SeaSaltMarketMakingNotes

$ESPORTS This drop is relatively gentle; waiting for breakout confirmation—once momentum returns, I’ll get on board.

A 22% jump looks great, but high volatility means it could plunge at any time—don't FOMO.

Linger near the range low for a bit—don’t rush to go all in; first see whether buyers step up.

AI Agents are all rushing on-chain now, helping people rebalance, snipe launches, and even provide automated liquidity. I’ve tried several with automated trading features, and honestly, they can handle simple commands, but fall apart when judgment is required. Take market-making parameter adjustments: a lot of people hype AI as being able to automatically hedge impermanent loss, but in my own tests, the strategies suggested by the model didn’t match my risk tolerance—it sees high volatility and narrows the range too much. Impermanent loss is lower, but so is the yield.
This still requires watc

Sometimes your trades keep getting jumped and sandwiched—not necessarily because of bad luck. First, check whether there are a lot of bundles in the block. You don't need to understand builder mechanics; a rough look is enough. Lots of bundles means smart money is competing for trades, and you're already facing the other side. Exit when you should.
I’ve made markets in small pools and specifically studied parameters and errors. Later, I realized on-chain it’s the same story: the key is always the signal, not the strategy. AI agents are all the rage now, bringing automated trading and on-chai

Accumulate in the 4.628–4.632 range; if it holds above 4.7, target 4.75; stop-loss if it falls below 4.55. Follow the rules.

Just sent a USDC transfer back from Osmosis to the Ethereum mainnet, and while I was at it I went through the trust nodes in the process. Honestly, every time you bridge cross-chain, it feels like you’re playing a game of trust relay. And once the bridge on the other end is off, you basically can only pray.
Take this IBC case, for example—the trust chain is actually pretty straightforward: first, it relies on the Light Client on the target chain to correctly verify the other chain’s block headers; then it requires the Relayer not to intentionally drop packets or tamper with messages; and final

I just withdrew from the pool one. The other day I was looking at a small position of 0.03 ETH and hesitated about whether to cut the loss. I thought, if I just ride out the volatility a bit more, it’ll come back. But before going to sleep, I glanced again—and it dropped another chunk. Forget it. I’ll accept the loss and exit. 😅
It’s basically the same as a breakup. The longer you drag it out, the worse it feels. Cutting it earlier actually saves you interest and effort—especially lately, with hardware wallets out of stock and phishing links flying all over the place. It’s the same on the sec

Over the past couple of days, I’ve looked at several RWA on-chain projects—the data stack looks pretty impressive, but this liquidity—plainly put, it’s a thin ice layer. Some pools show TVL of a few million, but in reality their trading depth can’t even handle a 100,000-order size; the slippage is absurd. And let alone the redemption terms: some projects write it like “you win,” but when you dig in, you find the redemption time is extended, fees are quietly increased, and there’s even a disclaimer like “subject to market conditions.” What else is this but a liquidity mirage?
Yesterday on the t

Lately I’ve been watching a few DAO votes, and the more I look, the more interesting it feels. On the surface, the proposals are about community governance, but underneath it’s all a game of incentives and power. For example, some proposals cram in a bunch of incentive parameters that seem designed to attract new users, but in reality they feed short-term arbitrageurs with disproportionately large voting power. By the time long-time users react, the proposal has already passed. I used to do market-making in small pools, so I’m especially sensitive to this kind of “turning impermanent loss into

I was flipping through data on stablecoin supply and ETF inflows, and a thought popped into my head—don’t rush to treat correlation as causation. It’s like the tide or like a thermometer: when the tide is rising, everyone thinks it’s a good thing, but the thermometer itself is only an indicator. Lately, everyone’s been talking about rising rate-cut expectations and the U.S. dollar index moving up, and risk assets moving along too. Honestly, I’m a bit confused—I’m not sure whether the money is really coming in; you have to look at the actual on-chain wallet activity, not just a few numbers. Whe

