Lao Bai: This Is Crypto's Last Decent Bull Market! Three Main Themes: RWA, Perp, Privacy

The market has entered the "post-narrative era." The low-hanging fruit has been picked, and the competitive focus has shifted from "storytelling" to product-market fit (PMF) and cash flow.

Author: Mr. Z, 168X

The low-hanging fruit has been picked: Alpha is shifting from narrative to cash flow, distribution, and PMF

Sept. 23, 2026, 7 p.m. Crypto is warming up again. Robinhood, perpetual DEX, privacy-focused projects, and a batch of real-world assets (RWA) are starting to attract capital again. But the most obvious difference in this market cycle is that it's hard to find a completely unfamiliar new story like the 2017 ICO, the 2020 DeFi Summer, or the subsequent NFT and GameFi waves. The easy fruit is becoming scarcer, and the market is shifting from "what story to tell" to "who has really found product-market fit (PMF)."

This episode of 168X welcomes back an old friend, Lao Bai (@Wuhuoqiu). Recently, his main focus has been on Robinhood. In his words, he's "fishing where the fish are." This episode covers Bitcoin (BTC), Ethereum (ETH), HYPE, BNB, SOL, then moves on to Robinhood, stock tokens, artificial intelligence agents (AI Agents), GPU real-world assets (RWA), Fomo, perpetual DEX, Hyperliquid Improvement Proposal 3 (HIP-3), and then breaks down why the cryptocurrency primary market is almost dead. In the end, he has one conclusion: if this is the start of a bull market, it could be Crypto's last "decent bull market," and this cycle has only three main themes: real-world assets (RWA), perpetual contracts (Perp), and privacy-focused projects (Privacy).

1. ETH Position Will Be Higher Than BTC: The Long-Term Downtrend Line Has Finally Been Broken

Mr. Z: Let's start with Bitcoin and Ethereum. You just said that if there really is a bull market this cycle, would you personally lean more toward ETH or BTC?

Lao Bai: In terms of asset quality, I still think BTC is better, but personally, I will probably allocate more to ETH this cycle. I don't understand complex technical indicators like candlestick charts, but looking at the Ethereum/Bitcoin (ETH/BTC) weekly chart, that long-term downtrend line is very obvious. You can see it just by drawing a line, and it broke out around 0.03. If this bull market can continue for a few more months, I don't believe it will break through such a long-term trend line and then immediately fall back. So my position allocation should be more ETH than BTC.

Mr. Z: What about BTC this cycle? Many people now give three scenarios: bearish, neutral, and bullish.

Lao Bai: In the bullish case, I see $200,000. Neutral is around $150,000. Bearish is around $120,000, roughly back to the previous cycle high. If BTC can reach $150,000, then since ETH has already broken its long-term downtrend against BTC, I think its gain should be 50% to 60% more than BTC's. But how high ETH can go fundamentally depends on how high BTC can go. It's hard for ETH to lead the entire market on its own.

2. The Low-Hanging Fruit Has Been Picked: The Only Clear New PMF This Cycle Is RWA

Mr. Z: My biggest feeling this cycle is that there's no disruptive new thing like the 2017 ICO, the 2020 DeFi Summer, or the later NFT and GameFi waves. It seems like a truly new story is missing.

Lao Bai: I've said many times before that the low-hanging fruit has been picked. In the past, Crypto innovation was pushed outward from technology. We first invented ICO, then DeFi, researched how to improve TPS, and then pursued various new technical routes. At that time, we could temporarily ignore data, users, or real PMF. As long as there was a technological breakthrough, the market was willing to give a lot of room for imagination.

Lao Bai: But in this cycle, the technical routes have basically all been tried, and PMF has basically been explored across every sector—gaming, privacy, ZK, Layer2, finance, sports, cards. Everything that could be explored has been explored. I think 80% to 90% of the directions are roughly settled. What's left is for everyone to focus on a few directions and continue to break through, truly finding PMF.

