The Bull Market Question No AI Can Answer: XRP, ADA, and PI Through an Order Flow Lens

A 17% weekly gain in a token still down 73% from its high should feel like relief. For Cardano, it looked like a pulse: whales accumulating, traders returning, and a cluster of chatbots suddenly arguing it was the safest bet among three wounded assets. But I have stopped trusting relief rallies. The numbers didn't lie, but my trust did. Long before I learned to read order flow, I learned to read the silence of a dead treasury: a subtle reentrancy loophole in a contract I audited in 2017 drained $1.2 million in ETH, and the project vanished. Since then, when a media article tells me three AIs predict the next bull market winner, I don't see prediction. I see metadata. The question itself is a signal of where the speculative energy is hiding. The article in question was simple on its surface. CryptoPotato asked three AI models—likely OpenAI's ChatGPT, Anthropic's Claude, and Perplexity—to rank XRP, Pi Network's PI, and Cardano's ADA for the next bull market. The answers were predictable: one liked ADA's low dilution, one leaned XRP's institutional momentum, and another flirted with PI's 100x potential if the ecosystem finally lands. But the article gave us no contract addresses, no fee revenue, no validator data, no audit history. It gave us narratives. And after 18 years of watching markets turn narratives into liquidity, I know that narratives are the cheapest form of consensus. So let me slow down and look at the actual positions. XRP has dropped 65% in the past year. ADA is down 73%. PI is also down roughly 73%. Yet all three are being discussed as if the next cycle will lift them evenly. That is the first misconception. Bull markets do not lift assets evenly. They rotate through stories, and the rotation is driven by institutional order flow, on-chain activity, and regulatory arbitrage—not by how many Twitter polls a token wins. Ripple, the company behind XRP, has been doing something interesting. It acquired Hidden Road, a prime brokerage platform, which is not something a token project does when it plans to remain a retail-only pawn. It partnered with South Korea's KBank to expand crypto payment corridors. It secured a MiCA license, giving it a regulatory passport across the European Union. The SEC's lawsuit is in the rearview mirror. That collection of events forms a structural story: XRP is trying to become the settlement layer for institutions that want to bridge crypto and legacy finance. That is not a meme. It is logistics with a ledger attached. Cardano, by contrast, has a different profile. The token's supply is largely already in circulation, which means future inflation comes from staking rewards rather than large cliff unlocks. ChatGPT specifically praised this: ADA's dilution risk is lower than XRP or PI. That is a legitimate point. But lower dilution does not create demand. It only removes a supply-side headwind. The ADA rally we saw—17% in a week—was accompanied by whale accumulation, according to the article. Yet no data was provided on Cardano's total value locked, active developer count, or DApp transaction volumes. Without those, I treat the weekly rally as sentiment noise unless it is confirmed by on-chain activity. Then there is Pi Network. This is the one that makes my skin crawl. PI has one of the largest claimed community bases in crypto—something that sounds impressive until you realize that community size is not market demand. It is not on any major exchange. There is no reliable price discovery. The token's value is essentially a placeholder in a mobile app's internal ledger. The 100x potential prediction is conditional on two enormous unknowns: the mainnet ecosystem becoming real, and a top-tier exchange like Binance or Coinbase listing the token. Those are not catalysts. They are hurdles. And in the history of crypto, we have seen dozens of massive communities die waiting for a listing that never came. I need to be careful here, because there is a version of reality where PI eventually surprises everyone. But as a trader, I don't trade the version of reality that might happen. I trade the order flow that is actually in front of me. I see no order flow for PI because there is no liquid market. I see no audit trail because none has been published. I see no governance structure because the project has not been transparent. Silence is the loudest audit. And PI is screaming. Let me get into the core analysis that the original piece missed. The first thing I look at in any asset is the incentive stack. When I built an arbitrage bot for Curve pools in 2020, I learned that the best technical analysis in the world means nothing if the economic incentives are misaligned. The protocol that eventually tried to manipulate yields did so because its incentive scheme rewarded short-term extraction over long-term health. My strategy survived because I mapped the game, not the code. The same logic applies here. XRP's incentive stack is now institutional. Ripple's acquisition of Hidden Road means XRP could eventually sit inside prime brokerage workflows, where institutions get access to one of the few tokens with regulatory clarity. That creates a different kind of demand—not the feed of retail buyers chasing a green candle, but the slow accumulation of treasuries and market makers. Institutional flows move slower, but they are stickier. The risk is that if the headline narratives around a Bitcoin ETF or an XRP ETF fail to materialize, the entire thesis gets postponed, not abandoned. The patience required is uncomfortable. Art burns hot; patience burns colder. ADA's incentive stack is the classic permissionless chain: stakers secure the network, developers deploy applications, users pay network fees. That is a healthy base. But in a bull market, the price of ADA depends on whether Cardano becomes a venue for marginal speculation. The article gave me no data on Cardano's DeFi TVL. From what I know from public dashboards, Cardano's TVL still lags far behind Ethereum, Solana, and even some smaller L1s. Its positioning as an academic chain has earned respect