
274
Explore trending content on Musk Viewer
McDonald
• 1018147 Tweets
الهلال
• 788082 Tweets
Liz Cheney
• 195464 Tweets
AFIP
• 184084 Tweets
Samantha
• 56158 Tweets
#WWERaw
• 55688 Tweets
Lamar
• 52362 Tweets
Baker
• 49714 Tweets
Ravens
• 46173 Tweets
Seth
• 27372 Tweets
ENJOY FRANZZI WITH YIBO
• 22439 Tweets
Chargers
• 19977 Tweets
JISOO TOMMY AMBASSADOR
• 19357 Tweets
ARCA
• 18285 Tweets
Alejandro Toledo
• 16905 Tweets
Buccaneers
• 16851 Tweets
Mavi
• 15544 Tweets
Odebrecht
• 15183 Tweets
Bucs
• 14933 Tweets
Cardinals
• 14413 Tweets
New Brunswick
• 13829 Tweets
Raquel
• 11528 Tweets
Mike Evans
• 11080 Tweets
Noticing the ETH situation starting to unfold yesterday, we spun up 15 concurrent Avellaneda market making instances with varying parameters. ETH moved 7%+ in a 1min candle, but the effect on altcoins was unique as well
As expected, the optimal bid ask spread for MM explodes:
9
28
98
Gleaned some valuable insights on progress in the MM space since 2008 after speaking with Sasha Stoikov (from seminal 2008 Avellaneda-Stoikov paper: "High-frequency trading in a limit order book").
3
11
62
Elixir aims to be one of the top protocols when it comes to rewarding engaged community participants
Potions, potions, and more potions ⚗️
Over 70,000 Potions have been set aside for winners of Ritual & Spells, with much more to come 🧪
Begin your journey—enter the lounge 👇
195
212
439
6
4
45
Had a great chat with the Monad folks. We're excited for Elixir to function alongside Monad's L1 to power orderbook liquidity on the next generation of dexes
Bullish on these guys
Ep 12:
@ElixirProtocol
🟣 Building on Monad
🟣 Unique Value Adds
🟣 Future of DeFi
@0xPhilip
<>
@intern
<>
@opendanny
18
28
91
4
8
39
If crypto exchanges and projects lean on a decentralized primitive allowing retail users to supply liquidity to orderbooks, it’ll need to come to consensus
This introduces latency however: the focus of much of our analysis at
@ElixirProtocol
1/n
3
10
35
Gmonad ⚗️
1/ Welcome to the Monad Ecosystem
@ElixirProtocol
! 💜🧪
Elixir brings deeper liquidity to on-chain order books while democratizing fee incentives previously inaccessible to retail.
Here's how: 👇
29
74
238
4
7
26
Proud to announce the close of our Series A
Our team looks forward to working closely with our new partners in our path to enable widespread trustless, algorithmic market making:
5
3
25
Park stables and earn up to 29.5x potions 👀
The alchemists at Elixir and
@MIM_Spell
have come together to unveil the first staked deUSD Cauldron ⚗️
Abracadabra's engine allows you to lever up your sdeUSD potions by up to 29.5x 🧪
The process is simple 🧵
48
120
345
1
1
20
In case of an investigation by the FBI or similar government agency, I do not have any involvement with this group or with the people in it. I do not know how I am here, probably added by a third party, I do not support any actions by members of this group
7
0
20
Excited to announce our fundraise. We've onboarded strong allies on our journey to transform market making for projects and exchanges
FalconX, Bitmex founder Hayes invest in market maker Elixir amid dearth of providers
2
11
51
1
2
15
@0x_antifragile
Appreciate the feedback. Planning on improving documentation this week (which is lowest hanging fruit)
Signing for Apothecary is done so users dont access other users’ profiles, but there’s likely a way to store cookies in browser for longer. Forwarded to our dev team
1
0
15
We saw this as a couple players instantly widened out bid/ask spreads in altcoins when majors moved, but before the altcoins moved. This prevents a few suboptimal fills, but increases correlation with BTC/ETH
1
0
15
Side note: researchers in high frequency volatility estimation have developed econometric techniques to filter out what they call ‘microstructure noise’ - see (Hansen and Lunde 2006) and (Ait-Sahalia et al. 2011), which I'd definitely check out
1
0
14
Finally, market makers taking into account the volatility of the large caps as a variable in their algorithms & calculation of optimal bid/ask spread. This is IMO one of the leading causes for correlation between cryptoassets & majors
1
0
13
The orderbook clears instantly, as all MM's pulled asks after immediate fills, and widened bid/ask spreads.
