Kool Krypto: The Story Behind The Anon DeFi Investor and What He's Betting On This Cycle

Summary

• Kool Krypto’s core message is that crypto’s “fundamentally mispriced tokens” may be entering massive cap-expansion reratings, and the costly error is selling early — “learn to believe a little bit again.” Institutional adoption is reaching a tipping point after products such as on-chain options took far longer to build and mature than expected; options teams may have thought “this year is going to be our year” for five years. He still expects to round-trip some positions.
• His highest-conviction bag is Derive (DRV), which he calls “probably the purest expression of regulatory arbitrage”: desks such as Wintermute and Flowdesk cannot face retail options flow directly, so Derive’s RFQ system lets them quote retail orders while giving users access to an instrument they previously lacked. He found it around $50 million FDV, says roughly 99% of his Derive orders now use RFQs, and sees AI agents as an unpriced tailwind — echoing founder Nick’s “this isn’t billions, this is trillions” framing. He says Derive has about 95% market share, while also noting that almost nobody trades HYPE on Derive and there is no HYPE price discovery there.
• The episode references a roughly $3.5 million BTC win on a September 70K/80K call spread bought when Bitcoin ranged around 62K–65K and implied volatility sat near its 1st–5th percentile — about $300K of premium for roughly a 20:1 maximum payoff. He also put on 10,000 units of a March 27 ETH 5,000/7,000 call spread at about $30 per unit, risking roughly $330K for up to $20 million. The ETH logic: an all-time-high retest “is not that interesting”; the asymmetry is an ATH break plus FOMO rotation — and if ETH merely hits $4,000, the $5,000 calls could still appreciate roughly 10x.
• Near-term, he is risk-off through month-end and bullish on October, with structures targeting HYPE moving from $85 to $100. He expects a possible year-end breather from tax selling, funds locking in gains, midterms and rates. He says there are no hikes possible in October, and considers an October blowoff top that traps the four-year-cycle crowd a particularly thematic outcome — while stressing that these are path-dependent views that should change as price changes.
• Kinetiq is his asymmetric HyperEVM bet. Consensus treats HyperEVM as “a total dud,” but he doubts Jeff would accept a defunct HyperEVM under the “house of finance.” He says Elysium could move the bullish Kinetiq case from roughly $1 toward $5; at 10% of Robinhood Chain’s sequencer fees, he estimates a rerating above 10x to about $2.60–$2.70. He and DeFi Dad disclose Kinetiq exposure, and he says it is difficult to be bullish on HYPE without considering Kinetiq.
• He publicly reversed on Lighter: after claiming an airdrop of roughly 275–280 units, he sold it all around launch, then rebuilt a position near a $1.40 average basis and rode it toward $5. Fee compression was an overemphasized thesis — moving from 1 basis point to zero saves little — but Lighter’s architecture may suit an options order book better than Hyperliquid’s. He expects Lighter options by year-end or early next year and sees a possible Hyperliquid/Lighter/Derive division of labor.
• The Hyperliquid story explains his sizing philosophy: rudimentary market making on the perp spot market cost about $2,000 in fees and generated 20,000–30,000 points, now worth “tens of millions.” At TGE, he overrode his partner’s proposed sale into a $2.5–$5 billion FDV and broke the fund’s 10% position cap as HYPE rose through $4, $8 and $20, arguing for a minimum $20 billion valuation. His guiding principle is “use everything, own what you use.”
• On TradFi, he sees strong understanding of stablecoins but continuing blind spots elsewhere. His fund seed traces to a poker-table thesis that the US would promote stablecoins to preserve dollar dominance and create demand for short-dated Treasuries. He says Robinhood’s chain economics were visible on DeFiLlama for two or three weeks before an article preceded a 10% HOOD move, while some metals professionals with multibillion-dollar positions did not know the funding rates on copper, gold and silver.

Deep dive

1. The bookend thesis: people forgot how to believe

• Kool Krypto opens and closes on the same note: people have become so battle-scarred by the recent stretch that they forgot what they had been building toward for five, six or seven years is coming to fruition. Options are his exhibit A — teams such as Derive’s may have thought “this year is going to be our year” for five years, because products take longer to build and mature than expected.
• The trading corollary: “everybody’s so quick to sell” a 10% pump, but on fundamentally mispriced tokens “the really big mistake is selling too early.” He accepts the cost — “I’m sure I’ll round-trip a bunch” — because these are “massive, massive, massive cap-expansion” bets that he considers relatively easy to underwrite.