but not overwhelming market share. The whale accumulation signal is real, but it is also based on a handful of large wallets. I have seen whale accumulation fail to hold support when the broader market turns. I need to see sustained TVL growth, active address growth, and a compelling new application before I adjust my view. PI's incentive stack is almost completely suspended. The project rewards users with PI for tapping a button on a phone, which means the incentive is directed at app engagement, not at economic value creation. Without a liquid token, without major exchange listing, and without a functioning on-chain ecosystem, the community is an audience, not a user base. The 100x prediction is not entirely absurd in the sense that any token with zero liquid market can theoretically pump from a near-zero base. But it is not an investment thesis. It is a lottery ticket. And the lottery odds are made worse by the lack of team transparency. If I cannot verify who controls the supply, I assume the supply controls me. The original article's biggest weakness is that it treats three AI answers as if they were research. They are not. These models are trained on the same public internet that contains years of Ripple hype, Cardano devotion, and Pi Network evangelism. Their outputs are statistically likely to reflect the dominant narrative in their training data, adjusted for some caution. When two of the three models leaned toward XRP, that tells me XRP has the most coherent public story—not that the token will outperform by March. AI models are sentiment aggregators, not alpha generators. I can get the same information by reading a month of crypto Twitter, but I would rather read the order book. Let me apply the game-theoretic lens that we use in my copy trading community. We ask one question: who is the marginal buyer? For XRP, the marginal buyer is becoming an institution—or an ETF structure that pulls in institutional money. But there is a catch. Institutions are patient. They do not need a token to go 10x in three weeks. They need it to settle faster, clear regulatory checks, and keep liquidity deep. That means the XRP bull case might actually be a slow grind, not a parabolic spike. Retail traders who want to buy XRP because Ripple is winning might be disappointed by the payback calendar. The upside is real, but it is measured in years, not quarters. Flows change, but the current remains. For ADA, the marginal buyer is a contrarian retail trader who sees a 73% drawdown and thinks the same thing that happened in 2021 will happen again. The whale accumulation adds a second layer: large holders positioning ahead of a possible narrative rotation. But the absence of a fresh catalyst bothers me. Without a protocol upgrade, a major partnership, or a sudden surge of DeFi capital, ADA's rally lacks a foundational event. It is an alligator moving in shallow water: visible motion, but no deep displacement. For PI, the marginal buyer does not exist yet—at least not in a liquid form. There is a large pool of community members who want to sell at $10 when they bought at $0. The 100x is a transaction that happens only if exchanges open the floodgates. Once that happens, the real question is not whether it pumps, but whether the team and early miners dump on the newly-created market. I have personally seen this pattern in the ICO era. A project with millions of Telegram followers lists on a single exchange, the token spikes briefly, then the unlock schedule begins. The narrative wins, the price loses. I cannot prove PI will follow the same path, but I have a strong prior. This brings me to the contrarian angle. The conventional reading of the article is pick the best of the three. My reading is different: the question itself is a trap. The next bull market may not reward any of these tokens in the way their communities expect. If the market is truly at the tail end of a long bear market, the assets with the cleanest institutional access and the deepest liquidity—like Bitcoin and Ethereum—will likely lead the first phase. Then capital will rotate into higher-beta assets only after the main indices stabilize. XRP and ADA are high-beta, but so are a hundred other tokens. The differentiation will come from actual product usage and revenue. Based on the article, none of the three is showing a compelling, verifiable revenue narrative. XRP's real-world payment flows are not disclosed in the article. ADA's DeFi volumes are not disclosed. PI's future revenue is hypothetical. We are being asked to choose among narratives, not fundamentals. Some observers argue that ADA is the safest because its dilution risk is lower. I respect that. But in a crypto bull market, dilution is a secondary concern. When everyone is playing offense, the lowest-dilution token does not always win. The highest-momentum token wins until the music stops. I remember this from the 2017 ICO era: tokens with absurd unlock schedules still outperformed for months because the demand side was overwhelming. Dilution matters at the margin, not at the decision level. The decision level is always about how fast the flow enters and where it exits. Another contrarian observation: the market's preference for AI-driven predictions is itself a signal of low conviction. When traders are confident, they do not ask a chatbot what to buy. They watch order books, track cumulative volume delta, and check funding rates. When they lose confidence, they outsource the decision because they are hoping for certainty. I see that in the article: the author is writing because three AIs agree is a better headline than I did the research. It is not. The market rewards people who do the research. Let me talk about something the article completely missed: regulatory geography. XRP has MiCA. That is a huge advantage in Europe, where institutions cannot simply buy the most volatile unregulated tokens. A European bank that wants to offer crypto custodial services is more likely to choose a token with a licensed company behind it. Ripple's acquisition of Hidden Road also gives it a presence in the prime brokerage world, where institutional clients can