从职业扑克选手到交易巨头：走进 Selini Capital 的思维世界
- 散户已经离开加密货币山寨币市场。 Jordi说，资金流已经证明了这一点：散户下单流量“大幅”下滑，分散到Micron、SanDisk等个股、体育博彩和预测市场。Grass就是信号：团队告诉投资者自己能产生约5,000万美元现金，代币却在利好消息下单日跌去1/3。“什么样的垃圾币，会因为利好消息在单日跌去1/3？”问题出在结构：Binance约1年的锁仓期，最终会留下“一批亏得很惨的VC”，只能卖掉任何还有流动性的资产，直到散户放弃。
- Selini现在交易的是股票的加密版本。 SpaceX在Binance上的日成交额达到20亿美元；Selini在Binance、OKX、Lighter和Bitget上交易这些资产，并在传统金融一侧做对冲。Jordi的判断是，“在SpaceX上找到alpha并不难”，因为加密货币的交易剧本可以整套搬过来：KOL（Elon、CNBC）“胡扯出一套故事”，8月有“荒谬规模的解锁”，随后是下一批解锁，2周后召开财报电话会。“这些我们在加密货币里见过，知道事情会怎么发展。”
- 被动资金是指定的接盘者，而且流向高度可预测。 Nasdaq即将把大量资金配置到刚上市、低流通盘的SpaceX（不在S&P 500成分股之列；按Jordi的说法，S&P 500规模约为Nasdaq的13–14x）——这可能是被动资金第一次在如此大程度上被“利用”。Millennium在指数再平衡上的28亿美元、37亿美元和27亿美元数字，本质上是同一笔交易；加密货币版本则是抢跑Michael Saylor——他本质上就是傻钱的“Nasdaq”。
- 加密货币已经形成一个“震荡底”，当前动能为零。 从AI向加密货币的轮动必须慢慢完成——资金得先退出SpaceX、内存股和KOSPI赢家。行情一旦启动，会突然且高度反身性地发生：“一切都会毫无理由地发生……我们怎么又涨了10%？因为昨天涨了10%。”
- 10月10日让Selini损失了个位数百万美元。 一些套利团队“表现极佳”，但使用3–4倍杠杆的账户——把10–15%的全额抵押回报推向60%——来不及重新平衡，最终被清算。永久性变化是：总体上避开非美元抵押品；USDE和wrapped SOL一旦“暂时脱锚”，就会遭到激进清算；更多采用组合保证金，也更加保守。回测永远看不到这种每3–4年才发生1次的事件。
- 加密原生的微观结构优势仍然扛得住Jane Street。 永续合约资金费率在下一个窗口之后无法预知——“公式只会告诉你下一个窗口，不会告诉你再下一个。”随着永续合约未平仓量扩展到SpaceX并进行24/7交易，“这是我们的世界……他们进入了我们的地盘。”Jane Street涉足ETF级别的博弈，可能只会“削掉几个点”。
- 按Jordi的说法，所有金融的底层原语都是“你总是在用风险交换预期价值”。 去风险者（创始人、锁定期股份持有人）会为了落袋为安，把优势让出去；赌徒则愿意让出优势购买波动——“他们想要的是那个1,000倍”。对散户的实际推论是，费用会复利式累积：5%的优势损耗“会让你从盈利变成不盈利”；而Kalshi等平台会刻意把你引向市价单。
1. 大过滤器：散户离场，代币结构失灵
- Jordi对本轮周期的开场判断是：“大过滤器正在启动”——低梯队创始人已经拿不到融资，市场上剩下的只有有纪律的资金。恰恰印证这一点的例外是：Selini在加密AI领域的被投公司Prime Intellect宣布融资约3亿美元。但在他主持的加密AI圆桌上，没有人愿意明确为某个具体项目站台，包括Grass和Tao在内；大家都“非常看好AI，也非常看好加密货币”，但“很难把两者连在一起”。
- Grass是诊断案例：团队在投资者电话会上说，自己大约能产生5,000万美元现金，代币却跌去1/3——而且是在利好消息下。据转述，创始人Andre抱怨说：如果它是一只股票，按这样的增长速度，应该以很高倍数交易；但现在的FDV却只有约380。“这完全说不通。”
- Jordi归咎于这一机制：Binance为锁住流通盘，强制设置约1年的锁仓期；1年后，“一批亏得很惨的VC”只能卖掉仍有流动性的资产，持续把代币价格压低，直到散户退出。散户下单流量已经“大幅”下滑，转向Micron、SanDisk等个股押注、体育博彩和预测市场。“散户已经没钱了。”
- 关于AI向加密货币轮动的问题，Jordi认为它必须慢慢完成：人们得先退出SpaceX、内存股和KOSPI赢家。加密货币涨跌两边都是动量游戏，而现在“完全横盘”，形成一个“震荡底”；真正的行情会突然、反身性地启动——“一切都会毫无理由地发生……我们涨10%，只是因为昨天涨了10%。”投资者仍在追逐FOMO，他也不知道这种状态何时会改变。
2. 传统金融变成了加密货币：SpaceX成了新的主流品种
- Selini正在Binance、OKX、Lighter，甚至Bitget上交易“股票的加密版本”，并在传统金融一侧做对冲；仅SpaceX在Binance上的日成交额就达到20亿美元，超过Ethereum市场。Jordi过去并不是股票交易员（固定收益、大宗商品、外汇），但“游戏就在这里进行。说实话，这里有alpha。”
- 他用一句话概括这套迁移：“这和我们在加密货币里学到的东西一模一样”——KOL包括Elon和CNBC的那些评论员，拿一套故事胡扯来影响散户；8月有“荒谬规模的解锁”，随后是下一批解锁，再过2周召开财报电话会。“这些我们全都在跟踪，知道事情会怎么发展。”