2. Since 2010: mining Bitcoin was easy, keeping it was the trick

• He is 29 and has been in crypto since 2010 — more than half his life in the industry. A PC-building teenager competing on Tom’s Hardware forums and a self-described anarcho-capitalist libertarian making spreadsheets to fix Social Security, he mined Bitcoin from 2010–2014 and stayed “for the love of the game.”
• His survivorship-bias warning: making money then was easier than keeping it — Peercoin, Dogecoin, Mt. Gox and other pitfalls meant that even with today’s information, many people would still have lost everything in one of the era’s schemes, coins or hacks.
• The DeFi aha came with early Ethereum DeFi, against his experience getting a small-business loan: “I can load up a million bucks of ETH into a wallet and take a $500,000 loan at 2% interest… this is going to change the world.” Self-taught, with no college or financial background, he concluded that if he could learn these primitives, they could eventually be accessible to anyone.

3. The poker game: stablecoins as dollar strategy, arbitrage as Trojan horse

• In 2021 he met high-net-worth individuals and TradFi hedge-fund managers at a poker game; they connected him with a prominent New York family office that seeded the fund he has run since.
• The thesis that hooked a contact with a Fed background: the United States would promote stablecoin proliferation to preserve dollar dominance and create a sink for short-dated Treasuries while selling out of duration. He points to Bessent, Trump, War and others as examples of pro-stablecoin positioning.
• The pitch that reduced career risk for allocators was to concede that crypto looked silly, then emphasize the arbitrage opportunities created by its inefficiencies: “we’re actually going in to take advantage of other people’s stupidity.” That made the trade easier for TradFi professionals to accept.

4. TradFi gets stablecoins, misses everything on the ground floor

• Stablecoins were the second-most-understandable thesis after Bitcoin as a store of value or gold alternative, which is why the Circle IPO did so well. He says TradFi views on stablecoins can be superior to crypto-native views because of their regulatory and distribution expertise, and calls it an exceptionally attractive business.
• The HOOD example: Robinhood Chain’s TVL and sequencer-fee revenue were visible on DeFiLlama for two or three weeks before the stock re-rated. After a Friday article brought the information to wider attention, HOOD rose more than 10% on Monday. He qualifies that institutions were days, perhaps a week or two, late rather than months late.
• The residual edge is on-the-ground usage: he cites professionals with multibillion-dollar precious-metals positions who did not know the funding rates on copper, gold and silver earlier in the year. TradFi understands the emerging P/E and P/S framing, but many participants still do not use crypto-native infrastructure for signal.

5. From GLP carry to value accrual: the perps arc

• The fund’s carry mandate found GLP early — “you literally got to beat the house” — with roughly half the fund in GLP, alongside carry and basis trades that anticipated what Ethena later popularized.
• His repeated investor-letter refrain was that value-accrual mechanisms and tokenomics did not matter yet but eventually would. Gains performed acceptably and GMX performed well because he saw the total addressable market for perps as enormous, while the associated tokens could internalize trading fees, liquidations and other value through mechanisms such as buybacks, burns and distributions.

6. The Hyperliquid make-it trade

• The fund had to mark its points internally for investor accounting. He marked them on the high side — recalling values of $2, $4 and $8 per point — while privately believing they could be worth more.
• The engine was rudimentary market making on Hyperliquid’s heavily incentivized perp spot market: roughly $2,000 in fees produced 20,000–30,000 points, which he says are now worth tens of millions.
• The 4 a.m. TGE disagreement was well reasoned on both sides. His partner expected a $2.5–$5 billion FDV and a typical airdrop selloff; Kool argued for at least $20 billion. As HYPE rose through $4, $8 and $20, he kept adding, overrode the fund’s previous 10% position cap and told the team to keep holding. Internally, he viewed it as a $20–$50 billion opportunity.

