Art of Trading - with Light, Su Zhu, and Hasu

Uncommon Core podcast (~October 2020). Hasu sits down with Light, an anonymous prop trader who moved from equities and options through poker into crypto, and Su Zhu, CEO of Three Arrows Capital. Two halves: a trading process masterclass from Light, and a macro argument for Bitcoin's structural dominance over altcoins from Su Zhu and Light together.

Part One: Trading as Process

Markets as a changing zero-sum game

Light frames crypto as closer to poker than to investing — a zero-sum game of incomplete information — with one critical difference: in poker the rules are fixed and only the meta evolves; in trading both the rules and the meta change. That double change rewards people who can hold frameworks loosely and update them as the game morphs. It is why he argues generalists still outperform specialists in crypto: the space is soft enough that a disciplined generalist can harvest edge across multiple simultaneous games rather than being trapped in one.

He blends three sources of signal: behavioral economics (narratives drive reflexive prices), asymmetric information (edge vs. less-informed participants), and tape reading — sitting with live order flow long enough to build a pattern library of market tells. "If you spend enough time staring at market orders coming in from an aggregate of different exchanges, you'll get a feeling… I've seen that before, and usually this starts and then something else follows."

The three-part trade requirement

Every trade needs: (1) a thesis, (2) both a price-level and thesis-level invalidation, and (3) an asymmetric payoff. All three are required.

Without price invalidation there is no structural stop. Without thesis invalidation there is no test for whether the reason to hold has broken down independent of price. Without an asymmetric payoff the risk-reward profile is structurally unattractive even if the thesis is right. Su Zhu's payoff benchmark: a 3:1 reward/risk at 50% win rate is a great trade. Most beginners either skip invalidation (hoping the trade recovers) or skip the payoff check (entering without quantifying what winning looks like).

This belongs with position-sizing — size should always be derived from the distance to invalidation, not from conviction or excitement.

Ego-defense as the biggest beginner leak

Society trains people to avoid being wrong. Markets punish this hardest. The refusal to be wrong early leads to a cascade: oversizing, averaging down without structure, thesis drift (the trade silently reclassifies as a long-term investment), and eventual ruin.

Light: "Everyone gets what they want from markets. Most guys want to come in and be right." Making money is boring — it is tedious, patient, disciplined. Spectacular trades feel like gambling. Good process feels like sitting at a nine-handed poker table and folding most hands.

Timeframe conflation

A distinct failure mode: starting a position as a trade and then reclassifying it as a long-term investment once it moves against you. Psychologically seamless, structurally lethal. In auto-correlated markets — where down moves follow down moves — adding size after a loss without a structural reason (not just a lower price) compounds the original error. "There's nothing more dangerous than having that happen to you because the human mind is pretty amazing in what it can rationalize."

Nullification as the right approach

The correct scientific process: try to find all the reasons the trade is wrong first. If you cannot nullify the hypothesis, then trade with more confidence. Most traders do the reverse — they collect reasons to justify holding what they already own. Light: "Whenever I have a trade idea that feels good, the way I actually spend time with it is by trying to figure out all the reasons why this is a terrible idea. And then if I can't come up with something conclusive, I feel more confident."

This connects to inversion and second-order-thinking: the goal is to put yourself in the shoes of the counterparty and ask why they are right and you are wrong. Without that exercise you have no real basis for the trade.

Trade journal as the feedback mechanism

Cognitive biases from volatility — loss aversion, anchoring to high-water marks, prospect theory — make memory unreliable as a learning tool. Without a recorded log of trades (entry thesis, invalidation level, size rationale, exit) a discretionary trader cannot do leak-finding. Light's comparison: trading without a journal is like poker without hand histories. You keep making the same mistakes because you never see the pattern.

The leak-finding priority matters too. Find the structural leaks first — the spots where you are hemorrhaging EV in obvious ways — before worrying about fine-grained execution improvements. A trader obsessing over whether to call or fold on a specific run-out while systematically opening weak hands from bad position is fixing the wrong problem.