- 主持人补充说，通过Goldman获得的IPO配额（可能是Goldman，听起来像“Oldman”）就是Binance Launchpad的对应物——定价偏低，拿到配额的人纷纷转手。两人过去总爱开玩笑说加密货币会变得更像传统金融——它确实变了；结果传统金融又变得像加密货币。
3. 被动资金接盘：流向机械且可预测
- Jordi最尖锐的结构性判断是：Nasdaq即将把大量资金配置到刚上市、低流通盘的SpaceX；SpaceX不在S&P 500成分股之列，而按Jordi的说法，S&P 500规模约为Nasdaq的13–14x。“最终接盘的总是这些被动资金，而这些自动化资金流太容易预测。”
- Millennium据报的指数再平衡交易数字——28亿美元、37亿美元和27亿美元——在Jordi看来并不是什么天才交易：“他们真有那么聪明，聪明到没人能看出指数再平衡？我不这么认为。”这只是资本和注意力共同押注机械化的傻钱流。加密货币市场围绕Michael Saylor做了同一笔交易；他本质上就是被动买入的“Nasdaq”：资金在他之前买入，在他之后卖出，而当他停止买入时，资金也不得不退出。
- 主持人用了一个值得保留的熊市比喻：熊市就是一头真正的熊——你不必成为场上最快的人，只要比最慢的人快；最慢的是养老金和指数资金。难点在于抓住行情突然上冲的时点：“你昨天以110美元卖出Intel，今天醒来它已经是17美元……我为什么要卖？”Jordi的经验法则是，绝不追求任何单月或单年的收益最大化：留着现金，保留smart beta，靠时间复利。
4. 10月10日：每3–4年才发生1次的交易日剖析
- 这是喜忧参半的一天，不是灾难：一些套利团队“表现极佳”，但资金不足、使用3–4倍杠杆的账户——这正是10–15%的全额抵押回报与约60%之间的差别——来不及重新平衡，仓位变成单边，最终被清算。净损失为个位数百万美元，“对我们来说这不算什么大日子，无论好坏。”回测也救不了你：“此前从未发生过。”
- 市场的第一反应是：Jordi醒来时收到了50条Telegram消息，大家问“你们还好吗？”人们以为Wintermute或Selini已经倒闭，因为“那是唯一的解释”。Wintermute如今已经“死过”20多次；Selini只有2次或3次——“后面还会死很多次。”
- 永久性变化包括：通过组合保证金进一步统一流动性，并总体上避开非美元抵押品——10/10已经证明，USDE或wrapped SOL一旦“暂时脱锚”，就会遭到“非常激进”的清算。Bitcoin或许在一定程度上可以例外。“我们仍然使用杠杆，但已经更懂得如何使用，也更保守了一点。”
5. 金融的底层逻辑：以风险换预期价值，以及谁在为优势买单
- Jordi的第一性原理框架是：“你总是在用风险交换预期价值”——本质上，只要你持续找到想降低风险、愿意为此让出预期价值（EV）的人，你就能赚钱。创始人IPO、锁定期股东、解锁事件——凡是存在结构性因素的地方，都可以设计交易，包括买入锁定股并进行对冲。Selini自己的做法，是为交易所提供紧密流动性，换取留存费，同时在基础设施上投入数千万美元——使用从纽约连到日本AWS数据中心的微波数据链路。
- 反向交易定义了加密货币：赌徒放弃EV去买波动，而不是降低风险——“他们只想要那个1,000倍，愿意让出优势来获得波动。”最成功的加密货币产品卖的正是这一点，和彩票一样。主持人举了Yankees的50/50抽奖：一半给慈善机构，一半给1名赢家，“这是这个星球有史以来最差的彩票”，但每个球场仍能筹到25万–50万美元，因为好玩。于是Robinhood把交易游戏化：“向交易里注入的多巴胺越多，能抽取的优势就越多。”主持人一贯的警告是：“不要用手机交易。”
- 面对Jane Street，微观结构优势仍在：永续合约的用户体验很好，但“你永远不知道第二天的资金费率会是多少……公式只会告诉你下一个窗口，不会告诉你再下一个。”随着永续合约未平仓量扩展到SpaceX并进行24/7交易，“这是我们的世界……他们进入了我们的地盘”，Selini的机会将大幅扩张；Jane Street开始进入ETF级别的博弈，可能只会“削掉几个点”。
- Jordi自己在Kalshi（可能是Kalshi，听到的发音像“Kashi”）下注世界杯，暴露了散户如何被收割：界面把你引向市价单，把被动限价单入口藏起来——以14美分而非15美分成交，需要“多点几下”——平台有意将对价格不敏感的赌徒分层，类似Coinbase与Coinbase Pro。复利效应在于，如果5%的优势损耗“会让你从盈利变成不盈利”，它就会改变你的整个轨迹——就像2021年的交易者为40万美元利润支付5万–10万美元的Ethereum Gas费，“几年后反正还是会把这些利润交还回去。”
Guest
How do you create value? What is finance? How do you create value with trading and markets? There's someone who wants to be risk-averse with an asset that they have. They're willing to give edge to secure the bag. That's always the case.