Mr. Z: So what do you think has actually been found now?

Lao Bai: The clearest one is RWA. Perp is more of a continuation from dydx and GMX, except Hyperliquid has made the experience better and expanded the space further by combining HIP-3 with RWA. What's really newer is stock tokens, the coin-stock pairing brought by Robinhood, and the form of buying stocks on-chain. Privacy, I even think, only counts as half a track. It's related to the narrative that institutions need a certain degree of privacy after going on-chain, but this cycle it has risen so well that everyone has been educated by the price.

3. How to Choose Among HYPE, BNB, and SOL: Three Sets of Odds—Stability, Imagination, and Elasticity

Mr. Z: After BTC and ETH, what do you think about SOL, HYPE, and BNB?

Lao Bai: The most stable ones, I think, are HYPE and BNB. BNB has been around for too many years, and Binance is the absolute leader. It's a bit like Coca-Cola in beverages, or NVIDIA in AI. It's definitely very stable, but the upside won't be as large as some newer assets.

Lao Bai: HYPE is relatively just the leader in the Perp niche. Although it's already very expensive, if Hyperliquid really achieves its originally envisioned state—permissionless trading of everything—and HIP-3 continues to expand, and HIP-4 launches with good data, then its theoretical upside is still larger than BNB's. And in the short term, I don't see a particularly clear reason for a decline.

Lao Bai: SOL, on the other hand, has the most elasticity among the three. It went from $2 in the last cycle to $250, then fell back to single digits, and then rose back to over $200. This chain has a very strong grassroots base. The Foundation, developers, and a group of Western Solana KOLs and users have always been very loyal. Its biggest advantage and disadvantage are actually the same thing: after FTX collapsed, there is no centralized exchange providing an absolute exit path, so many things are done by native on-chain DEXs themselves. This actually makes it very much in line with the feeling of Crypto Native, cyberpunk, and grassroots power.

Mr. Z: So if you had to summarize simply?

Lao Bai: BNB is certainty, HYPE is imagination, SOL is elasticity. I also hold some SOL myself. In the short term, over the next one or two months, I'm looking at around $150. But whether it can truly break its previous high ultimately depends on BTC.

4. Robinhood Is the Endgame for Exchanges: One Account to Trade Everything

Mr. Z: Let's come back to Robinhood. You previously wrote that "the ultimate form of all exchanges in the world is Robinhood." What exactly does that mean?

Lao Bai: Robinhood as a company, and Vlad himself, are quite special. They are one of the few teams that understand both stocks and Crypto native culture. The retail short squeeze culture like GME came out of Robinhood users, so they understand retail investors, market makers, Crypto, and stocks.

Lao Bai: Look at how Robinhood Chain has developed—it fits their style. At first, they used Memes like CASHCAT to grab attention, then quickly found Launchpad, paired Memes with stocks, and then developed various new things on top. At least on Robinhood, I've seen many new plays that don't exist on traditional Layer2s.

Lao Bai: I said before that the ultimate form of all exchanges in the world should be Robinhood. At that time, Robinhood's stock was only $76, but I didn't buy it myself. Later, I bought it at over $100. I finally practiced what I preached.

Mr. Z: So you mean traditional stocks, Crypto, Perp, and prediction markets all in one place?

Lao Bai: Yes, an exchange for everything. Traditional stocks, Crypto spot, Perp, prediction markets—whatever you want to trade, I'll give it to you. If all other derivatives are added, it's basically an Everything Exchange.

5. What Will Stock Tokens Ultimately Compete On? When Everyone Can Access US Stocks, the Advantage Returns to Top Exchanges

Mr. Z: Now offshore exchanges like Bitget, Gate, and Binance are also starting to offer stock tokens. When everyone has them, what will these exchanges ultimately compete on?

