The Scoreboard Appeared on September 9
On September 9, the scoreboard went public. David Hoffman, Bankless co-founder, announced on May 21 that he had sold his entire ETH position. Proceeds rotated into five tokens: VVV, NEAR, ZEC, HYPE, and LIT. The reported gains by publication date: LIT +288%, ZEC +121%, HYPE +55%, VVV +55%, NEAR +33%. ETH, across the same window, managed roughly +17%.
The arithmetic mean of those five returns, if equally weighted, lands around +110.4%. The coverage almost writes itself: a Bankless founder abandons ETH, crushes his benchmark. Then you audit the ledger and gaps appear. No position weights. No entry dates beyond one announcement. No stop-losses, no partial sells, no record of whether the five coins were still held when the recap ran. One window of roughly four months. Five assets. A claim of verified skill.
In my work—backtesting yield strategies and auditing where value accrues in DeFi—a result I cannot reproduce from public data is a hypothesis, not a proof. This is a hypothesis wrapped in a victory lap.
Context: When the Preacher Leaves the Church
Hoffman matters because Bankless has functioned as one of Ethereum's loudest belief amplifiers. Its audience skews toward crypto natives who treat ETH as the core holding. When a co-founder sells 100% of his ETH publicly, it stops being a routine portfolio change. It becomes a statement inside the ecosystem's attention economy.
The destinations deepen that statement. Look at what the five tokens represent:
VVV—Venice Token, one of 2025's AI-crypto listings. Decentralized AI inference. Small float, recently launched.
NEAR—a sharded Layer-1, originally generalist, repositioning around AI and chain abstraction. The reported entry near $1.40 sits close to the bottom of its historical range.
ZEC—zcash, the 2016-era privacy coin built on zk-SNARKs. The reported purchase near $560 is historically high. Narrative: privacy revival, halving supply math, possible ETF speculation.
HYPE—Hyperliquid's token, bought around $45. Narrative: high-performance order-book L1 for perpetuals. The on-chain CeFi thesis.
LIT—Lit Protocol, decentralized key management and access control. Smallest cap in the basket. Infrastructure story.
Five tokens, four narrative families: AI, privacy, high-performance execution, and key management. Conspicuously absent: Ethereum-based DeFi and every established blue-chip with a yield loop. This is a rotation out of the ETH universe, not a dip-buying exercise.
The pivot matters beyond one wallet. Bankless spent years telling listeners to accumulate ETH. Now its co-founder sits in AI tokens and a privacy coin. Every future Bankless segment on these topics carries a question the audience will ask: does the coverage follow the money, or does the money follow the coverage?
Core: Reading the Incomplete Ledger
Portfolio return is a weighted average. The public data gives per-token returns but hides the allocation. Change the weights and the story moves dramatically.
Take three baskets from the same five names. Equal weight: roughly +110%. A 70% tilt toward NEAR and VVV—the two laggards at +33% and +55%—collapses the return into the mid-50s. A 70% tilt toward LIT and ZEC pushes it toward +140%. Between those versions, the gap spans a 3x multiple. Hoffman either tripled ETH's outcome or barely left it ahead. Without weights, quoting +110.4% is averaging a portfolio that never existed.
The ledger also omits the rest of the book. Did Hoffman buy other tokens in May that failed? Did he exit ETH in one print or a staggered ladder? The article measures ETH from May 21, but it never shows his actual ETH sell prices. Survivorship bias cuts both ways. A trader who published a winner-only page could hide an entire portfolio of losers. Code doesn't—humans are the weak link—and a 100% winner-only subset is not data.
Unrealized marks are not realized returns. LIT's +288% is a paper gain if he never sold. By the September 9 recap, a later drawdown could erase half of it without any follow-up article. My operating bias comes from surviving the Terra collapse in 2022: between an on-chain anomaly and the eventual failure, forty-eight hours turned thousands of green accounts into zeros. Exit cadence, not entry genius, is what preserved capital there.
Small-cap winners have asymmetric exits. LIT's triple may coexist with thin books. VVV carries a small float and scheduled unlocks that can overwhelm demand. ZEC bought near $560 sits far above its long-term holder basis, and privacy tokens carry exchange-overhang risk in several jurisdictions. HYPE's book is deeper but futures-driven. Entry is easy. Exit is a different market.
A publicized entry matters further. Hoffman disclosed May-level prices; anyone reading in September is buying post-attention prices. Publication itself creates residual buying. My 2024 ETF-arbitrage work taught me this directly: I extracted a 3% dislocation in the GBTC-BTC-ETH complex, but the edge decayed within days once it became visible. KOL scoreboards work the same way, only faster. Retail inherits the tail.