So there's always money to be made by founders or shareholders that have all their money in the company they founded. They want to sell it; they want to IPO it. And there's this meat on the bone for everybody to eat. There's the banker who's going to make money, and there's the guy who's going to buy it.
And why are they giving up all this money? It's just because, at the end of the day, you're always exchanging risk for expected value. And all the investors, whether they're doing credit or equities—it doesn't matter what they're doing—are, in essence, going to make money if they consistently find someone that wants to reduce risk and will give EV for it. So if you understand that that's really the primitive of the game, then you start finding all the ways that that kind of materializes.
Today, we're interviewing Jordi, who is the founder of Selini Capital, one of the sharpest minds in the space. I've known Jordi for a long time. We've been in a ton of trades together, and he always has interesting things to say. Most importantly, we're filming this at his conference, the Selini Summit.
1. The Great Filter: Crypto's Washout Is Healthy
Okay, you know what? This is phenomenal. We are actually in Cannes right now at the Selini Summit, which is an amazing conference that Selini and Jordi put on every year. It is just in the most beautiful setting. We're currently in the Carlton Hotel. Thank you for joining me, and thank you for hosting the summit.
Guest
Yeah, as proper vampires, even though it's nice and sunny outside, we're all inside. So it's okay. We are quite literally in a dark room. It might seem like it's light, but that's just because we have all this professional lighting around us. Yeah, we're NEETs at the end of the day. They're like, “Why go outside?” “No, we'll go later. We'll go later.”
We've got some fun stuff planned, right?
Guest
Yeah. What's been the craziest thing you've heard today?
I don't know. I actually think maybe the most fun thing that I've heard is that we interviewed Alex Karp. He was talking about his time over at Palantir, and he basically gave us the blueprint for how Palantir found Osama bin Laden, which was kind of fun. I think that was pretty cool.
It has absolutely nothing to do with markets. In fact, he transitioned into saying that Palantir is probably a short at these prices. But, Jordi, what have you found most fun or interesting—from the talks, from the people, from what's going on at this conference? Anything that struck you as different?