Lao Bai: I think it's user experience and resources. With stock tokens, if everyone ultimately connects to brokers like Alpaca and then completes trades on Nasdaq or the NYSE, then the same stock is hard to differentiate. You're buying NVIDIA on every platform. The real differences are only the frontend UI, UX, fees, and where you already keep your assets.

Lao Bai: So the earliest platforms to launch can grab the first batch of early adopters, because others don't have it and you do. But once everyone is connected, the advantage returns to top exchanges. If my funds are already on Binance, and I want to trade both Crypto and stocks, and Binance still has the best coin selection and depth, then I'll most likely keep my assets on the top platform.

Lao Bai: Unless a user only plays stock tokens, in which case they can completely ignore Crypto depth and liquidity and just pick the platform with the lowest fees or the one they feel is safest. When everyone has the same asset, the asset itself is no longer a moat.

6. How to Find Alpha on Robinhood: First Understand 5 to 10 Leading Assets, Don't Scan the Whole Chain

Mr. Z: If someone has never researched Robinhood before and suddenly jumps in now, what's the first thing they should understand?

Lao Bai: First understand the leading assets issued on different Launchpads. For example, which are the leaders on PONS, which are the leaders on LONG(), and understand the trading patterns, trends, and what three to five major assets actually do. I think researching 5 to 10 coins will give you a good idea of what this chain is doing. You don't need to scan the whole chain randomly at the beginning.

Mr. Z: Can it also be divided into two approaches? One is more like stock valuation—looking at protocol revenue, market cap, TVL. The other is purely capturing attention and narrative.

Lao Bai: Yes. Data-driven people look at AI and PONS, where you can clearly see daily burns, TVL, trading volume, and various dashboards. You can roughly estimate what something should be worth. But for new things like SHROOM, DELTA, HOOKR, ORBIO, ZZZ, you often can't estimate. It's just attention and narrative.

Lao Bai: Like ZZZ—once everyone confirms that some impressive Founder or Dev is associated with it, the price might double immediately. But whether it has actually built much at that point is uncertain. So on Robinhood, you first need to distinguish whether you're buying a business with data that can be valued, or a pure Attention Trade.

7. The Biggest Problem with AI Agents: To This Day, No Product Has Made Me Comfortable Putting Large Amounts of Money In

Mr. Z: AI Agents have also been hot recently. Mojo AI, Travix, SmartX, and others are emerging. Do you think AI Agent Trading really has PMF now?

Lao Bai: I haven't seen a particularly mature, truly breakout product yet. I've used so-called AI Trading Platforms myself. I put in a few hundred dollars, set my own strategy, and let it run. Overall, the experience still has many things that feel tricky to me.

Lao Bai: Sometimes the Agent misunderstands your intent, sometimes it doesn't fully execute the framework you set, and there can be bugs. The bigger problem is that you don't dare entrust too much capital to it. I might be okay giving it $500 or $2,000, but if I had to put in more than $10,000, there's currently no AI Agent Trading product that makes me feel comfortable.

Mr. Z: So what do you think is a good combination of AI and Crypto right now?

Lao Bai: It's actually not trading. One is VVV, which does decentralized private inference. If you have data or questions you don't want to make public, or don't want OpenAI or Anthropic to know, you can use this privacy-focused version of OpenRouter. I think this is a very clear, hard need.

Lao Bai: Another one is ORBIO, which turns AI inference credits into an economic asset. The idea is clever, but I don't 100% buy in, because it ultimately depends on ORBIO's own trading volume. Trading generates fees, which are then used to buy inference credits on OpenRouter, and those credits are distributed to token holders. If no one trades this token, the tax is small, and the inference credits that can be bought are also small. So essentially, there's still a layer of Attention Economy flywheel.

8. The Real Winner of the Agent Economy Might Not Be Any Specific AI Coin

Mr. Z: So what is the real intersection of AI × Crypto? If the Agent Economy really arrives, who will capture the most value?