7. Fund construction: exposure bands and “own what you use”

• The mandate began at roughly 50% exposure, with the remainder supported by yield strategies. Because crypto routinely experiences drawdowns of 80% or more, he views maintaining dry powder for extreme lows as preferable to being fully exposed at all times.
• The design fits crypto’s rhythm: euphoric periods often bring attractive delta-neutral carry, including funding above 15% annually and triple-digit rate dislocations on dYdX lasting for weeks. The moments when the fund wanted to reduce risk often provided the best yield opportunities.
• They do some liquid-VC, points-farming and liquidity-farming deals, but their edge is in liquid markets. His guiding principle is “use everything, own what you use”: he traded heavily on GMX, Hyperliquid and Derive, so he wanted to own each.

8. Memecoins: a casino he refuses to dress up

• He says he has no edge in memecoins and doubts people understand how much of the apparent success is survivorship bias: the winners are visible, while the billion-plus people who lost money are not. His broader objection is that memecoin attention distracts from an industry that has cleaned itself up.
• “Let’s just be honest, this is a casino.” He has nothing against people who trade them, and views the protocols that monetize the activity as potentially strong investments, but rejects treating the revenue as automatically durable. Pump.fun’s low valuation reflected uncertainty about sustainability, unlike his greater confidence that people will keep trading on Hyperliquid over the next several years.
• His market read is that memecoins absorb excess crypto-native liquidity in the same way Bitcoin can absorb excess liquidity in the broader economy. When hot money is sloshing around and people start blasting it into memes, he often reads that as a top-ish signal.

9. Derive: the bull case is regulatory arbitrage, not just better UX

• Conviction came from operational pain. Deribit was among the last venues his fund could access, requiring a prime broker and adding fees and complexity. OTC options meant messaging Wintermute, Flowdesk or Galaxy, waiting hours — sometimes receiving no response on weekends — and then explaining every leg to fund administrators and auditors. Derive makes the process visible and settled on-chain.
• Why it sat around $50 million FDV: repeated failures among on-chain options projects such as Ribbon left investors jaded, while views such as Flood’s led people to treat perps as having obsoleted options. Kool calls that an oversimplification; he sees perps and options as complementary instruments.
• The load-bearing structural point is that options desks generally will not face retail directly because of regulation, infrastructure and account-size requirements. They can instead onboard to Derive as RFQ partners and quote retail orders. He calls Derive “probably the purest expression of regulatory arbitrage”: retail gets access to an instrument it previously lacked, while desks gain distribution without having to build enormous retail-facing operations.
• The flywheel is spinning: his trading size attracted makers, competition tightened quotes, and better pricing generated more volume. He says liquidity has reached a crescendo over the last two or three months. He says Derive “feels like” it owns the HYPE market, but also notes that almost nobody trades HYPE on Derive and that there is no HYPE price discovery there. He adds that some OTC desks are not permitted to trade ZEC because it is a privacy coin.

10. Anatomy of the roughly $3.5M Bitcoin call spread

• The episode references a roughly $3.5 million BTC win. The setup was Bitcoin ranging around 62K–65K, with implied volatility near its 1st–5th percentile because volatility pricing is heavily influenced by recent price action. Having climbed the wall of worry around the Saylor overhang, he wanted upside exposure without a downside leg: buy September 70K calls and sell 80K calls to finance them.
• Spreads let him forgo the extreme tail above 80K and take leveraged exposure to the 70K–80K move. He says buyers are overpaying if they retain exposure to a highly remote, seven-standard-deviation outcome. Market makers also quote spreads more cheaply because they do not need to collateralize naked short calls with tens of millions of dollars on a decentralized platform.
• The trade cost about $300,000 and had roughly $5 million of maximum upside, around a 20:1 payoff. The equivalent 500-Bitcoin position would have required enormous capital at 2x–3x leverage to avoid liquidation. Bitcoin then rallied to around 80K and above; he trimmed and booked the trade’s P&L, while retaining some short 80K calls into a rangy, slightly bearish month-end view.
• His correction to a common misunderstanding is that the expiry P&L graph is not the whole trade: options reprice dynamically, and he generally does not hold them to expiry. The position could have been sold or hedged after Bitcoin reached 80K even if it later fell back to 65K.