Guest
The quality of the people left in this space has gone up because there's such a washout happening of all the lower-tier people. I think they can't even get funded now. That's the most positive thing. There's a big filter. The Great Filter is underway.
I think that's what's really going to create a foundation, because the reality is that over the last years, it's just been one terrible idea after another, with bad founders and bad ethics. Now, the only things that are getting funded—because the only people who have money left are very disciplined about it—are things that at least make sense, or where the founder is giving it a proper try.
Clearly, there are some good things. I saw Prime Intellect, one of our portfolio companies, announce a $300 million raise or something.
What is Prime Intellect?
Guest
They're doing crypto AI. It's a category that could be anything, right? It could be anything that uses both technologies. It doesn't need to be constrained within AI. Your panel was a crypto AI panel, and one of the questions was, “What's it going to take to get a top-10 market-cap coin?” People were saying that they could definitely get into the top 10 and maybe even be something bigger than Bitcoin, because it has real usage if it's done well. But we're still figuring out what is a crypto AI scam and what is potentially going to be promising.
It's funny. I asked the people on the panel that I was moderating, “Is anyone willing to stand up for any crypto AI project?” So, like, Virtuals, Vana, Grass, and TAO. Basically, everybody demurred, and they were like, “I don't know.” I'm very bullish on AI, and I'm very bullish on crypto, but it was kind of hard to get people to join the 2, I actually think.
2. Retail Is Gone
Guest
Yeah, Grass has been a very interesting one. We've been seeing the reaction to the investor call where they said, “We were making $50 million in cash.” The token sold off by a third. What kind of shitcoin goes down by a third in a single day on good news?
It kind of shows that the liquidity in crypto for altcoins is bad. We've talked to the founder a bunch. Andre says that, in many ways, if he were an equity play, he would be valued at some very high multiple. And in crypto, imagine trading at a $380 million FDV, with your kind of growth rate, when you're on track for massive numbers. This doesn't make any sense.
I think a lot of the reason is that retail has been so rinsed by the structure of the tokens, where Binance kind of forces everyone to be locked for a year. They want to get their float, and then after a year, they're just a bunch of VCs that are down bad. They have to sell whatever's doing okay, whatever liquidity they can get.
And then you just retail those tokens. Inevitably, those tokens just get dragged down. Retail has just kind of gotten tired of it. Retail doesn't have any money.
3. Selini Is Trading SpaceX: Same Playbook As Crypto
Retail's kind of gone in a way.
Guest
They are gone.
You run a market-making firm, anyway, so you'd be able to see the flows.
Guest
Yeah.
But I assume that retail flows, altcoin flows, are down massively.
Guest
Massively.
They've all gone, presumably, to equities, or they've zeroed out, or they're sports betting, or they're on Polymarket. I mean, where—
Guest
I think it's all of the above. There's a distinction between betting and the people who want to just take concentrated bets on a specific name. There are those kinds of people. They want to bet on this name, that name. Those are going to the “I'm going to do Micron, SanDisk” crowd. They've moved on to other things.
Thankfully, we're capturing some of that perp flow, potentially.
Are you doing this now? Is Selini trading equities?
Guest
We are trading the crypto version of equities and hedging it with some of the TradFi side, but we're trading on—it's not just on Hyperliquid. You have Binance, OKX. There are large amounts of liquidity on Lighter, even Bitget now. A lot of the exchanges are getting into these assets, and liquidity's not bad.
So Selini's exiting crypto.
Guest
No, we're trading crypto, really, the same as always. But if they're adding assets that are trading—SpaceX was trading $2 billion a day on Binance. $2 billion. I mean, that's not small. We can make some money on that: the spreads, the flows, everything.
I was never an equity guy. I was trading fixed income, commodities, and FX. This single-name thing is kind of new to me, but getting into it, yeah, because that's where the game is being played. Honestly, there's alpha there. It's not that hard to find alpha in SpaceX or something.
It's the same things we learned in crypto. We learned the playbook. First, you've got to think about the whole thing. There's the KOLs. They've got the KOLs. They're, in this case, Elon and whatever people on CNBC. There are all these KOLs, and they're bullshitting some story. They affect retail. We've seen that.
We've seen the unlocks. SpaceX is going to go through a ridiculous amount of unlocks—just in August, and then the next one, the next tranche, at the earnings call 2 weeks later. We're tracking all this stuff. We know how this plays out. We've seen this in crypto.