Lao Bai: I've actually thought about this question for many years. In late 2023, I talked to over a hundred AI + Crypto startup teams—various decentralized inference, ZK verification, game theory. Basically, they were all holding a hammer looking for nails. After three years, very few have truly impressed me as a good combination.

Lao Bai: Instead, what Crypto has already done well is the payment rail that AI Agents will need most in the future. We have USDC, Ethereum, and Solana. When the Agent Economy truly emerges, agents will most likely use stablecoins to complete payments and transactions on some public chain.

Lao Bai: So the ones that truly capture the value of the Agent Economy may not be any specific project called an "AI Token" today. It might instead be Circle, or public chain infrastructure like Solana or Ethereum. X402 is similar—it's just one possibility for Agent Payment. In the future, there will be Google's, Visa's, and Stripe's own protocols, and various on-chain, off-chain, and semi-on-chain options will emerge.

Lao Bai: AI Agents will likely need Crypto in the future, but needing Crypto doesn't mean needing a specific AI coin.

9. GPU Can Also Be RWA: USDAI and Dawn Are Betting on "Semi-Standardized Assets"

Mr. Z: Besides stocks, you've recently been looking at GPU and AI infrastructure assets going on-chain, like Dawn Protocol and USDAI. What do you think of this?

Lao Bai: Dawn can be understood as an expanded version of USDAI. USDAI is more focused on GPU collateral, while Dawn may expand to memory, edge computing, and other revenue streams. The underlying business model is basically the same: using AI infrastructure-related assets for collateralized lending and project finance.

Lao Bai: I would divide RWA into three types. Treasury bonds, stocks, gold, and crude oil are standardized assets and the easiest to put on-chain. Previously, Centrifuge did music copyrights and various single cash flows—those are very non-standardized assets, and I think they're the hardest to do.

Lao Bai: GPU is in the middle. It's a "semi-standardized asset." Because there are only a few models like H100, A100, and B200, you can create a unified framework, but each model has different pricing, payback periods, and depreciation. So it can be used for collateralized lending, easier than fully non-standardized assets, but it definitely won't reach the scale of treasury bonds and stocks.

Mr. Z: So it has potential, but it won't become a particularly huge sector?

Lao Bai: Right. It's essentially still fixed income and project finance, earning the spread. But AI infrastructure itself will be very large in the future, so even if you can only capture 1% or 2% of the financing demand, you can still form a market of a certain size.

10. What Fomo Really Wins Isn't Apple Pay: It's Compressing the Time from Shilling to Ordering into a Single Moment

Mr. Z: Fomo has been very hot recently. My own feeling is that its most powerful feature is that an ordinary person can just press Apple Pay—no need to worry about wallets, cross-chain, or gas fees—and immediately buy a Meme. What do you think?

Lao Bai: I agree with that summary. But I think what Fomo really wins isn't just Apple Pay—it's Social Trading. You can see the leaders you follow, what they're buying and when, and that information is real-time.

Lao Bai: In the past, for example, if you followed Dayu (@BTCdayu) and he bought a coin, he might post his view on X only after some time. After you saw it, you still had to understand his logic, open a trading platform, and place your own order. Fomo compresses almost all of that time gap. The moment he buys, your phone pops up: "Dayu bought $200,000 of a certain Token," and you can follow immediately.

Lao Bai: In the future, it could even do Copy Trading directly. That experience is on a completely different level from "I read his tweet, verify it myself, and then place an order." I think this is what Fomo really wins.

Mr. Z: So it and X might eventually end up in the same place?

Lao Bai: Yes. Fomo is a bit like a simplified version of X. Once X has X Money, token candlestick charts, and trading, it will become more and more like a complex version of Fomo. In the end, it's about putting three things—shilling, viewing charts, and buying—on the same platform.

11. Latecomers in Perp DEX Have Only Two Paths: Attach to a Big Player, or Push Execution to the Extreme

Mr. Z: Hyperliquid is already so strong in Perp DEX. How can later projects differentiate?

