so, if you know me, you would how much i love to geek over comparison analysis... so from time to time, I gotta bring it out... this time, i got @Enter_Elysium vs Hyperevm, cause that gotta be interesting to see right... so i made a dune dashboard for that comparison: dune.com/sarcastinatooor/ely… hyperevm is the useful baseline here because both environments can run the same evm contract call. if an app builder is choosing where that call should run, fees and observed inclusion time are worth comparing under the same workload... i left hypercore out because it is the native trading engine, with different actions and execution mechanics. running a contract-call benchmark against its order books would tell us nothing useful about either one... in the initial testnet sample, elysium came in at 1.91s median observed inclusion versus 2.22s on hyperevm. at p95, it was 3.11s versus 3.75s. 1,000 identical calls cost about 0.000357 test HYPE on elysium, against 0.002886 on hyperevm... the dashboard shows the hourly results, source timestamps and individual transactions so you can inspect the sample. it ran from one client host, and observed inclusion includes network and polling time. this does not establish final settlement speed, maximum capacity or mainnet fees... next i want to remove benchmark traffic from elysium’s activity and see what people are actually building and using there... go check it out and leave some love if you like it...

a simple way to farm it, is to run a small two-sided market-making strategy through the @tread_fi mm bot: > set a conservative inventory limit and small order size > quote both sides of the book instead of taking oversized directional risk > use a tight but realistic spread so your orders provide genuine liquidity > enable cancel-and-replace so stale quotes are removed as the market moves > monitor fills, inventory, fees, and unrealized PnL. > stop the bot before the market settles and close any remaining exposure the goal is to create real market activity while keeping risk controlled. the $400 is an incentive pool, not guaranteed income, and rewards may depend on outcome’s eligibility rules, volume allocation, fees, slippage, and competition from other traders...

track it live here... hack-trail.vercel.app/incide…

i think it's about time we call @MagicEden a joke... they've been getting drained, as we speak, almost 3,832 NFTs moving out of hundreds of wallets... guess, no one is really surprised here... the initial report flagged a suspected Magic Eden approval path. shortly after, @0xQuit identified the collector wallet as whitehat recovery infrastructure and said the NFTs would be returned... that changes how i read the transfers, but it does not close the case. we still need the exact spender contract, the approval or signature that allowed the moves, a collection-by-collection reconciliation and proof that the assets reached their owners again. the dollar loss is unconfirmed... for now, i would check old NFT approvals on any wallet that interacted with the suspected path and revoke permissions i no longer need. i would also ignore anyone offering “recovery” through a DM... my build for this exploit is a lil different, first time I added the option to revoke your permissions too, so this should be an elevated experience for you... hack-trail.vercel.app/incide…

ethereum:0x31eea0edeb85b0cb65c1400bd2f4e487fd61b537 keeps on cooking... this should bring-out more mm-terminals to do a better job...

this one's for all the polymarket maxis out there doing moon math on trader profiles: i saw one advertised as making $149,221 while losing only $96 across 12,000+ trades... the profile showed 12,697 “predictions” and about $5.5k in open positions. polymarket’s documented trade count measures distinct markets, not fills or completed round trips... on one football match, two unders were marked down $88 and $82, while two spreads were up $76 and $6. four markets, one underlying event, roughly -$88 across the group. the two losing legs alone exceed the claimed $96. perhaps that figure refers to closed losses in a particular window, but it needs a definition... before you term this as a recursive system, i want realized p/l from exits and resolutions separated from open inventory, fees, maker rebates and rewards. then group the bets by event: several lines on the same match are not several independent risk decisions... and replay the order book at entry. a fill at 58.5c tells you little about what someone else could buy at $1k size after spread and slippage... the trader may be excellent. i just want to know which profits are settled, which are still at risk, and whether anyone else could actually get those prices...

if @variational_io lands of @tread_fi it is game-over for mm's... it is simply gonna hike-up the volume by atleast 4x and OI by 2x...

we've been so obsessed with perps that we started hurting spot... we may not have killed spot per-se but it is definitely in a rut... perps share of all defi volume, by year: 15.8%, 27.8%, 42.2%, 50.5%, 61.6%, and 72.0% so far in 2026. it crossed half in 2024 and never looked back. the ratio went from 0.19x spot to 2.57x... mark it, i am not spot is ded yet. $5.64b a day in 2021, bottomed at $2.47b in 2023, peaked at $13.46b in 2025, and is running $8.44b this year. that is still 3.4x the 2023 trough. spot did not collapse, it just stopped capturing anything new... 2026 feeling like a turning point tho. perps are running $21.66b a day against $21.59b last year, so flat. spot is down 37.3%. the share shift this year is not perps winning, it is spot bleeding while perps stand still... open interest changed shape though. $0.43b average in 2021 against $9.19b this year, $15.5b right now. and turnover has been falling, 2.91x a day over the last year, 1.73x over 90 days, 1.64x over 30. more capital parked, less churn per dollar of it. leverage holding is only going up with OI being heavily pushed by all perps... it is also not a market, it is one venue and a queue. hyperliquid is 36.4% of all perp volume on its own and the top five are 72.7%... where it ends up if nothing changes. the perps to spot ratio has compounded 1.66x a year and doubles every 1.37 years. extend that and 2027 is 82% perps, 2028 is 88%, 2029 is 93%. spot ends up as a settlement layer that exists so perps have something to price against... i do not think that line holds, mostly because turnover is already rolling over and a market at 93% leverage has nothing left to lever. but the direction has been intact for five straight years and i have not found the year it broke... the thing i would actually watch is not the ratio. it is spot volume in absolute dollars. as long as that holds near $8b a day the ratio is just perps doing their job. the day spot goes back under the 2023 trough while perps stay flat would probably a warning sign for us...