There's actually 1 extra part there, which is that the Binance Launchpad equivalent is getting an allocation to the IPO through Goldman or through one of these other investment banks. Then everyone is just able to flip it now because they're pricing these things pretty low.
Guest
Yeah, it's kind of nuts.
And it really does—I mean, we always would joke about the fact that crypto is going to become more like TradFi, and it has. Then TradFi has become like crypto, and it almost seems like you guys should be in there cleaning up.
As an individual trader, that's all I'm thinking about now. I'm thinking in terms of—trading's almost gotten easier in equities, in a sense, for people who are willing to chase narratives and willing to put up volume on it. The way that we did it in crypto is you just park your capital in things that you think are going to catch a bid, because you know that attention is going to go there at some point in the future.
Guest
Right. And you're not really sure when.
Mhm. You don't actually know exactly: 1 month, 2 months, 3 months.
Guest
Yeah. Even for the most sophisticated of us, we never predict, “Oh, this is the day that it's going to happen.”
It's just like, over the next 1 month, there's a 75% probability that this particular sector catches a bid because there's a hot ball of money running around. And that's kind of what's happening, I think, now in equities, although the hot ball of money is basically just in all these AI stocks and some downstream stuff.
I'm curious: What are you seeing in terms of flow in the crypto markets? People are talking about it right now: Maybe AI is down a little bit over the last few days, and maybe there's going to be a rotation where people cash out of AI trades and put it into crypto. Are you seeing any of that?
Guest
Well, I was going to say that has to play out slowly. First, you need people to exit the liquidity that they have, exit their SpaceX, exit their memory stocks that are up infinitely—all the Koreans trading KOSPI and up so much. They need to exit. They're not going to immediately put it in crypto.
But as things pick up momentum, it's a momentum game. With crypto, it's always been momentum up and momentum down. We have no momentum right now; it's all sideways. We've kind of bottomed, I think. It's sort of making this choppy bottom, and we'll have a very sudden move up. We just don't know when it is.
4. Passive Flows Always Hold The Bag
It'll all happen for no reason. It'll be like, “What? We're up 5–10% today?” And then the next day is like, “Why are we up 10% again?” Well, because yesterday we were up 10%, so now people are piling in. We'll get this reflexive effect where people start believing again that they can do well and make money.
Sadly, investors are still just chasing FOMO. It's still going to be the same thing. I don't know when that's going to change. It's either that or just stupid passive flows piling into automatic strategies.
I think SpaceX might be the first time that gets exploited to such a degree where the Nasdaq is about to pile in a bunch of money into a stock that just got listed with a low float. It's not the S&P 500, which is, I think, 13 or 14 times bigger than the Nasdaq, but it is a decent amount of money that's just going to go from Nasdaq holders into a single stock. These passive flows are always going to be the ones that, in the end, hold the bag, and these automatic flows are so easy to predict.
I don't know if you saw the story about Millennium making $2.8 billion, $3.7 billion, or $2.7 billion on index rebalancing.
Yeah, there was the Millennium pump—$2.7 billion on index rebalancing. What does that even mean?
Guest
Well, I'm sure these guys are smart. They're doing it, but are they that smart that nobody else can figure out how to do an index rebalance? I don't think so. I think having a large amount of capital and just having attention on this specific kind of mechanical thing, where you have dumb-money flows moving and you just front-run them, is the whole game.
Obviously, in crypto, sometimes we have Michael Saylor be the dumb money. He's, in essence, like the Nasdaq that just passively puts money in. We saw how many funds in crypto were trying to buy before him and sell after him. When he stopped buying, everybody had to exit that, and we had a bit of a dip there.
I think that's all these guys are doing. They're just tracking the mechanics of how these things work and piling a ton of money in.
Well, I think this is why it's so good to be a retail trader right now and actually very good to be nimble and small. I view the bear market—the scary bear market—as a literal bear that's chasing you and chasing everybody else. You don't have to be the fastest person on the pitch; you just have to be faster than the slowest person.
5. Surviving October 10th & The Leverage Lesson
The slowest people are all the pension funds, passive capital, and index funds flowing into this. You kind of just have to outrun them. Once the quote-unquote bubble pops, you'll probably be okay.
Now, the hard part is that nobody has any idea when the quote-unquote bubble is going to pop, right? So you sell Intel at $110 yesterday, you wake up today and it's at $17, and you're like, “What am I doing? Why did I sell?” That's the tough part that people are trying to navigate right now.
