“Buyback tokens averaged +62.2% this year, while the broader market was down 19%”—this set of numbers has been getting repeated by a bunch of accounts over the past couple of days. The source is a long post published by @wallstreetbets on 11/12, titled “The Crypto Revenue Meta Is Just Starting.” I translated the original, and these numbers don’t hold up under closer scrutiny.
First, look at how the sample was put together. The original says it itself: it selected the top 30 protocols by revenue over the “past 30 days.” In other words, only protocols still making money in November made the list. Those whose revenue—and token prices—collapsed this year weren’t included at all. Calculating gains since the start of the year using a year-end list of winners means the conclusion was already baked in when the sample was selected.
Next, look at the control group. Of the 30, 14 don’t have a tradable token, leaving 16: 12 with buybacks and 4 without. That “-31.6%” control group consists of just 4 tokens. Four samples account for a gap of 93.8 percentage points. Make of that what you will.
The mean is 62.2%, while the median is 32.6%—almost half as much. That shows a few massive gainers are pulling up the average. I made a rough calculation using Binance perpetuals’ opening prices on January 1 (UTC) through early this morning: HYPE went from 25.45 to 72.81, about +186%; PUMP was up about +84%; JUP about +40%; UNI only around +4%. Over the same period, BTC went from 87608 to 68974, about -21%. The original says buybacks and burns totaled more than $900 million this year, with HYPE alone accounting for more than $500 million—the bulk of it. The so-called “buyback camp outperforming” is largely just “Hyperliquid outperforming.”
The causal relationship may well run in the opposite direction, too. Buybacks are funded by revenue; high revenue means the business and trading volume are growing, so it’s only natural for the token price to rise along with them. Did buybacks drive up the token price, or did a strong business lead to both buybacks and a higher token price? This table can’t tell the difference. Even funnier, the same article says that in 2025, 28 projects spent more than $1.4 billion on buybacks, and among the top ten, only AAVE, HYPE, and SKY outperformed BTC during their buyback periods. Something that was disproven last year becomes a “meta” this year just by changing the sample.
Now look at where the article is headed. WLFI is excluded from the statistics because it “just launched,” and then gets its own section praising it as the “most direct attempt” to link stablecoin issuance revenue to a token. It’s absent from the data but present in the conclusion. Then, over the next few hours, a dozen or so accounts with “nft” or “web3” in their names repeated the same numbers under the banner of “WSB’s analysis.” I have no proof anyone was paid, so I won’t draw any conclusions, but old hands in crypto have seen this kind of promotion before.
Buybacks themselves aren’t a bad thing. Projects like HYPE, which use actual transaction fees to buy tokens based on volume, really are absorbing sell pressure. But to judge whether buybacks are effective, you need to compare buyback amounts with unlocks and emissions, and see whether revenue can hold up in a bear market—not rely on a retrospective table of cherry-picked returns. If you use this table to chase “the next buyback token,” you’re buying someone else’s narrative.
#回购 #HYPE
First, look at how the sample was put together. The original says it itself: it selected the top 30 protocols by revenue over the “past 30 days.” In other words, only protocols still making money in November made the list. Those whose revenue—and token prices—collapsed this year weren’t included at all. Calculating gains since the start of the year using a year-end list of winners means the conclusion was already baked in when the sample was selected.
Next, look at the control group. Of the 30, 14 don’t have a tradable token, leaving 16: 12 with buybacks and 4 without. That “-31.6%” control group consists of just 4 tokens. Four samples account for a gap of 93.8 percentage points. Make of that what you will.
The mean is 62.2%, while the median is 32.6%—almost half as much. That shows a few massive gainers are pulling up the average. I made a rough calculation using Binance perpetuals’ opening prices on January 1 (UTC) through early this morning: HYPE went from 25.45 to 72.81, about +186%; PUMP was up about +84%; JUP about +40%; UNI only around +4%. Over the same period, BTC went from 87608 to 68974, about -21%. The original says buybacks and burns totaled more than $900 million this year, with HYPE alone accounting for more than $500 million—the bulk of it. The so-called “buyback camp outperforming” is largely just “Hyperliquid outperforming.”
The causal relationship may well run in the opposite direction, too. Buybacks are funded by revenue; high revenue means the business and trading volume are growing, so it’s only natural for the token price to rise along with them. Did buybacks drive up the token price, or did a strong business lead to both buybacks and a higher token price? This table can’t tell the difference. Even funnier, the same article says that in 2025, 28 projects spent more than $1.4 billion on buybacks, and among the top ten, only AAVE, HYPE, and SKY outperformed BTC during their buyback periods. Something that was disproven last year becomes a “meta” this year just by changing the sample.
Now look at where the article is headed. WLFI is excluded from the statistics because it “just launched,” and then gets its own section praising it as the “most direct attempt” to link stablecoin issuance revenue to a token. It’s absent from the data but present in the conclusion. Then, over the next few hours, a dozen or so accounts with “nft” or “web3” in their names repeated the same numbers under the banner of “WSB’s analysis.” I have no proof anyone was paid, so I won’t draw any conclusions, but old hands in crypto have seen this kind of promotion before.
Buybacks themselves aren’t a bad thing. Projects like HYPE, which use actual transaction fees to buy tokens based on volume, really are absorbing sell pressure. But to judge whether buybacks are effective, you need to compare buyback amounts with unlocks and emissions, and see whether revenue can hold up in a bear market—not rely on a retrospective table of cherry-picked returns. If you use this table to chase “the next buyback token,” you’re buying someone else’s narrative.
#回购 #HYPE