11. Who’s on the other side: desk dispersion and DIY makers

• OTC market makers fill most of his trades, but anyone can quote RFQs. He sometimes bids when he sees an order, and two users can potentially meet in the middle and eliminate a market maker.
• Desk dispersion is a core reason for RFQs. He says Flowdesk complains about weak business development while Wintermute is particularly effective at sourcing family-office flow. In one example, Wintermute paid 75 cents for January $6 LIT calls he was selling, versus 55 cents from Flow Traders. RFQ systems replace copying requests across multiple chats and waiting through different time zones with a competitive quote process.
• Beginners can be discouraged by the mark: an option marked at $2 might be bid at $1.60 and offered at $2.40, producing an immediate negative mark-to-market. But a retail trader with a different volatility view can also become a maker and quote the order.

12. AI agents are the unpriced tailwind — and the tooling is accessible

• DeFi Dad connects Variant Fund’s AI-startup exposure to the AI-agent thesis he discusses; Kool separately says AI tools have helped him learn which option structures fit different market environments. A mutual friend is described by the Derive team as the only person they know to have built an AI agent that can open RFQs and close open trades on Derive V2, though Kool leaves open the possibility that others exist.
• The larger claim is that DeFi yield is drying up: the era of putting money into SushiSwap or Anchor for 20% APR is gone, so yield strategies may need to express a view on volatility. He expects user-generated vaults and AI agents to route more carry through options, which is the context for Nick’s “this isn’t billions, this is trillions” framing.
• The stack includes Derivatives Monkey, which he says should not be free and compares favorably with institutional data products costing $50,000–$60,000 per year, plus implied-volatility and historical-realized-volatility checks through Derive, DeFiLlama, Claude or ChatGPT. If volatility is around the 10th percentile, he would lean more toward long-volatility trades, while noting that short volatility tends to pay more frequently and long volatility is less frequent but more asymmetric.
• His practical advice is to ask an AI agent for a structure matching a view, such as a ratio spread that buys one 78K call and sells two 80K calls for a capped-rally thesis. But “there’s no magic in options”: the trade still requires a fundamental view.

13. Perps and options: hammer and screwdriver, best used together

• He compares them to a hammer and screwdriver rather than competing products. India’s zero-DTE options dominance illustrates how traders stick with the form factor they know. Perps can replace zero-DTE options for extreme leverage and a pure delta view, but they are poor tools for expressing ranges or views about the amplitude of a move.
• His workflow is to enter an options position once and hedge its delta with perps instead of paying a chunky options spread both entering and exiting. Against a 70K call, he might ladder perp sells at 72K and 74K as delta increases, preferably on extreme wicks. Perps provided instantly available liquidity during the cited 10/10 example, while an RFQ might not fill during a move lasting only seconds.
• He explains gamma with an Alabama football analogy: when a team is ahead 81–3, another field goal barely changes the prediction-market line. Deep in- or out-of-the-money options behave similarly, while gamma peaks near the strike, where price changes alter delta most sharply.

14. Outlook: bullish October, then uncertainty

• His regulatory hot take is that failure of the CLARITY Act could be a slightly negative headline but a positive catalyst. He sees the bill as better for consumer protection but potentially too punitive toward builders and protocols, while bespoke SEC and CFTC rulemaking might be more accommodating to crypto builders.
• He holds very bullish October option structures, including a bet on HYPE moving from $85 to $100, tied partly to USDC from AQA V2 actually beginning to flow toward Hyperliquid. He also notes that no hikes are possible in October. After that, he sees possible pressure from tax selling, funds locking in gains, midterms, rates and the “50th peace deal with Iran.”
• He frames the market through the question of what would be most painful for participants. The four-year-cycle crowd waiting for October could be trapped into buying an October blowoff top, an outcome he says fits crypto history. But these are path-dependent views: if price moves 10%–20%, the thesis should change, as he argues it should for any token view, including Arthur Hayes’s.