I don't know if you have any heuristics for that.
Guest
Our heuristics are always like this: We're not going to make the most return on any specific trade, or any specific month or even year, but we always keep a good amount of cash. We keep a good amount of beta that we think is smart, and just compound over time.
They really try to kill you in crypto many times. You have liquidation events and things like that.
Was October 10 a good day for you?
Guest
No, it could have been. Some of our teams did spectacularly well, some of the arbitrage teams and stuff like that. But some of our trades were a bit undercapitalized because we try to run—we want to make a return on capital.
Let's say if we run 1:1, fully collateralized, on some trades, it only makes a 10–15% annual return, right? Doing a lot of these trades by making a little bit of spread. Maybe if we run it at 3–4x leverage, we make 60%. So we were running 3–4x leverage, and some accounts just couldn't be rebalanced fast enough. You end up one-sided and get liquidated.
I think we ended up losing single-digit millions, which is not a huge day for us, good or bad. It's always capital—how much leverage you want on your capital is a decision for us. It's less safe if you push it a little bit to get a higher return.
So it was a mixed day. It could have been a very good day if we had everything dialed in properly, but those are the things that happen once every 3 or 4 years.
Right. And your systems aren't fully prepared for that. Your backtest isn't going to catch this because it hasn't happened before. You don't have this level of liquidation in the data.
Guest
It was funny hearing all the—I woke up and had 50 Telegram messages from people asking, “Are you guys okay? Are you guys okay? Are you guys okay?” I'm like, “What happened?”
People just assumed that some firm, either Wintermute or Selini, would have gone bust or something. That was the only explanation.
By the way, that's everyone's favorite thing to do all the time. The number of times that Wintermute has died is, I think, more than 20.
Guest
Yeah, they've died a lot, actually. We've done it maybe 2 or 3 times.
Yeah, you have a lot more deaths to go.
Guest
To catch up.
So any time the market goes down, I need all the listeners to tag Jordi on Twitter and say, “Are you alive?”
Guest
Yes.
Let's meme that into existence.
Guest
More of that. Haven't had enough of that.
Have you permanently shifted the amount of leverage that you use now, post-October?
Guest
I mean, I would say yes. Also, in some cases, instead of segregating it into subaccounts, you try to just have more unified liquidity. Exchanges are making it a little bit easier with portfolio margin.
I think 10/10 did show that if you're using coins for margin—even if they are USDe or whatever, wrapped SOL or whatever—all that stuff that temporarily de-pegs will liquidate you, and it's pretty aggressive. We don't really use non-dollar coins as collateral. It just seems like a recipe for disaster.
Maybe Bitcoin you can use to a certain extent for collateral, but I would say we use leverage, but we've gotten smarter about how we do it and just a little more conservative with it.
So, just no more coin collateral?
Guest
You can, but you shouldn't, because it will suddenly mark it down and liquidate your whole subaccount.
It's amazing. I think crypto also taught us a lot about how to take advantage of market microstructure. There's a lot of that in crypto, and I just don't think there were enough people taking advantage of it.
6. "They're In Our Land": Where Selini Still Has Edge
I've kind of been out of that game for a little bit, but one thing I was thinking about the other day is: Does that inefficiency still exist in crypto, or are there things that you take advantage of on a day-to-day basis? Or have the Jane Streets come in?
Guest
There are. Yeah. Even things like perpetuals, which are becoming very popular, are a good user experience. You don't have to trade futures, roll them every 3 months, and figure out how to get the spread and the basis with them. Perpetuals are a very amazing UX. You just put it on and leave it, right?
But you're dealing with this funding rate, and you never know what the funding rate is going to be the next day.
It could be anything. You can't predict it. And that's the big disadvantage compared to a calendar future, because a calendar future is just going to expire on a specific day. You don't have to worry about any of the funding until that day arrives. You really just focus on the expiry date. There are shenanigans happening around the expiry, but that's it.
With perpetuals, it's just this constant guessing about what the funding is going to be.
Well, there's a formula.
Guest
Well, you don't know what the basis is going to be between the spot and the perp tomorrow. You just know what it is right now. The formula just tells you the next window, not the one after that.
There are things like this that the Jane Streets and stuff have no idea about. I don't think they do. This is our world. They're in our land. As perpetuals get bigger and OI on perpetuals includes SpaceX and all these other trades, I think we understand it much better than these guys, and we'll continue to have an edge on this kind of microstructure stuff.