Options mispricing in crypto

A structural supply/demand imbalance: yield-seeking traders sell covered calls (capping upside for consistent income), and structured product providers sell derivatives to miners and whales who prefer steady cash flows over convexity. Both systematically underprice upside volatility.

Light and Su Zhu positioned themselves as consistent call buyers — "we show up, people message us because they know we're the marks" — and treated it as recurring structural edge. The heuristic: when Bloomberg runs an article about crypto survivors finding a "rare lifeline" in selling Bitcoin calls at the bottom, that is the signal that upside is cheapest. The specific numbers decay as markets mature, but the mechanism (behavioral yield-seeking creates mispriced asymmetry) is a durable lesson about where asymmetric payoffs can hide.

Part Two: Bitcoin Macro Thesis

Bitcoin as crypto's shelling point

Su Zhu and Light argue that Bitcoin is the shelling point of crypto monetary coordination: investors with different reasons — inflation hedge, digital scarcity, censorship resistance, macro optionality, Lindy durability — converge on the same asset. Altcoins fragment this coordination. For any alt-positive thesis (DeFi, smart contracts, rollups) there are many competing tokens, which dilutes the coordination game. Light: "If you were to go into a coma tomorrow for 10 years and had one chance to set your crypto portfolio, would you own Filecoin? Would you own Compound?" The answer was implicit.

Trust minimization and fair launch

Trust minimization is why monetary coordination is easier around Bitcoin than around more technically capable alternatives. Systems where the rules are legible, constrained, and hard to change are easier to coordinate around because participants do not have to renegotiate who controls what. Ethereum's constant roadmap evolution signals to institutional players that the rules are not settled. Bitcoin's "nothing happening" is the best signal for that audience: the 50-year-old family office manager who puts Bitcoin in his portfolio needs to know what he's going to get.

The ETH bear case

Su Zhu's argument: Ethereum was pricing in two things it would not hold. On money: only one base money is needed; Bitcoin already owns that coordination point. On smart contract platforms: rollups and competing L1s (Polkadot, Cosmos, Solana) would fragment the application market. Ethereum as one of many smart contract platforms is not worth what Ethereum as the smart contract platform is worth. And the narrative problem: "I don't even know what the narrative is anymore that can be sold to people, and that narrative is key for price appreciation." In reflexive assets, a fragmented narrative is a bearish catalyst because narrative clarity is a large component of value.

Historical Outcome

The podcast was recorded around October–November 2020, with BTC trading near $13k.

BTC all-time high call: correct. Bitcoin cleared its 2017 ATH in late 2020 and hit ~$69k in November 2021. The shelling-point and institutional-adoption thesis played out faster than options markets implied. Light's call to buy cheap upside was a textbook asymmetric trade.

ETH bear call: wrong in the short run. ETH went from ~$400 to ~$4,800 in 2021, massively outperforming Bitcoin on raw returns during the DeFi and NFT cycles. Su Zhu's structural argument about ETH's narrative problem was valid as a long-term concern, but the near-term narrative (DeFi summer, NFTs, Ethereum as the application layer) was powerful enough to drive a historic run first.

Alt season timing: roughly right. Su Zhu said "a year or two away." The 2021 cycle delivered one of the most extreme alt seasons on record — SOL, LUNA, AVAX.

3AC's collapse in 2022: the hardest outcome. Three Arrows Capital, Su Zhu's fund, was one of the largest casualties of the LUNA/UST collapse in May–June 2022. 3AC had taken on massive leveraged exposure and reportedly continued holding through the collapse — a brutal real-world demonstration of the exact failure modes Light described in Part One: no structured invalidation, thesis drift, refusal to be wrong early, and eventually ruin. The fund that helped articulate the principles of process discipline destroyed itself by violating those principles at scale under real pressure.

The evergreen framework in Part One aged cleanly. The specific market calls in Part Two aged with mixed accuracy. The historical coda is its own lesson: knowing the principles is not the same as applying them.

Connections

Sources

Transcript provided by user