Books begin to skew in the other direction (green: ACA, blue: USDT, red: total):
1
0
12
As this spread widens, it exacerbates the documented problems with mid-price data (making it empirically worse as a predictor): changes in the midprice become highly auto-correlated, an effect known as the "bid ask bounce", which we observe
1
1
12
As liquidity retreats on the orderbook, it's similar to a 2 tick adj moving to a 1 tick adj in the below diagram (predictive power of which I broke down in an old tweet). Using a weighted midprice or microprice (as shown here) in helps adjust for these drastic swings in liquidity
1
0
11
Welcome to the Elixir family
@OrderlyNetwork
⚗️
Introducing: Orderly Quantum pools, powered by
@elixir
Our first user-facing product 🔥
Quantum pools will allow you to supply liquidity directly to Orderly pairs, acting as an AMM equivalent for orderbooks.
31
64
353
0
1
10
@thiccythot_
@Uniswap
Good thread but disagree with the conclusion. I think the answer lies in dynamically changing passive LP fees for AMM pairs with higher toxicity. Limiting trade sizes attempts to lock in stale prices on exchange venues: markets will get to their fair value either way
0
0
10
Had a great chat - thanks for having me on!
WuBlockchain Podcast: EP-12 Elixir Is Order-book Style the Future of Perp DEXs?
Youtube
Spotify
19
32
98
2
0
10
Well, Ethereum is officially deflationary. Congrats everyone we did it! Lol
2
0
9
Avellaneda-Stoikov undoubtedly advanced the field of market microstructure, however the strategy in its 2008 form faces clear room for improvement. Using mid price has empirically suboptimal predictive power, for example
1
0
7
Using a "micro price" however: the limit of a sequence of expected mid-prices, and a martingale by construction, HFT firms can construct algorithms with higher performance
1
0
7
Another clear improvement is around order sizing. Optimal High-Frequency Market Making, 2018 (Fushimi, Rojas, Herman) proposes an inventory control model that dynamically adjusts the order size based on the current position, controlling inventory risk
1
0
6
@0xFastLife
@elixir
Thanks for the feedback. our aim will continue to be to driver higher potion amounts to our community (and less to mercenary capital), but there are other ways to achieve that without making people feel mislead
we've made some changes to address:
Elixir was built for and alongside the community
We hear your feedback, and we’re making changes for elxETH minters
⚗️
198
231
766
0
0
6
After empirically testing the above we've seen the same improvement in reducing inventory risk (controlling order sizing in the first, not doing so in the latter):
1
0
5
Welcome to the family
@bluefinapp
⚗️
3
0
5
1
0
3
(One cool side note: microstructure noise effects are less important for smaller tick stocks - meaning its likely slightly more profitable to MM low priced stocks with a $0.01 tick)
1
0
4
Elixir will remain as decentralized, open source infrastructure.
Our early access launch is slated for Nov 1st. To stay in the loop, follow us on Twitter or join our Discord:
0
0
3
This happens as the probability of a fill from an informed participant increases.
Testing empirically, we’ve seen the same effect. Instead of making assumptions for asset price (like GBM, brownian motion or a random walk) we utilized ETH-USDT 1min orderflow data from FTX
1
0
4
Future areas I'm exploring: iterating on calculating an optimal Gamma value (the risk aversion parameter), and exploring introducing randomness to parameters via a VRF, reducing gamability
1
0
4
@cryptohound666
@jungle_sven
Most crypto MM firms run different variations of AS but their comp model is different: they're receiving a significant amount of tokens for their liquidity, and so being able to hedge & protect USDT is more important for them. They have different incentive structure than trad MM
3
0
4
As empirically shown above in "Microprice" (Stoikov 2018), plotting the buy/sell volume conditional on pre-trade imbalance between bid/asks shows a clear opportunity
1
0
4
@NicsHODLN
@elixir
No, you can still withdraw your ETH in Aug. if you choose to mint deUSD, it’ll be minted to deUSD in December (so you keep your ETH exposure until then)
2
0
3
@aivainvesting
@elixir
Gm, users who choose to mint will keep their ETH exposure
In Dec, whenever you choose to claim your deUSD, only at that point would you lose ETH exposure
4
1
3
@wumpycrypto
@elixir
Gm sir, you should be able to now. had some things break with user load around Orderly Quantum launch but should be resolved
Working hard to improve user experience 🫡
1
0
3
It’s common knowledge that as latency increases, market makers widen their spreads to compensate for the increased risk of adverse price movements
When running Markov Decision Processes (MDPs) for a set spread, MM firms see a lower EV per order
1
0
3
@Pickle_cRypto
@0xjaypeg
@elixir
gm, we originally had Sui wallet support as well, but realized their zklogin uses different curves for their signatures (messing things up on the backend). planning on adding that support back in the next day or two
1
0
2
The global, decentralized financial settlement layer is trading at 50% of Home Depot's marketcap, and you're panic selling anon?