Well, doesn't that mean it's great for you as a company, then?
Guest
Yes.
I mean, if that market expands massively, then presumably your ability to make money also expands massively.
Guest
Yeah.
Especially if it trades 24/7.
Guest
It trades 24/7. And there are these huge overlaps now where the SpaceX market is so big. It's much bigger than the Ethereum market or whatever. If you start finding opportunities to buy locked shares, but then you kind of know how to hedge, there are a bunch of interesting structural things that you can do—
No, no, no, no. Keep going.
Guest
No, no, I can't.
Don't stop there. Keep going. What do you do? You buy locked shares?
Guest
So, in essence, if you buy locked shares, how do you create value? What is finance? How do you create value with trading in markets? It's always that there's someone who wants to be risk-averse with an asset that they have. They're willing to give edge to secure the bag. That's always the case.
So, there's always money to be made by founders who are shareholders and have all their money in the company they founded. They want to sell it; they want to IPO it. And there's this meat on the bone for everybody to eat. There's the banker who's going to make money, and there's the guy who's going to buy it. And why are they giving up all this money? It's just because, at the end of the day, they're de-risking out of it. They're willing to give EV—to give you an expected value that's positive—because the utility is not just the EV; it's also the risk. So they're taking risk off. You're always exchanging risk for expected value.
All the investors, whether they're doing credit or equities—it doesn't matter what you're doing—are, in essence, going to make money if you consistently find someone who wants to reduce risk and will give you EV for that.
So, if you understand that that's really the primitive of the game, then you start finding all the ways that that kind of materializes. You can do that with crypto, you can do that with equities, you can do that with a combination of the two. At the end of the day, anytime there's an unlock happening, anytime there's something structural there, you can structure something.
7. The Casino Economy: Variance, Dopamine & Getting Fleeced
Obviously, we do a lot of that. We provide tight liquidity where we just make a tiny bit of spread. As a company, we make a lot of money by working with exchanges that need liquidity. They'll sometimes pay us a monthly retainer to make sure that their books are liquid, and we don't just make the money and keep it. We have to spend a lot of money on infrastructure. We have to get the fastest feeds, microwave feeds from New York, and send them to Japan, where the AWS data centers are. There's a bunch of costs. I mean, we spend tens of millions of dollars just on infrastructure and networking. Our networking bill is massive. Then you have to pay all the smart traders to sit around.
Right.
Guest
It's a very high-intensity business. And obviously, like you said, the Jane Streets of the world step into more of the huge games, like the ETF-level games, and we'll try to probably shave off a few points.
Probably shave off a few points.
Guest
Probably shave off a few points. Yeah.
Okay. I mean, I think you—I want to repeat this because it was so beautifully put: you trade with people who are willing to let go of edge because the utility to them of whatever capital they have is reasonably high.
And that's actually probably a big opportunity for people who got very wealthy in OpenAI, for people who got very wealthy in Anthropic, SpaceX. But I go back to crypto and say that was kind of the whole market in many ways.
Guest
So, in crypto, you have both sides. You have a lot of people who are up 1,000x on Bitcoin, ETH, or whatever their bags, and at some point they want to distribute. So you can still find value if they're distributing.
And then it's funny: in crypto, and in general in speculation and gambling, you have the opposite, where the EV people will give is not to de-risk but to pro-risk. They just want that 1,000x, and they're willing to pay edge to have the variance. They want the variance, right?
You point to the camera and go, “You hear that? 1,000x.”
Guest
[Laughter.] Yeah, the more successful products in crypto have been the ones that are just giving people the chance to get that extra variance. It's like the lottery. The lottery has been very successful for, I don't know, hundreds of years, right? It just gives people the chance to change their life, and they'll exchange it for some edge. And we have a lot of that as well.
And that edge is entertainment, and that's kind of the thing of our whole industry in many ways. Not to repeat myself with the crypto thing, but it strikes me because I go to Yankees games a lot. I'm a big baseball fan. One of the things that they do there is hold this thing called the 50/50 raffle, which is maybe the worst-EV thing that I've ever seen in my entire life. The way that it works is you buy a ticket, 50% of that ticket gets donated to charity, and then 50% of that ticket gets distributed to one person who won a ticket.
