kel - How to Make it Without Leverage
Summary
Thread Guy’s Zcash thesis rests on three converging forces: rising fear of surveillance and capital controls, growing doubts that Bitcoin will add the privacy and resilience features its original mission requires, and the possibility that fiat and sovereign debt lose credibility together. If those risks become real—or are merely perceived as rising—capital seeking an asset outside governments and custodians could become “very significant.”
The opportunity existed because ZEC spent roughly a decade as crypto’s laughing stock, making an early buyer feel foolish even when the facts were improving. Kel looks for precisely that gap between shared information and biased interpretation: at $30, ZEC was an “antimemetic” idea that resisted circulation; after the rally, everyone could suddenly call it “the next Bitcoin.”
Kel is overwhelmingly a spot investor, with ZEC now accounting for most of his crypto exposure and leverage largely excluded from the process. He ring-fences most of a high-conviction position as a long-term bet—“I won’t touch it; I won’t think about it”—while reserving a much smaller sleeve for sentiment and positioning. The aim is to survive a 90% fall without forfeiting a potential 10x or larger right tail.
Thread Guy’s core warning is that correctly trading every intermediate move can make less money than holding an asset whose endpoint the market still underestimates. HYPE could travel from 2 to 30, back to 8, then 40, 20 and 90; most early bulls would still sell repeatedly. “Every successful investment will have periods of relative underperformance,” so part of the book must remain deliberately unmanaged.
Kel is less convinced that 2026’s onchain boom represents a new technological cycle than a better-packaged continuation of the post-2021 casino. Social trading looks durable, but repeating the same memes may keep producing lower highs unless platforms bridge viral onchain behavior with the large positions retail traders already take in equities, commodities and names such as Sandisk. Hyperliquid’s stickiness also weakened the thesis that zero-fee competitors such as Lighter would automatically seize share.
Decentralized AI offers enormous optionality, but Kel’s base case is blunt: “nothing works.” Pluralis, Pendle, Bittensor/TAO and Venice are attempting technically difficult combinations of distributed compute, model training and reflexive tokens; any winner could resemble Hyperliquid emerging from the crowded 2023 perps field, but protocols may fail technically, abandon tokens or leave crypto entirely.
Kel rates his execution only “B-” but places his idea generation near the top tier because he reads broadly and searches for ideas before they become socially legible. Solana in 2023, Hyperliquid before launch and Zcash before its revival fit that pattern. His closing recommendation was not another trading manual but the antimemetics book he attributed to Nadya Aspanakhova: investors need systems for finding what is valuable precisely because it is hard to remember, share or search for.
Deep dive
1. Kel hunts for misunderstood information, not undiscovered information
Kel entered crypto after becoming disillusioned with traditional political and economic systems during COVID, then joined Messari as a research analyst in 2021. TITAN, the Polygon fork of Luna, supplied the initiation ritual: he bought for maybe $1, sold around $50 thinking, “I’m a genius. I am a value investor,” and watched it go to zero overnight.
His cycle map separates prices from technological conviction. In 2021, “cultish optimism” attached itself to figures such as Vitalik and Su Zhu; by 2023–24, bullish trades revolved around ETFs, Trump, Treasury purchases and memes, financially successful theses that were technologically pessimistic. In 2026, sentiment is worse even as real use cases may finally be appearing.
Kel’s governing question is, “Where is the biggest misunderstanding in the market?” Everyone can now point bots at the same data feeds, so the edge lies in identifying the bias that distorts interpretation. Zcash qualified because buying a decade-old punchline felt idiotic; Thread Guy cited a tweet he thought was from MoonOverlord comparing its revival with Cardano suddenly rising 1,000-fold.
2. Conviction exists to protect the right tail from its owner
Position creation, in Kel’s framework, cannot depend on which insider claims to be buying. “Some guy in Dubai” promising to acquire a huge block is useful only if his word is trustworthy; Kel would rather build an independent thesis strong enough to withstand volatility than borrow conviction from a network he does not possess.
Most of a conviction position is classified as a long-term bet, not an actively managed trade: “If it grows 10 times, great. If it falls by 90%, that’s also great.” A smaller sleeve can respond when traders are maximally long or “the trumpets sound,” but the core must remain outside the daily psychological process.
Kel avoids leverage. After seeing celebrated leveraged traders and large funds use it, he does not feel a need to do so himself. His Bitcoin example was that even someone who lost 99,900 of 100,000 coins but retained 100 could still have preserved a consequential position.
Thread Guy’s pushback—worth keeping—was that holding through repeated moves from 3 to 20, back to 8, then 40, 10 and 60 is brutally difficult when other tokens are multiplying. He cited Connor’s short post, “The Belief That One Should Stick to a Strategy”: anyone unwilling to endure relative underperformance should accept that they will never hold a great investment.
3. Thread Guy’s Zcash thesis combines political fear with dissatisfaction over Bitcoin
Thread Guy resists fixed price targets because they create anchors that overpower new information. Saying ZEC can reach an extravagant number may help express the right tail, but it can also recreate the last cycle’s mistake: remaining leveraged long because Bitcoin once approached an anticipated target, even after the underlying setup had changed.
