Nvidia reports in an hour. First report without Jensen Huang, and half of CT is trading his comeback instead of the numbers.
This afternoon Huang said he'd be "honored" to return if the board asks. Within the hour the board said it isn't considering it. Polymarket shrugged at the board: odds of a return by year-end went from 8.5% to 12% and haven't come back down. The stock was less romantic. Nvidia's perp on Hyperliquid added 1.2% on his statement and gave back 0.8% on the board's answer.
To be fair, there's a reason the crowd wants him back. The perp opened November at $204 and sits at $153 now. The low, $148.86, came the day he resigned. And everyone remembers Disney running as much as 10% the morning after Iger's return was announced in 2022.
But tonight the board stays quiet and the guidance does the talking.
If the guidance holds, Nvidia without Huang starts to look like a business rather than a founder story, and the comeback bet loses its point. Tech bounces and drags crypto along. Today BTC, ETH and Hyperliquid's S&P 500 perp all bottomed in the same minute, 15:54 UTC.
If it disappoints, "bring Jensen back" gets louder and the resignation-day low of $148.86 stops looking like the bottom.
Which makes it a strange trade. The better the report, the less anyone needs him.
And a crowded one. $NVDA is the 4th biggest market on Hyperliquid, $461M of open interest, more than SOL. trade[xyz] cut leverage to 10x for the print. Options price about 8%, enough to liquidate a 10x position on either side.
P.S. If he does come back, Nasdaq won't be the first to price it. Last Sunday this perp moved almost 4% in minutes on a weekend headline.
1
5
125
At 10:28 UTC the attacker asked Lighter to withdraw $5,096,588
At 18:03 Lighter paid out exactly $5,096,588.02. Just not to the attacker. To a wallet first used at 16:04
Same withdrawal, new owner
1
2
138
How the keys leaked, per Lighter's postmortem: the victim's own signing code reused nonces, and Lighter's signatures were public
In a Schnorr-style signature s = k + e·x. Sign twice with the same k and anyone can solve x = (s1 − s2) / (e1 − e2). No access to Lighter needed
1
38
Where it went: the new wallet got $50 USDC from an address Blockscout tags as Abraxas Capital Management, received the $5.1M at 18:03 and sent it to a Binance deposit address 4 minutes later
Good outcome for the victim. It also means Lighter's Security Council can change who a pending withdrawal pays
37
Lighter says the protocol and smart contracts are unaffected. Technically, that's true. Nobody broke anything this morning. Someone with a trader's leaked API keys used Lighter's own features, in exactly the order they were built
09:35 UTC, a fresh wallet deposits $10,200. Five minutes later a new public pool appears on the victim's account with a 0% fee. The victim's $1,000 seed is pulled out and replaced with $1
So the attacker's $9,999 buys 99.99% of the pool
Then a dress rehearsal. The victim's account sends $1 in, the attacker cashes out to check the math and buys straight back in
09:45:01, the real thing. Ten transfers of $1M and one of $383,864 from the victim into the pool in 33 seconds. At 09:45:38 the attacker redeems for $10,392,826. The victim's 0.01% is worth $1,039
At 10:25 the same script ran on a second account for another $5.1M. That withdrawal is the one Lighter is now trying to claw back
37 seconds, $10.4M
1
89
The wallet behind it: 0xb3fffca23d4444eE73349825ACD84f7b5020e387
Seed money bridged in on Nov 12, pushed through Aevo's bridge and out to a fresh address on Nov 13, which breaks the trail. Then at 09:00 today gas came straight from the pre-Aevo wallet anyway. A $29.96 test deposit hit Lighter at 06:47
After the withdrawal, $10.36M went into ETH on CoW in 4 minutes. 1,920 ETH left for Arbitrum via LayerSwap in 160 ETH chunks, ~420 ETH to what look like THORChain vaults, 10 ETH into Tornado. 3,404 ETH is still sitting there
1
32
I can't tell you who did it. A wick like this needs leverage on both sides, and Binance OI added $27M on the way up and lost $22M in 15 minutes on the way down
If you farm perp DEXs anyway, Variational is my main one: omni.variational.io/?ref=OMN…
77
Catch: these are live estimates. Every new wallet dilutes them, and OI is time-weighted, so a position opened today counts for less than one held since Nov 10. Funding on the two legs can differ too
RISEx changed referrals this week: whoever refers you now also gets 4-15% of your fees in USDC, on top of 10% of your points, and neither comes out of yours. My code liwikruk gives +15% points: rise.trade/invite/liwikruk
1
28
Fear & Greed is at 13 and everyone's trying to catch the bottom. I'd rather catch the funding
On Thursday XMR longs on Hyperliquid paid ~490% a year to stay in. Now the shorts pay them, 20 of the last 24 hours. On Binance the same coin's longs are still paying
Long on Hyperliquid, short on Binance: $10K made ~$15 in the last day while XMR fell 10%
2
159
How I'd play a market like this: same size on both legs, low leverage so a wick can't liquidate one side. Hyperliquid pays every hour, Binance every 8h
Fees in and out on both legs come to ~0.19%, so the first day or so just pays for them. Still cheaper than guessing the bottom
1
83
Catch: funding moves by the hour. XMR went from longs paying ~490% a year on Thursday to shorts paying 235% at the worst hour today, and it flips back once the shorts close. XMR is thin too, so keep the size small
If it turns positive again, I'd close both legs the same hour. P.S. The bottom callers can keep the bottom
42