I just came across a post explaining how block builders package bundles, with technical details all the way through—it gave me chills just reading it.
Honestly, do retail users really need to know this? My thoughts are simple: you only need to know that “someone can see your transactions in advance and jump the queue,” and the rest is handled by slippage settings and private transaction channels. I tried to study those parameters, but the more I looked, the more I felt like a headless fly. In the end, I realized it’s better to spend the time on risk management.
I thought knowing how bundles ar

Recently I tried a few AI Agents to monitor on-chain interactions. To be honest, some parts do save effort, but there are plenty of traps too. For example, when it comes to wallet signing, the Agent can automatically recognize the contract address and provide a risk warning. But once it encounters a lookalike contract masquerading as a common project—especially a newly deployed one that hasn’t been included by the labeling system yet—it just freezes and directly shows “unlabeled.” Last time I nearly got pulled into an interaction by a fake GMX Agent—luckily I checked the creation time, and it

I just saw a friend say they chased a privacy-sector token, and it pulled back by a dozen or more percentage points—now they’re saying it hurts. Honestly, lately the discussion around these privacy coins and mixer protocols has been pretty heated in the community: one side says compliance is a matter of survival, while the other says anonymity is the original intention.
Putting positions aside, I personally think when you get itchy and want to chase a run-up, the first question you should ask yourself is: are you actually seeing new information, or are you just being pushed along by emotion?
M

Just after I plugged the charger in, when my phone’s battery jumped from 5% back to 20%, it suddenly hit me—NFT liquidity isn’t it basically the same as this charger? The floor price looks low, but when there’s no volume, nobody will pick up your bid, no matter how far down you place your order. Royalties are another matter entirely—project teams rely on community narratives to keep things afloat. Then on the chain-game side, inflation blows up, the studio runs off, the coin price spirals downward, and the community directly turns into a dead group chat. Plainly put: with liquidity, it’s not a

Every time the market’s hot-spot rotation kicks in, the first thing I’m used to doing is flipping through my own recap notes from before— not to check how much I’d made, but to look back at the “why didn’t I manage to hold back” or “why was I hesitating” from back then. Plainly put, I’m afraid I’ll keep repeating the same kind of mistake. This time, during the meme rotation, once a celebrity starts pushing a call and people in the group begin charging in again, I instinctively pull up that last trade I blew in a similar emotional mood and take a quick look— and I’m instantly calmer. I’m not tr

When liquidity dries up, it’s the easiest time to get carried away. Watching the pool like a dead stagnant pond, with everyone on-chain canceling orders, people still keep shouting “buy the dip, buy the dip.” I’ve seen too many people who still have a bit of U left in their hands and feel they have to rush in to catch falling knives—only to end up trapped even deeper.
Actually, when I was building pools, I understood it: when liquidity is gone, even if you pick the safest trading pair, it can still double down on impermanent loss. The recent NFT drama has been equally lively—royalties were cut

They say market making is effortless profit. I used to think so too—back then I even felt it was nothing more than placing funds there and waiting to collect fees. Now looking back, I can only say it’s the curve that matters—something about it has fooled plenty of people.
That AMM curve looks smooth, but it hides some tricks. What you think is impermanent loss is what the math formula calculates, but what’s really painful is this: the coins you put in might go up by 30%, and when you take them out, most of the gains get wiped out by slippage and losses. In plain terms, market making isn’t like

I just glanced at a certain liquidation transaction—gas shot up like crazy, but the quotation timestamps from the oracle data feeds inside it were actually off by three blocks. Later I found out it wasn’t the chain that was congested; it was the data source of the feed node itself refreshing slowly, causing the liquidation bot to wait there for half a day, nearly smashing the pool wide open.
Honestly, this is even more covert than impermanent loss. You think it’s market volatility messing with you, but it turns out that a feed delay turns you into someone else’s liquidity withdrawal machine. R