Statistically, this is a toy sample. Five assets. One window. One regime. No drawdown disclosure, no volatility, no Sharpe ratio. If the basket fell 45% between May and September before recovering, its risk-adjusted performance is worse than ETH's smooth 17% grind. During my 2020 Curve liquidity experiment, rebalancing cadence alone shifted annualized results by 14% in volatile conditions. Manager choices—weights, rebalance timing, exits—dominate asset-selection signal. Hidden weights and hidden timing turn every benchmark comparison into false precision.
The genuine, durable information here is regime information: four months in which high-beta alternative narratives outperformed the largest L1. That tells you the market was risk-on toward newer stories, not that a Bankless founder possesses repeatable skill as an asset picker. I would slot the window into broader altseason behavior: capital rotating from “too big to triple” assets toward the sub-$10 billion range, chasing AI-agent catalysts, order-book L1s, and privacy revival themes while Ethereum fee markets stayed depressed.
ZEC deserves a separate comment. It is the only non-2025 narrative asset in the basket. Bitcoin-halving-adjacent supply math and ETF speculation are plausible catalysts, but privacy assets have spent years trapped between regulatory gravity and limited retail distribution. If Hoffman bought ZEC as a hedge against a macro tail—rather than as an innovation trade—his basket is not purely a narrative rotation. That internal inconsistency suggests multi-thesis portfolio construction, and it raises the same question: without notional sizes, we cannot infer which thesis sits at the center of his book.
ETH's absence is as loud as the alts' presence. Ethereum's roadmap has fragmented into L2-specific paths; fee flow that once accrued to the base layer migrated upward. For a capital allocator holding ETH, the demand side now depends on ETF inflows and institutional settlement narratives. This is not a bear thesis—it is a velocity and capital-efficiency problem. In a choppy market where ETH volumes stagnate, smaller assets with independent catalysts can outperform. But that condition is cyclical, not permanent.
Then layer Hoffman's media position over his investments. Bankless is a brand with a DAO, a media arm, and affiliated investment activity. His disclosed ledger does not clarify whether his holdings overlap with the projects Bankless editorial covers. Whether intentional or not, public holdings steer attention, and because attention is crypto's scarcest resource, the disclosure itself becomes a market event. At minimum, followers need to know: the same brand that amplified Ethereum's narrative now personally owns five alternatives to it.
Contrarian: The Trade Was Not the Story
The contrarian angle is not whether he is right about Ethereum. It is that the trade shifts belief infrastructure. Ethereum's premium is partly reflexive; it lives in the conviction of allocators and media. When a flag-bearer of that conviction rotates publicly, the action itself contributes to the very underperformance it profits from. The feedback loop is the trade.
The conflict surface is the real flag. Bankless is a media brand, and Hoffman is its co-founder. If editorial coverage of VVV, HYPE, or LIT expands in the coming quarters, the line between content and personal position blurs. No accusation is needed; the incentive structure is visible on-chain. A scoreboard is not a disclosure policy. The audience deserves to know whether the media arm and the personal wallet move in the same direction.
There is also a self-fulfilling risk. Hoffman's publicized buy may have influenced the very returns he later displayed. If Bankless's content and its founder's allocations move the same crowd, the 110% figure is not a validation of the trade; it is the trade doing what visibility does. It resembles a contest where the musician also controls the playlist. Every marginal buyer who enters after September 9 is paying for that visibility, not for the original edge.
What would make me update? If ETH develops catalysts that offset its fragmented L2 ecosystem—capital-efficiency recovery, real ETF flow maturation, a base-layer fee revival—parts of this alt basket could rotate back violently. Then the contrarian play exits. In a sideways tape, tokens run on narrative rails, and the rails can be pulled out the moment liquidity thins.
Takeaway: Chop Is for Positioning
In a consolidation market, this kind of scoreboard functions as positioning, not proof. Yield is the interest paid for patience and risk—and the relevant patience is in design discipline, in verification, in asking which weights, which exits, and which risk-adjusted returns sit behind the headline.
Before acting on any KOL ledger: trust the audit, verify the stack, ignore the hype, then check where the early liquidity actually sits. The market rewards those who read the source code—and the source code here is incomplete. Ask what the median follower of such published trades truly earns, and then decide where the next attention flow goes. Usually, the best answer is the one your own backtest can reproduce.