15. The ETH lottery ticket: March 5,000/7,000 call spread

• On the morning of recording, he put on 10,000 units of a March 27 ETH 5,000/7,000 call spread. The $5,000 calls cost about $40 and the sold $7,000 calls brought in about $10, for a $30 net cost. He wanted 20,000 units and might add more. The position risks roughly $330,000 for up to $20 million, although he stresses that the odds of reaching the full payoff are low.
• An ATH retest is not sufficiently asymmetric: if ETH merely returns to its high while other assets are well above theirs, that is not much of an ETH resurgence. He wants an ATH break followed by FOMO rotation. If ETH reaches $4,000 before expiry, he estimates the $5,000 call could trade near $300 — roughly a 10x on that component without reaching the strike.
• He says the Solana thesis has weakened somewhat because Robinhood Chain may take over some of the activity he associated with Solana. Robinhood Chain is also an ETH bear case in the near term: ETH earns roughly “$6 and a pair of shoelaces” against Robinhood’s reported $6 million per day in sequencer fees. Still, staking yield and activity-linked gas make ETH a relatively digestible institutional thesis compared with Bitcoin’s lack of productive cash flow.

16. Kinetiq: the free-options basket on HyperEVM

• His frame is that altcoins resemble options positions: over a long enough horizon, they may go to zero or return 10x, 20x or 50x. Kinetiq is one of the few viable HyperEVM beta plays. Kool and DeFi Dad say they hold Kinetiq, while DeFi Dad separately discloses a small DRV position.
• Consensus treats HyperEVM as “a total dud,” but Kool doubts Jeff would leave a defunct HyperEVM under the “house of finance.” His analogy is a car driven through the founder’s living room and simply left there.
• Elysium changes the target: its connection to HYPE is intended to avoid the cannibalization problem seen in prior L2 models, moving his bullish Kinetiq framing from roughly $1 toward $5. At 10% of Robinhood Chain’s sequencer fees — itself a bullish target — he estimates a rerating above 10x to approximately $2.60–$2.70 at the current revenue-to-market-cap ratio. On staking revenue alone it may be fairly valued or rich, but the other potential sources of value are effectively free options.
• The opportunity exists because the thesis is complicated and not fully front-runnable. Investors may wait to see Elysium generate fees, then buy once the case is obvious. He says it is difficult to be bullish on HYPE without considering Kinetiq. He also says the Markets app is currently generating substantially more Kinetiq points for him than simply holding kHYPE and using other methods.

17. “What if Hyperliquid does options?” — the moat answer

• The structural answer is that one perp is one order book, while a BTC options market requires many strikes and expiries, each with its own book. He says Hyperliquid’s current infrastructure is not well suited to that task. Deribit remained the leading options venue for years despite Binance, OKX and Bybit also offering options.
• The moat is more than technology: Derive has spent months or potentially more than a year building market-maker integrations, distribution and user education, and Kool says it has about 95% market share. Listing an asset is not enough: he says Lighter had “stretch” and SpaceX listed before Hyperliquid, but those markets attracted little volume. He reads rival emphasis on binary options and RWAs as a willingness to pursue adjacent markets rather than immediately compete head-on.
• The unpriced reverse case is Derive building a larger perp business. Kool says it has long been a cheap place to borrow against HYPE and has paid yield on deposits longer than Hyperliquid. If it becomes a broader derivatives venue, including for major assets and RWAs, it could warrant the kind of valuation currently associated with perp DEXs. He compares waiting for a future Hyperliquid options launch to asking what would happen if Nasdaq listed perps: it can be addressed when it happens.

18. Lighter: a reversal told against himself

• He claimed an airdrop of roughly 275–280 units and immediately sold it, believing the launch timing was poor. The platform also had instability issues: during TGE and other periods, some orders did not go through or were canceled. He also thought traders were overemphasizing fee compression: moving from 1% to 10 basis points matters, but moving from 1 basis point to zero saves little relative to the value of better liquidity and execution.
• What changed his view near $1 was Lighter’s possible advantage in building an on-chain options book, its early connection with Robinhood and a valuation that looked cheap relative to Hyperliquid’s revenue. He also cites Lighter bulls’ theses around a possible CFTC license and an Ethereum-based institutional “escape hatch,” rather than presenting those outcomes as established facts. His average basis after adding was around $1.40, and he rode the position toward $5.
• He now sees a possible CEX-like structure: Hyperliquid as the dominant venue, Lighter as another large venue and Derive specializing in options, analogous to Binance, Bybit and Deribit. He says he is more concerned about Lighter as a competitor than Hyperliquid in the options context, and expects these market roles to develop for structural reasons.