borrow, trade, and settle across asset classes. This is not a small detail. It is the difference between a token that retail traders hold and a token that financial infrastructure is built around. PI has none of this. If you ask me which of the three has the best risk-adjusted position for the next five years, I would say XRP, despite the lack of a 100x trajectory. The risk-adjusted answer and the lottery answer are different. But I also need to check the downside. XRP has been above $1 for a total of about a week in its entire existence. Its liquidity is heavily concentrated in retail exchanges, and institutional adoption does not automatically translate into price appreciation. In fact, if Ripple builds all those corridors and continues to sell XRP from its treasury to fund operations, the institutional narrative could coexist with persistent sell pressure. The acquisition of Hidden Road may not even be designed to push XRP up. It may be Ripple building a crypto brokerage business that uses many assets. That is a critical nuance: the company is becoming a financial services conglomerate, not a single-token booster. The token might be only a part of the story. Do not confuse company success with token success. I also want to challenge the ADA bull case. The 17% weekly gain sounds strong, but in the crypto universe, a 17% gain after a 73% drawdown is a speck. It takes a 370% gain to get back to the old high. The whale accumulation could be a pre-planned accumulation before a catalyst, or it could be a large buyer averaging down into a dead cat bounce. Without wallet-level analysis, the signal is too weak. I would only trust it if we see a sustained increase in active addresses and a rise in Cardano's DeFi TVL. That data is public. Peers like DeFiLlama and CardanoScan can confirm it. The article did not cite any such data, and that omission tells me the rally is narrative-driven. What would change my mind? For XRP, a formal XRP ETF filing, confirmed integration with Hidden Road's prime brokerage platform, or a major European bank announcing XRP liquidity flows. For ADA, weekly growth in TVL above a threshold, a surge in stablecoin minting on Cardano, or a catalyst like a large institutional tokenization partnership. For PI, a listing on Binance or Coinbase with functioning on-ramps, a published tokenomics audit, and a credible mainnet security review before the listing. If those occur, I will adjust. Until then, I cannot recommend PI except as a speculative position with a small amount of capital you can afford to lose entirely. Even then, I would be watching the exit before I watched the entrance. The article also fails to mention that these three tokens occupy different liquidity bands. XRP has a market cap in the tens of billions. ADA is comparable. PI has no meaningful market cap because it is not trading on major venues. Liquidity profoundly changes how a token behaves in a bull market. A smaller asset with small float can pump harder, but it can also crash faster. The best performer in the next bull market is likely to be the token with the highest percentage of holders in profit entering the market, the lowest overhead supply, and the strongest new demand catalyst. None of these can be inferred from the article. We are blind. Now, let me bring it back to the trading room. My community was built on a simple rule: trust no one, verify everything. But that rule is too short for deep analysis. The long-form version is: separate the story from the structure, watch the flows, and never let a compelling narrative replace a balance sheet. I have been burned by a beautiful story dressed as a technical architecture—the 2017 audit failure taught me that. I have also profited from boring structures with correct incentives—the 2020 Curve arb taught me that. The difference has never been the brand. It has been the game theory. When I look at XRP, I see a game that is being played in boardrooms, courtrooms, and prime brokerage terminals. When I look at ADA, I see a game being played on trading charts, Telegram groups, and academic forums. When I look at PI, I see a game being played inside a mobile app, but the scoreboard is not connected to the market. Three games. Three different risk profiles. One bull market that will not treat them equally. Let me offer a synthesis. The safest conclusion from the original article is not a token pick. It is a monitor list. If the bull market starts in the next six months, watch three signals. First, any regulatory or ETF event related to XRP. Second, Cardano's TVL and active address trends over the next four to eight weeks. Third, any top-tier exchange listing announcement for PI. These events are far more valuable than the AI predictions that prompted this article. Because in my experience, the pattern becomes visible before the price moves. You just have to know where to look. I see the pattern before the price does. The next bull market will eventually arrive. It will bring enough greed to make everyone believe a 100x is destiny. But three AIs asking which token will perform best is not a prophecy. It is a Rorschach test. The market will likely choose an asset with actual inflows, confirmed product-market fit, and a leadership team accountable to someone other than a chatbot. And if I had to place a cautious bet right now, it would not be on the loudest community. It would be on the asset that can survive a regulatory audit, settle a payment in ten seconds, and keep its supply schedule honest. That asset is XRP. But I would not pay 100x for it. I would pay exactly what the market offers on the days when the noise fades and the order book is quiet. The numbers didn't lie, but my trust did. So now I trust flows, incentives, and the patience to wait for the signal. Art burns hot; patience burns colder. The same is true of assets. The one that burns brightest in the first month of a bull market is rarely the one that survives the whole cycle. The one that survives is the one that has aligned incentives, regulatory clarity, and a use case that does not depend on a single listing. That is the truth the AI models cannot see. Yet.