Lao Bai: It's very hard. You either have a distribution advantage or particularly strong institutional capabilities. The two examples I've seen that actually work are Lighter and Variational.

Lao Bai: Lighter initially focused on ZK and zero fees, but that alone didn't create a qualitative change compared to Hyperliquid. Zero fees attracted a lot of farmers early on. What really took its valuation to the next level was when Robinhood later chose it as one of its underlying layers. You suddenly attached yourself to a big player, and an external platform is directing traffic to you. That's order flow Hyperliquid can't capture—that's differentiation.

Lao Bai: The other is Variational. You have a very strong institutional background and capital capability, and you can push slippage, execution, and institutional trading experience to another level, carving out your own market.

Lao Bai: So you can't make something "similar to Hyperliquid" and expect to steal its users. You must have a distribution channel it can't replicate, or push the execution experience for a certain user group to a level it's unwilling or unable to reach.

12. The Biggest Misconception About HIP-3: Grabbing a Ticker Is Useless—Tradexyz Can Copy You the Next Day

Mr. Z: What about HIP-3? Entropy, Tradexyz, and various HIP Deployers are all grabbing tickers. What do you think of this model?

Lao Bai: HIP-3 theoretically gives many teams opportunities, but the ticker itself has no moat. Today you discover demand for a stock and rush to list the ticker. Tradexyz can list the exact same ticker the next day.

Lao Bai: And it already has more users, deeper liquidity, and stronger market makers. For the same ticker, why wouldn't users go to Tradexyz? It's a bit like centralized exchanges—once Binance has the best depth, everyone naturally goes to Binance for the same coin, because the slippage and depth are the best.

Lao Bai: So I believe what HIP-3 Frontends should really compete on isn't tickers or liquidity, but Distribution and User Experience. If your frontend can offer a form that Tradexyz completely lacks, or if you control a certain group of users it can't reach, then you have a chance.

Mr. Z: So being first doesn't really help—you're just validating demand for the leader?

Lao Bai: Right. You list first today, Tradexyz sees you have volume, copies it tomorrow, and as soon as it launches, its liquidity immediately surpasses yours, and you're in trouble right away. Grabbing tickers isn't an edge—Distribution is.

13. Why This Is the Most Brutal Bear Market in History: VCs, Market Makers, and the Primary Market All Failed Together

Mr. Z: Why does this bear market feel so bad? I think it's completely different from before, especially with the primary market almost dead.

Lao Bai: That's the biggest difference. In the past, no matter how bearish it was, the secondary market was terrible, but the primary market was still innovating. In 2018 and 2019, VCs still invested in Uniswap, Compound, and other infrastructure that later powered DeFi Summer. In 2023, I joined ABCDE. At that time, I looked at an average of 4 to 5 projects a day, and on the most extreme day, I talked to 9. Over more than two years, I talked to over a thousand projects.

Lao Bai: But after the last cycle, a large number of Asian VCs shut down or shifted to the secondary market. Second- and third-tier VCs in Europe and the US also basically shut down. Only top players like Dragonfly and a16z still have money to invest. Small projects can't raise money, and small VCs can't raise from LPs. This bear market is especially terrifying.

Mr. Z: Why did the old VC model suddenly stop working? Was the anti-VC narrative of Hyperliquid a catalyst?

Lao Bai: Yes. The typical playbook of the past seven or eight years was: VCs invest in projects, projects issue tokens, list on exchanges, market makers control liquidity and distribution, then pump the price, and finally let retail investors in. For the vast majority of projects, price had little to do with fundamentals.

Lao Bai: But the problem is that there are more and more projects. Of the over a thousand projects we looked at, we ultimately invested in more than 40, and many of them still went to zero. The scale of VC tokens to be unlocked in the coming years is too large, but retail capital and attention haven't grown accordingly. Too many wolves, too little meat.

