Buyback Yield
Buyback yield is the annual amount a protocol spends repurchasing its own token, expressed as a share of market cap. It is the shareholder-yield analogue for tokens: it converts a buyback headline in dollars into a rate that ranks across projects of very different sizes, and it only becomes a judgement once you read it against the revenue funding it and the ending the purchased tokens meet.
TradFi parallel: Like shareholder yield, which states buybacks and dividends as a percentage of market cap so a repurchase program at a small company and a large one can be compared on the same axis.
Key Takeaways
- 01Buyback yield is annual buyback spend divided by market cap, which turns a dollar headline into a rate comparable across tokens of any size
- 02It belongs in a valuation set, not on its own: the study pairs it with price-to-sales, price-to-fees, FDV/revenue and the revenue trend
- 03The same yield can mean three different things depending on whether the tokens are burned, held in treasury, or redistributed to stakers
- 04Read it against revenue and the revenue trend: PUMP's 36% yield sat on falling revenue, flagged as a value trap rather than a cheap asset
- 05Market cap is a float denominator, so a low-float token reports a flattering yield; HYPE was 15.5x sales on market cap and 70x on FDV
- 06Check the spend actually reached the open market: ASTER's headline was inflated by treasury tranches, and the on-market spend behind it is an order of magnitude smaller, so the real yield is a fraction of the headline
How It Works
Buyback yield is annual buyback spend divided by market cap. Tokenomist's buyback and burn study uses it as one of five valuation angles, alongside price-to-sales, price-to-fees, FDV/revenue and the revenue trend, precisely because the raw dollar figure ranks nothing: the largest program in the field is not the most material one. On trailing twelve month figures as of 2026-07-29, HYPE spent the most in absolute terms but registered about a 6% yield against a $12.2B market cap, while RAY at a $163M market cap posted about 15% and PUMP about 36%. Same metric, three very different claims on the token.
The rate alone does not tell you what happened to the tokens. Three endings sit behind the same number: burned, held in treasury, or redistributed to stakers. In the same table PUMP and HYPE burned, JUP's 9% was held, RAY's 15% was held at a public on-chain address, PENDLE's 7% was redistributed, and AAVE's 2% was held. A redistribute program returns tokens to circulation immediately, so its yield measures a transfer to stakers rather than a reduction in supply. Tag the ending before treating any yield as deflation.
Sustainability is the next test, and it is where a high yield can invert. Buyback yield divides by market cap; the money comes out of revenue. PUMP looked like the cheapest asset in the set at 2.4x sales until the 36% buyback yield was set against falling revenue, which the study calls a classic value trap rather than a bargain. An August 2025 review of pump.fun noted it was deploying more than 100% of revenue into buybacks, a higher rate relative to market cap and FDV than HYPE at the time. A yield that exceeds what the business earns is a drawdown with an end date, not a run rate. Verification matters too: ASTER's cumulative buyback figure was inflated by early treasury tranches, and the on-market spend behind it is an order of magnitude smaller: about $4M on Tokenomist's own accounting, or about $7.35M on DefiLlama's trailing twelve month series as of 2026-08-14. Either way the real yield is a fraction of the headline.
Two denominator cautions close the loop. Market cap counts only circulating supply, so a low-float token flatters its own yield: HYPE traded at 15.5x sales on market cap but 70x on FDV because roughly a fifth of its supply was circulating, and that locked supply is exactly what the buyback will eventually have to absorb. And yield is a flow, not a stock. Tokenomist's cumulative buyback total for pump.fun passed about $244M on 2026-03-25 and reached about $320M by 2026-07-29, with total supply down to 840,636,014,425 of a 1 trillion maximum, so roughly 16% has been retired. That stock is the accumulated position, a different figure from the annual rate. Read the flow against scheduled emission over the same window, and the stock against supply.
Real World Examples
PUMP: the highest yield in the set, and the least sustainable
View →On trailing twelve month figures as of 2026-07-29, PUMP carried a 36% buyback yield with the tokens burned, against a $778M market cap and $325M of revenue. It was also the cheapest token in the set at 2.4x sales. The study reads the two together and calls it a value trap: revenue was falling behind a buyback pace the business could not fund.
HYPE: the biggest program in dollars, a mid-range yield
View →HYPE ran the largest buyback in the field but at a $12.2B market cap that spend was roughly a 6% yield, with the tokens burned. Its multiples show the float caveat directly: 15.5x sales on market cap against 70x on FDV, because only about a fifth of supply was circulating. The dollars are the largest; the claim per token is not.
RAY: a high yield on a small cap, parked rather than retired
View →RAY posted about a 15% buyback yield on a $163M market cap, but the tokens are not burned: Raydium's documentation says bought-back RAY is held by the protocol at a public on-chain address, so the supply is parked and can return. It was not the cheapest on multiples either, which is the point: a high yield, a low price-to-sales ratio and an actual reduction in supply are three separate readings.
JUP and PENDLE: similar rate, different supply effect
View →JUP's roughly 9% yield was held rather than burned, and PENDLE's roughly 7% was redistributed rather than retired. Two programs in the same rate band, neither of which reduces supply the way a burn does: held tokens stay outstanding and redeployable, redistributed tokens re-enter circulation immediately.
ASTER: a headline that does not survive verification
View →ASTER's cumulative buyback was inflated by early treasury tranches, and the on-market spend behind it is far smaller: about $4M on Tokenomist's own accounting, or about $7.35M on DefiLlama's trailing twelve month series as of 2026-08-14. The buyback leg redistributes to veASTER stakers while a matched burn runs against the team allocation, and circulating supply keeps inflating. The lesson is to source the spend before computing the ratio.
Frequently Asked Questions
What counts as a material buyback yield?
There is no fixed threshold in the data. Across the programs Tokenomist studied with measurable on-market spend, yields ran from about 2% (AAVE) to about 36% (PUMP), and the highest reading in that range was the least sustainable one. Materiality is relative: compare the yield to the token's scheduled emission over the same window, and to the share of revenue it consumes. A yield that is small against emission is a demand injection, not a change in supply trajectory.
Is a higher buyback yield better?
Not on its own. A high yield can mean a protocol is returning a large share of real earnings, or that it is spending faster than it earns, or simply that its market cap is small. PUMP's 36% was the highest in the set and rested on falling revenue. Pair the yield with the revenue trend and with the share of revenue being spent before treating it as a positive.
How is buyback yield different from supply offset?
Buyback yield is a flow measured in dollars against market cap: what the protocol spent over a year. Supply offset is a stock measured in tokens against supply: how much of circulating supply the program has removed to date. pump.fun's cumulative buybacks passed about $244M on 2026-03-25 and reached about $320M by 2026-07-29, retiring roughly 16% of its 1 trillion maximum supply. That stock and its annual yield are two different questions, one about accumulated position and one about current pace.
Does a buyback yield tell me supply is shrinking?
No. It tells you how much was spent relative to market value. Whether supply shrinks depends on the ending (only a burn permanently retires tokens) and on what the vesting schedule releases over the same period. A program can post a headline yield while its token still runs positive net inflation, because nothing was retired and the unlock schedule kept running underneath.
Related Terms
Track on Tokenomist
Supply-side analysis for educational purposes. Not financial advice. Verify assumption and precision labels on the relevant token page.
Tokenomist.ai provides a complete solution for supply-side tokenomics data. Analyze future token emissions, track vesting schedules, and compare standardized tokenomics and allocation across projects to gain actionable insights