why no one has caught this for $JUMP? $100m NO here should be an easy buy here...

where can I trade $SI?

there are defintely some red flags around @QFEX i guess... this reminded me again to be extra-vigilant about where and what we wanna farm for the sake of clout... i did called out qfex distinctly unique for bringing new types of asian equities, but guess should have a deeper-dive... so this is where i am at, and it looks dicey exactly as mr. potato called out here: > defillama files qfex with the chain field set to "off chain", no tvl adapter, zero audits listed. it is a cex with crypto rails > custodial by design. client agreement 5.1, client money sits in segregated accounts with third party licensed custodians. you are a creditor of qfex, not a keyholder > clause 10.3 caps their total liability to you at the value of your own balance at the time of the claim, and per trade at that trade's notional. so the ceiling on what you can recover is roughly what you already had > any interest earned on your idle balance is retained by qfex as beneficial owner. your margin earns them, not you > deposits and withdrawals must go to a bank account in your own name, one nominated account at a time, and requests can be denied or delayed while compliance checks are pending. money in through a wallet does not mean money out through a wallet > they can freeze, restrict or terminate at sole discretion on suspected suspicious activity, and separately reserve the right to delay withdrawals for compliance > the oracle methodology is published properly, index sources, futures roll blending 20 points across the final five CME boundaries, mark price as median of bid ask trade with a 150s EMA. but no data vendor is ever named. equities are "real-time feeds from us stock exchanges", metals "an aggregate of the most liquid ECNs". that prices equity perps at up to 50x

how cool is the Kalshi fight, feels like the 2020 era is back again... reading everything that comes to table, send me more hard-hitting evidence.. Kalshi going down means an HIP-4 takeover is eminent...

setting my limit TP at 99.999 just to finally clown on all you dot hls that will teach you

i was not wrong on this part...

ngl this feels like a whole new turn-around for the hyperliquid ecosystem... i mean we know hyperliquid already has the orderbooks build around perfectly, but now i wanna see how much more we can build around them without every application having to run its own offchain trading desk too... with @Enter_Elysium to testnet now live, my hot take is: a whole new wave of applications will again be targeted to somehow burn so much $HYPE the published architecture uses arbitrum orbit, settles to hyperevm and uses $HYPE for gas. the targets are 100-200ms blocks and 300 million gas per second. this is gonna hard honestly... the app i would most like to see is a prop amm that adjusts its quotes using hypercore depth and its own inventory, then hedges fills on hypercore. knowing how much liquidity exists within a price band could help it decide how much size to quote and what the hedge might cost. faster execution becomes useful when you need to update that quote before somebody trades against an old price... options vaults could use a similar setup too. collateral and option accounting could live on elysium, with the vault adjusting its delta through hypercore perps. given the option spreads we have been discussing, i would love to see this tested. cheaper updates could help, but somebody still has to fund the inventory and absorb the risk between selling the option and getting its hedge filled... then there are automated basis strategies and lp vaults. an app could track its spot exposure, funding and available perp depth, then resize a hedge as conditions change. the useful test would be its net return after funding, execution costs and delayed hedges. i have spent enough time looking at bot dashboards to want that reconciliation from day one... the token route is more concrete: > launch and build liquidity on elysium > bridge a backed representation onto hyperevm > pursue a linked hypercore spot market > potentially add a separately deployed HIP-3 perp each listing still has its own deployment requirements. bridging a token does not automatically give it an orderbook or a perp... i know hyperevm would remain the settlement layer and the connection for token transfers. existing hyperevm assets could bridge into elysium applications, while hypercore supplies the trading books. lending protocols would still need to explicitly accept those assets; being in the same ecosystem does not automatically make them usable as collateral... and the rollout needs watching. the docs schedule native hypercore reads and the fast write path for a post-mainnet upgrade... this is where i see them heading, and honestly would make the most sense for elysium... i am bullish on more applications generating useful flow for hypercore. testnet would showcase us the evidence that quotes stay competitive, hedges complete reliably and capital can move without users getting stuck between environments. that would give developers a much solid framework to then build on...

how soon people will start to loose confidence in @extendedapp TGE? i might be wrong here, but I never farmed extended and maybe it wont cook as much as you guys want it to...