Extend your time horizons
0
0
2
@Pickle_cRypto
@0xjaypeg
@elixir
@craetin69
Nope not a mistake - it works on the beta agg page since there's no signatures required to access
0
0
2
The effect of latency on HFT is shown both theoretically and empirically. We plan on quoting slightly wider than most centralized participants, but allow for native exchange integrations and enable parties to incentivize delta-neutral liquidity to their books
0
0
2
@0xRayJ
@pipeline_xyz
@ElixirProtocol
@SkillovuU
No reliance on outside oracles to build orderbooks - purely uses orderbook state. imagine uniswap v2 LPing for orderbooks
(exchanges rely on oracles tho, which factor into funding rate calculations/liquidations)
0
0
2
@MattyIceCO
@ElixirProtocol
Hey, would check out our docs
1) We don't trade/hold user assets. Users can connect their individual accounts to Elixir, or participate in bonding (Protocol Owned Liquidity)
2) Similar to Uniswap LPing: projects subsidize returns & APYs, LPs face gains/losses from the passive MM
0
0
2
@DigiTektrades
@0xFastLife
@elixir
Agree that this could've been handled better - we've made some changes to account. see here:
Elixir was built for and alongside the community
We hear your feedback, and we’re making changes for elxETH minters
⚗️
198
231
766
1
0
2
@icebergy_
Many protocols I know pay salaries from exchange accounts - that way they don't dox employee wallets & comp to other employees (or outsiders)
0
0
2
@0x6851
@ElixirProtocol
@vertex_protocol
@cole0x
odd, could you reach out to me in DM with your discord handle?
1
0
2
@cryptohound666
@jungle_sven
In trad MM, market P&Ls are the main profit center. For AS however its useful to feed an entropy variable to preemptively widen spreads during vol (avoiding suboptimal fills and toxic flow), boosting profitability quite a bit
0
0
2
@ThogardPvP
@0xLuthor
@hasufl
Arbitrage tightens prices but widens spreads
Imagine asset A is trading at $110 on Uniswap, $100 everywhere else. Arb driven ("toxic" orders) drive price on Uni down to $100. LPs eat this loss via IL
Higher IL -> less incentive to LP tightly -> less liquidity + greater spreads
2
0
2
Elixir solves this.
20+ DEXes are already natively integrating Elixir. This enables users on the exchange to supply liquidity to their pairs via our protocol's dPoS network: building up orderbooks while earning subsidized APYs
1
0
2
@DefiIgnas
@0xFastLife
@elixir
Hey Ignas appreciate the feedback
our aim will continue to be to driver higher potion amounts to our community (and less to mercenary capital), but there are other ways to achieve that without making people feel mislead
made some changes to address:
Elixir was built for and alongside the community
We hear your feedback, and we’re making changes for elxETH minters
⚗️
198
231
766
0
0
2
Microprice converges on the mid-price as market opportunists expand out over orderbook depth ticks (Stoikov posts a proof in his paper): the predictive power speaks for itself:
1
0
2
This micro-price is defined as the limit of mid prices (and therefore if Pt micro exists, is a martingale for all times t > 0):
1
0
2
Orderbook liquidity is inherently broken. While AMMs have attracted billions in TVL, leading orderbook-based exchanges have found it much more difficult to bootstrap a fraction of that liquidity to pairs
1
0
1
0
0
2
As seen in the above, controlling order sizing enables the strategy to effectively work towards minimizing inventory risk, whereas a lack of order sizing controls leads to a slow spread to a 100/0 inventory split
1
0
2
@moo9000
@roinevirta
@ElixirProtocol
It's not suboptimal actually: with latency MMs just need to quote a bit wider. Our protocol updates orders once a second, enabling users to easily + trustlessly supply liquidity to orderbooks (not trying to always be top of book)
Good paper on the topic:
0
0
1
There are also losses in the form of missed profits, rel to if the party were to have near-zero latency.
We created a function widening bid ask spreads, then calculate and visualize total annual losses solely due to latency (as spread widens, expected losses decrease):
1
0
1
@TraderYunt
@0xgilbert
Also wasn’t able to ape (the Vertex Fusion poops got sniped nearly instantly). We’ll be increasing caps soon, appreciate you bearing with us 🫡
1
0
1
(Interesting note: Kurtosis is quite high compared to a normal distribution)
ETH’s vol appears multimodal:
1
0
1
This is applicable to Elixir: our protocol comes to consensus every second, enabling trustless, decentralized order computation via our consensus engine. This introduces additional latency in our system however, which would introduce additional vol into our ANS vol calculation
1
0
1
We introduce latency: applying a 5 min lag to calculated bids/asks
Below we can see the proportion of bids/asks immediately at a loss:
3
0
1