You're actively making it the worst lottery that's ever existed in the history of the planet. You're literally immediately losing half the value, right? It doesn't make any sense. But it is play. I mean, it sometimes gets up to $250,000, $500,000 in a stadium of 50,000 people.
Guest
Mhm.
You're like, “Well, why are they doing this?” And it's because it's fun.
Guest
Yeah.
And that's actually why you see Robinhood make trading fun. That's why you see all these games. That's why you see the gamification of finance, because the more you make it—the more you introduce dopamine into trading—the more edge you can actually extract from your average person, right?
Guest
And that's something that I think we actually—
I'm trying to warn against a little bit. Which is why I say one of the things that I always say on the show is: do not trade from your phone.
Guest
Like, just don't do it.
Yeah. It actually activates neurons in your brain. You're like a monkey: “I want to see the flashing lights.” Trade from your computer if you're going to trade. But that's something that I think about a lot, and it's also kind of inevitable.
Guest
A lot of these things are inevitable.
You might as well take the other side. Yeah. If it's inevitable, you take the other side.
Guest
I warn against gambling. I've spent more hours than I think most people in casinos in my life. I was a professional poker player. I spent a lot of time even playing blackjack semi-professionally, and sports betting as well. I really understand the psychology of the dopamine rushes. I've had friends who have just gone crazy and started martingaling the shit out of their bankrolls.
Foolproof strategy of infinite funds.
Guest
Yeah, it almost never—
[Laughter.] Almost never fails.
Guest
But on the other hand, you can't stop people from being monkeys, and at some point you just have to understand the microstructure and take advantage of it. I was doing some World Cup betting the other day. I went to Kalshi.
Edge?
Guest
No, no. Well, I could. I certainly could, but I don't spend any time on it. So I understand that this is for entertainment. If I'm going to watch the game, I just want to hope for a specific outcome. If it's Colombia–Switzerland, I love those countries, but I don't really have a horse in that race. If I bet on Switzerland to win in penalties, then I'll certainly care a lot more.
But the trader in me can't switch off. So I go to Kalshi and I want to bet, and I realize, first of all, they are directing you toward market orders. They want you to market order. They don't want you to place a passive bid.
If you just say, “Okay, I want this to happen,” of course, that’s intentional. They’re segmenting their customers in a way to get the least price-sensitive gamblers to just rip into the book, because that’s what they like—it makes fees, and they can charge the most. They realize you’re not a professional if you’re just ripping it in without even checking the liquidity.
They’re getting very smart about this. They’re forcing people like me who want to get 14 cents instead of 15 cents to figure out how to do it. It’s not that simple; you’ve got to really find it.
Yeah. I mean, this kind of reminds me of Coinbase Pro versus just regular Coinbase, because it’s the same concept as, like, let’s fleece the people that don’t care.
Guest
Yeah.
Right? And that’s my advice to you and anybody watching the podcast: make sure that you’re not the person getting fleeced, because somebody out there is. The best thing that you can do for yourself is just double-check that.
Guest
Yeah. Think about how much edge you’re losing on fees on any trade or on spread—just have a sense, because it compounds. You might have the right idea, but every time that you’re doing it, if you’re losing 2%, it adds up.
I think it’s actually hard for people to conceptualize this.
Guest
Mhm.
That’s because I think the fees of each individual trade are so low that they don’t factor into your decision analysis, but over time, obviously, that really, really, really stacks up.
Guest
That’s what people around me have told me for a long time now: only once they spoke to me and really thought about it. When they’re meme-coin trading, they’re going for the 100x, 1,000x. They don’t think about, “Oh, am I paying 5% or 10%?” They think, “Whatever, on this thing, it doesn’t matter.”
But you’re not always going to be right, and if you know that 5% takes you from being profitable to unprofitable, it just changes your entire trajectory.
Right. And that’s why there was so much edge back in the day as well, because I remember back in 2021 people were posting, “Oh, this is how much money I spent on gas on Ethereum,” and it’s just actually outrageous.
Guest
Yeah.
It’s—I mean, it’s tens of thousands of dollars, sometimes $50,000 or $100,000 of gas spent to make $400,000 of profit that you end up giving back 2 years later anyway.
Guest
Yeah.
So it’s like, that stuff actually really matters. And transaction fees—well, I’m not going to harp on them too much, because they sort of pay a lot of people’s bills.
Guest
Yeah.
But it is something to watch out for.