ok so after the massive success of @tread_fi i think the space is gonna blow-out and gonna grow out in parallel with perp-dexes... with only a handful of high-value perp-dexes currently being farmed by the ct, all mm-terminals are gonna fight and build on the same, which is only gonna be good for us farmers... i've been testing out this new terminal @vesper_mm cause they got all the key perp-dexes that i am farming right now... i was particularly drawn to their new strategy called Flux for @risextrade alongside a two-week trading competition. Flux can run XAU, QQQ and SPY together, though i started with XAU first because i wanted to see the execution before adding more moving parts... my initial session returned this: > roughly $2.4k routed > 35 round trips > 70+ fills > $0.18 session fees > $0.36 all-time fees > +$5.8k CPM shown by vesper the CPM figure is an execution metric from the session ledger, not my realised RISEx account PnL. i am keeping actual return claims out until the run is done and the venue balance has fully settled... still, the early read is pretty good... flux is built around a very particular kind of market making. entries and ordinary exits are post-only. it does not try to force a fill whenever the market moves. if one side of the book becomes thin or volatility starts becoming stupid, it pauses that side rather than continuing to stack inventory into bad conditions... a forced taker close only comes into play at the stop i set... that is pretty much what i want from a strategy running in these RWA books. it should be making the spread when conditions allow it, and be comfortable doing nothing when the book starts looking like it wants to donate your margin to someone else... the other part i liked is that they have made the ugly operational bits visible. RISEx has a toxic-hours guard around 13:00–16:00 UTC; their recorded fills showed roughly +98 CPM more cost in that period, so Flux stops building fresh exposure there while working any existing position down... the strategy is not pretending that every hour produces the same fills, or that more volume automatically means better market making... RISEx fees are 50% lower through september 25, which helps the current run and should be stated plainly. that is also why i will keep this running through the Flux Cup rather than turn a few thousand dollars of flow into a victory lap... i want to see how the CPM behaves once: > fee conditions normalize > XAU gets a proper volatility session > the strategy has more routed volume behind it > the guardrails actually have to intervene > and i add the QQQ/SPY legs instead of relying on one market but so far, vesper is giving me what most market-making terminals avoid showing: the fills, fee drag, round trips and execution output in the same place... that is enough for me to keep routing the RISEx and Arcus accounts through it for now... PS: the points are live as well...

am i the only one using @legendtrade still or it has gone down? if you aren't maybe try it out, its fun definitely... app.legend.trade/r/IXXX2HH5

what are we going to eif @variational_io points continue post september? asking for a friend...

i was so not a funding rate guys, untill i was... so for example here, my boros position is showing a green PnL number but it is soooooo not true... what i am holding is 1% of a bybit HYPEUSDT position i closed badly, long at 6.50% fixed with 11 settlements and 3.67 days left. the green figure on the positions tab is funding accrued back when the position was full size, when every settlement was printing 10.95% against my fixed. the realised loss from the close was nearly 2x that accrual in the opposite direction and it is simply not in the field. the market is net negative and the screen says otherwise... where the rest of it lands is already narrow... mark to market decays to zero at maturity so carry is all that is left. if bybit keeps printing its 10.95% baseline it settles +0.045% of notional. at the mean since i opened, 6.32%, it settles -0.002%. flat funding gives -0.065%, deeply negative funding gives -0.116%. the entire realistic range is about +0.045% to -0.116%, which is noise on a margin slot i would rather have back... next one is hyperliquid HYPEUSDC, 30 october expiry. implied 9.78% against a hyperliquid median of 10.95%, so i would be paying 11% below median to be long funding. carry is 0.167% of notional over 38.5 days... the problems with it are honest ones. the spread is 57.1bp on a 9.89% mid, so crossing hands over 36% of that carry immediately. and hyperliquid funding is hourly and uncapped, stdev 12.70%, 8% of prints negative, range from -20% to +133%. daily implied vol of 0.362% against sensitivity of 0.0010559 per YU means a daily mark to market swing 8.8x larger than a day of carry. on the bybit market that ratio was 1.0x... per unit of notional: carry to maturity 0.167%, a normal day moves 0.038%, a three sigma day moves 0.115%. one bad day is 69% of the entire expected carry, and that ratio does not change with size. sizing does not fix it, only holding period does... so i bid it rather than cross, size it so a three sigma day is boring, and write the exit before entering. out if implied closes above the hyperliquid median three days running because that kills the thesis, otherwise hold to 30 october... four things i would keep in mind on any funding position... 1. price carry off median settled prints over a window as long as you intend to hold. never off the underlying APR in the UI, which is an unsettled forecast that flips inside the period. bybit was showing me -3.66% while the prints around it kept landing at 10.95%... 2. multiply carry by remaining life before you get excited. a good annualised rate on three days left is nothing... 3. check the spread against that carry, not against the rate. a third of the edge can be gone on entry... 4. and run noise against carry. near 1x you are being paid to wait. near 9x you are taking a rate view whether you meant to or not, and the only thing that saves you is not touching it...