Written by: Xiao Bing

On May 21, David Hoffman, co-founder of Bankless and the most vocal evangelist of the Ethereum ecosystem for six years, tweeted early in the morning that "the atmosphere in crypto Twitter has really changed lately, and he sold his last bit of ETH."

Three and a half months later, he became a big winner.

What did Hoffman actually buy?

In early June, Hoffman fully disclosed his portfolio transition on X, deploying the funds after selling ETH in two batches:

The first batch of about 50% of the funds was immediately allocated to four targets after selling ETH: VVV (the governance token of Venice AI), NEAR, ZEC, and HYPE. He clearly mentioned in the tweet that the price of NEAR when bought was around $1.4.

The second batch of about 50% of the funds was reserved for DCA (Dollar Cost Averaging). His exact words were "saving to slowly buy something that hasn’t gone up yet." This money ultimately all went into LIT (Lighter, a token for an on-chain perpetual contract exchange based on zkRollup).

His logic for buying LIT was exceptionally clear: exchanges are always the best business model in the crypto industry; Lighter's repurchase rate is about twice that of HYPE; zk circuits allow users to verify whether the exchange complies with its own rules without permission; the product has lower latency, a better fee structure, and supports more assets, including the pre-IPO market.

When asked "How to choose between LIT and HYPE," he replied that LIT is both the Beta of HYPE and its Alpha, even tweeting at the end of June expressing regret for not buying more LIT.

This is a very clear investment framework: betting on privacy narrative with ZEC, betting on structural growth in on-chain derivatives with HYPE and LIT, betting on cross-chain infrastructure and AI Agent narrative with NEAR, and betting on decentralized AI inference with VVV.

The five targets cover four narrative tracks, none of which are related to the valuation of Ethereum L1.

Report Card

Comparing the prices before and after Hoffman's disclosure of holdings, he sold ETH in late May, and the first batch was built from late May to early June, with LIT's DCA continuing until mid-June. The following uses the approximate prices of each asset in early June as the basis for building positions, comparing with the latest prices in early September:

ETH: Selling price around $2100 → Current around $2450, an increase of about 17%.

ZEC: Entry price around $540 → Current surpassing $1200, an increase of over 120%.

On September 6, it once reached $1200, doubling in three months, with the catalyst being Grayscale ZEC Spot ETF (ZCSH) going live on NYSE Arca on August 25, with AUM growing from $300 million to $460 million in two weeks, combined with a short squeeze liquidation of $46 million.

HYPE: Entry price around $56 → Current around $87, an increase of about 55%.

On September 6, it refreshed its historical high at $89.54. Hyperliquid's token burn has accumulated over $4 billion, with daily protocol revenue maintaining around $2.26 million, and the cash-generating ability of this on-chain ATM is still accelerating.

LIT: Entry price range about $1.5-2 (DCA average price) → Current around $4.7, an increase of about 135%-210%.

On September 5, it refreshed its historical high at $4.95. As the single asset with the heaviest position for Hoffman (accounting for 50% of total funds), LIT contributed the largest absolute return in the entire portfolio.

NEAR: Entry price around $1.4 (confirmed by Hoffman) → Current around $2.37, an increase of about 69%.

VVV: Entry price around $16-18 (in early June, VVV was near its ATH, peaking at $21.32 on June 3) → Current around $17, basically flat.

This is the weakest performer among the five, and also the only holding for which Hoffman did not publicly increase his position or express regret.

Doing a rough estimate of the portfolio return: Assuming 50% of the funds are equally allocated to VVV, NEAR, ZEC, HYPE (each 12.5%), and 50% allocated to LIT, with median estimates calculated, the overall portfolio return rate is about 90%-120%. Meanwhile, ETH increased by about 17%.

Hoffman's portfolio outperformed ETH by at least 70 percentage points.

Where Did He Win?

Looking closely at this report card, the most noteworthy aspect is not how much each individual asset has increased, but rather how Hoffman's choice logic has been repeatedly validated by the market over three and a half months.

The surge in ZEC has traceable reasons. Grayscale's ETF application has been going through the SEC process since last November, and its formal launch on August 25 merely realized the long-developed institutional variable into price. When Hoffman bought in May, ZEC had already risen from over $30 at the beginning of the year to over $500; he chose to enter during the "already increased significantly but the ETF hasn’t landed yet" window, which essentially is betting on the certainty of the catalyst being higher than the short-term risk of the price.

The logic behind LIT is even more worth dissecting. When everyone was talking about how HYPE could become the Chicago Mercantile Exchange on-chain, Hoffman chose an earlier, smaller-cap competitor whose product structure might be more aggressive. His judgment framework was "looking for a higher elastic Beta in the same track," while also layering on structural reasons such as "faster repurchase speed, transparency premium provided by zk validation, and user migration potential from lower fees." In hindsight, LIT has risen about 500% from the bottom to the current level, validating this line of thought.

NEAR, with a return of 69%, ranks fourth in the entire portfolio and does not seem dazzling. However, combined with the current role of NEAR Intents as a "toll booth" in the ZEC market (ZEC-related trading pairs account for nearly 40% of NEAR Intents' total trading volume), Hoffman may have unintentionally built a self-reinforcing portfolio: the more ZEC increases, the larger the trading flow through the Zashi wallet and NEAR Intents, resulting in higher fee income for NEAR Intents, which strengthens the repurchasing of NEAR. There exists a hidden positive feedback loop between the two assets he bought.

VVV is the only position that has not realized gains. The narrative of Venice AI as a decentralized AI inference platform lost sustained momentum after a brief spike in early June, and Hoffman did not publicly increase his position.

The Real Question

Hoffman's report card provides a signal more worthy of exploration than "who increased how much": The center of gravity in the crypto market is shifting from L1 valuation to application layer revenue.

The ETH he sold is an L1 asset, with its valuation logic built on network effects, developer ecosystems, and gas burning mechanisms. Among the five targets he purchased, the valuation anchors for HYPE and LIT are verifiable protocol revenue and repurchases, ZEC's valuation anchor is institutionalized demand (ETF) and observable privacy usage data on-chain (proportion of shielded supply), NEAR's valuation anchor is transaction volume and fees in the cross-chain settlement layer, and VVV's valuation anchor is the actual invocation volume of AI inference services.

The common feature of these five assets is that their valuations come from independently verifiable on-chain activity data rather than future narrative promises.

Hoffman selling ETH is unrelated to the good or bad of Ethereum's technology; he himself has said, "still optimistic that Ethereum will win." He just realized something earlier than most people: In a market with scarce liquidity, faith does not generate returns; only verifiable cash flows and quantifiable demands do.