Lao Bai: Then Hyperliquid demonstrated another playbook. Revenue, buybacks, and burns are written directly into smart contracts. People started looking at tokens the way they look at stock P/E ratios. Combined with liquidity shocks, injured market makers, and the massive wealth effect of the AI market, these three things stacked together created this particularly miserable bear market.

14. Crypto Is Becoming More Like US Stocks: From Now On, Look at Revenue, Buybacks, and Cash Flow—the Four-Year Cycle Will Weaken

Mr. Z: If the old model of VC + market makers + narrative pumping has run its course, what will Crypto become?

Lao Bai: It will become more and more like US stocks. Now when people buy a project, they start looking at revenue, cash flow, buy and burn, and whether there are real users. You have to produce solid data for people to buy your token. In the past, many people valued projects using "dream multiples." Now it's becoming more like stocks.

Lao Bai: At the same time, US stocks are becoming more Meme-ified. Look at the rallies in some AI, semiconductor, and storage stocks this year—they're very similar to past Crypto market movements. So an interesting situation has emerged: Crypto is becoming more like US stocks, and US stocks are becoming more like Crypto.

Lao Bai: If the market really becomes like this, the four-year cycle of big bulls and big bears may weaken. Each project will be judged on its own revenue, growth, and data—rising and falling independently, like stocks, rather than all projects celebrating together every four years.

Mr. Z: So that's why you said earlier that this might be the last "decent bull market"?

Lao Bai: Yes, I'm not saying there will be no more bull markets. What I'm saying is that if we treat this as the start of a bull market, this could be Crypto's last decent big bull market. Because the technical routes have been mostly explored, and the PMF directions have also been mostly explored. After this cycle finishes hyping coin-stocks, Privacy, and Perp, what will the next cycle hype? It will likely return to lending, DEX, and Memes—recycling the mature stuff again.

15. Finally, Some Tickers: RWA, Perp, Privacy—What to Watch Next

Mr. Z: Let's be direct at the end. From a ticker perspective, what should people position in and watch now?

Lao Bai: The main themes this cycle are extremely clear. I think a consensus is gradually forming on X. The first is RWA, the second is Perp, and the third is Privacy. Whatever Launchpads eventually issue will also lean toward these themes.

Lao Bai: BTC and ETH don't need explanation—they're base assets. For Privacy, I'm currently watching Zcash, Zama, and NEAR. The popular narrative in the market right now is that Zcash is "privacy Bitcoin," Zama is more like privacy Ethereum, and NEAR is "what Ethereum should have been." These three are the more direct tickers in the Privacy sector.

Lao Bai: For Perp, it's HYPE and LIT, but neither is cheap anymore. HYPE especially is already very high, and I think further upside will start to get difficult. If there's a clear pullback in the future, I think it would be safer to buy, but at this level, I can't give specific price points.

Lao Bai: RWA is a bit awkward. ONDO, as the most typical tokenized asset project, has relatively weak value accrual for the token itself, so I don't like it much. More directly, I would look at UNI, because many of Robinhood's on-chain stock-related trades will go through it, and it can capture the growth from the combination of stock tokens and AMM.

Lao Bai: For AI, I would put VVV and Bittensor on my watchlist. Bittensor is the leader, but its narrative is stronger than its data. VVV's data is actually quite good now, but it has also risen a lot. If the price drops in the future but the data doesn't clearly deteriorate, that would be worth revisiting.

Mr. Z: So to summarize, this cycle isn't without Alpha—it's just that Alpha has shifted from "who can make up the next new story" to "who has really achieved PMF, who is really making money, and who can return value to the token." If this really is the last decent bull market, the most important thing may not be finding 100 coins, but looking deeply enough at the few pools that actually have fish. Thanks to Lao Bai for joining 168X again today. We covered a lot in this episode, and I hope everyone can find the opportunities most worth watching this cycle.