The first Zcash tailwind, in Thread Guy’s framing, is fear—of surveillance, capital controls, tax enforcement or custodians simply deleting access. He keeps the claim conditional: government risk “may increase or be perceived as increasing.” His concrete precedent is the United States banning private gold ownership in the 1930s despite its reputation for exceptional freedom.
The second is Bitcoin’s perceived stagnation. Thread Guy questioned Bitcoiners’ ability or willingness to implement privacy and quantum-resistant features and solve future fee problems, while noting that discussion has shifted dramatically from 2017’s ethos toward Michael Saylor’s financialized imagery. His categorical version was that Bitcoin “is indeed weakening” relative to the role it was meant to occupy.
The final link is sovereign debt. If bondholders no longer want Treasuries, but also reject storing gold in a foreign vault, the market returns to Bitcoin’s original capital-flight proposition—with privacy added. Thread Guy does not say this transition is guaranteed; he says that if fiat is perceived as “essentially a scam,” the resulting inflow could be enormous.
4. Onchain speculation needs a new growth engine
Kel cautions that it is too early to declare the 2026 onchain cycle sustainable. Faster pattern recognition can still produce exceptional returns in thin AMM liquidity, and ai16z exceeded Kel’s expectations, but the broader euphoria still resembles a continuation of the euphoric roller coaster from 2021 through 2024.
Better products such as FOMO and Phantom have lowered friction, yet Kel has not seen a leap in asset quality comparable to a new technological regime. Even the early ai16z run—its market cap seemingly moving from roughly $1 million to $700 million in two days—has not been repeated, leaving open whether current launches are early or simply making lower highs.
Meme trading can remain “eternally popular” while players enjoy the casino, but becoming the next MGM requires new participants. Kel would reverse his skepticism if college freshmen embraced meme trading; based on the people he encounters, however, recycling the same format through 2026–28 does not yet look culturally inevitable.
His more plausible future combines onchain social behavior with markets where retail already takes huge positions—Sandisk, yields, oil and other large positions. Scrolling $15 million Hyperliquid P&Ls is more compelling than another tiny memecoin scoreboard; the growth platform may therefore look more like FOMO or a competitor than a purely insular blockchain casino.
5. Platform stickiness matters more than headline fees
Kel’s Trade XYZ idea looked attractive because it appeared cheap relative to HYPE, but he admits, “I didn’t trade very well.” He prioritized ZEC because its stigma created a clearer psychological wall among the wealthiest potential buyers; the broader market felt more dependent on technical analysis and sentiment.
His framework comes from “The Gorilla Game”: determine whether platform economics become winner-take-all and which competitive levers actually move users. Early predictions said Hyperliquid would be vulnerable to cheaper rivals, yet priority fees attracted users willing to pay more, while Lighter’s zero-fee offer captured less market share than expected.
Thread Guy supplied the behavioral explanation: he likes Hyperliquid, trusts where his funds sit and will not move between venues to save a little. Kel agreed that this comfort is “very sticky.” Trade XYZ’s strange management posts and the failure of zero fees to prove decisive kept him from doubling down as its price weakened.
6. Decentralized AI is a high-upside thesis with a failure base case
Kel compares decentralized AI in 2026 with perps in 2023: GMX, GNS and dYdX were visible, while the eventual winner, Hyperliquid, was a pre-token product with little traction. He first encountered it through a friend.tech index, illustrating why today’s apparent AI leaders may not be the projects that ultimately matter.
The shared thesis is to redirect unused compute toward valuable inference or model training. The discussion describes the holy grail as performing AI work and mining Pendle simultaneously without extra cost. Token designs for Pluralis, Pendle and Bittensor remain unresolved; Venice appears to be purchasing its own compute, leaving strategies far from settled.
Kel’s honest base case is that “nothing works.” Pre-token protocols may succeed and never issue a token, as crypto-oriented projects can later move away from the sector; existing tokens face the harder task of combining two complex industries with reflexive monetary incentives. The optimistic payoff remains large enough to justify research, not certainty.
7. Reading is Kel’s edge, and antimemetics explains the trade
Concentrated ownership leaves mental bandwidth unused, so Kel fills it with adjacent sectors, science fiction, philosophy and opposing political voices. Broad reading develops the abstract pattern recognition he considers his specialty; it also keeps him occupied without inventing reasons to interfere with ZEC every day.
He grades his trade execution at “B-” but considers his idea generation close to S-tier. His public examples are Solana when the industry treated it as a zero in 2023, Hyperliquid when many expected another fading dYdX or an Arbitrum L3, and Zcash before its rehabilitation.
Kel does not expect to trade forever, though he distrusts the idea that parking everything in the S&P 500 is automatically safe. Experience finding major expectation gaps justifies actively managing some capital, but “trading forever” is not, to him, an inherently valuable life.
His reading list spans Reminiscences of a Stock Operator, More Money Than God, Market Wizards, The Devil Take the Hindmost, Thinking, Fast and Slow, Daemon and The Price of Time, which he described as a continuation of The High Price of Time. His final emphasis was the antimemetics book he attributed to Nadya Aspanakhova: learn to find ideas that resist transmission. “You want to learn about memory stocks when Sandisk is $30, not when it’s $1,000.”