Hyperliquid Assistance Fund: HYPE Buyback Tracker (2026)
How the Hyperliquid Assistance Fund buys HYPE with trading fees: address, real balance and daily buybacks from a dated snapshot, and a Python tracking script.
The short version: Hyperliquid sends trading fees to the community, and part of them goes to the Assistance Fund, a system address (
0xfefe…fefe) that automatically converts fees into HYPE as part of L1 execution. Hyperliquid's docs say HYPE in the fund is burned. In our snapshot on 25 October 2026 (about 08:05 UTC), the fund held 48.25 million HYPE, worth roughly $3.67 billion at $76, and over the six full days from 19 to 24 October it bought an average of 21,484 HYPE (about $1.69 million) per day. All of it is readable from the public API with no keys.
"HYPE buyback" is one of the most searched Hyperliquid topics, and also one of the most misquoted. Numbers get passed around on social media without a source or a date. The good news is that you do not need anyone's dashboard: the Assistance Fund is an ordinary address on HyperCore, so its balance and every fill it makes can be read from the same public endpoints you would use for any wallet. This guide explains what the fund is, what the docs actually say, the real numbers from a dated snapshot, how to read them correctly, and a short Python script to pull them yourself. If you would rather just look at the address, paste it into Hyperfolio: it is a free, read-only tracker, no wallet connection or API keys needed.
What the Assistance Fund is
Hyperliquid's fee page states that fees are "entirely directed to the community (HLP, the assistance fund, and deployers)". Spot and HIP-3 deployers can keep up to 50% of the fees generated by their own assets. For the Assistance Fund, the documentation says three specific things:
- It uses the system address
0xfefefefefefefefefefefefefefefefefefefefe. - It converts trading fees to HYPE "in a fully automated manner as part of the L1 execution". There is no team wallet placing orders by hand.
- HYPE in the Assistance Fund is burned, which removes it from circulating and total supply.
That is the whole mechanism. Traders pay fees, a share of those fees ends up at the fund, and the fund buys HYPE on the HYPE/USDC spot market. HLP, the protocol vault, is a separate destination; we cover it in the vaults guide.
Real numbers: snapshot of 25 October 2026
We queried the public info API on 25 October 2026 at about 08:05 UTC. The HYPE spot mid price was $76.00.
| Metric | Value | Source (info request type) |
|---|---|---|
| HYPE held by the fund | 48,251,014 HYPE | spotClearinghouseState |
| Value at $76.00 | ≈ $3.67 billion | allMids, pair @107 |
Cost basis reported by the API (entryNtl) | $1.385 billion, ≈ $28.70 per HYPE | spotClearinghouseState |
| HYPE total supply | 998,828,890 | tokenDetails |
| HYPE circulating supply | 302,002,103 | tokenDetails |
| Fund as % of total supply | 4.83% | calculated |
| Fund relative to circulating supply | 15.98% | calculated |
Two details are worth knowing. First, tokenDetails lists the fund under nonCirculatingUserBalances, so the 48.25M HYPE is already excluded from the circulating figure. That is why "15.98% of circulating" is a ratio, not a slice. Second, the same response still counts the fund's balance inside total supply. If you see a site subtract it to show a "burned" figure, that is its own adjustment, not a raw API field.
Daily buybacks: the last seven days
Every purchase is a fill on the HYPE/USDC spot pair (@107 in the API). Grouping the fund's fills by UTC day gives this:
| UTC day | HYPE bought | USDC spent | Average price | Fills |
|---|---|---|---|---|
| 18 Oct (from 08:36) | 11,111 | $859,707 | $77.38 | 464 |
| 19 Oct | 21,200 | $1,674,966 | $79.01 | 837 |
| 20 Oct | 22,445 | $1,791,104 | $79.80 | 982 |
| 21 Oct | 20,995 | $1,706,229 | $81.27 | 740 |
| 22 Oct | 27,614 | $2,162,084 | $78.30 | 1,071 |
| 23 Oct | 17,980 | $1,396,067 | $77.64 | 803 |
| 24 Oct | 18,667 | $1,410,826 | $75.58 | 583 |
| 25 Oct (to 07:57) | 3,431 | $262,317 | $76.46 | 147 |
Across the six full days (19 to 24 October), the fund bought 128,901 HYPE for $10.14 million, or 21,484 HYPE and $1.69 million per day on average. The whole seven-day window adds up to 143,444 HYPE for $11.26 million at an average of $78.52.
The daily amount is not a fixed schedule. Because the fund is fed by trading fees, the size of each day's buying moves with fee revenue, which in turn moves with trading activity. That is the useful way to read this table: the buyback works as a rough live meter of what Hyperliquid earns in fees. If you want to see how fees relate to volume, our trading volume guide breaks down where Hyperliquid's volume comes from.
How the fund trades
The fills show a consistent pattern in this window:
- Buy only. All 5,627 fills were buys. The fund does not sell.
- Passive orders. Every fill had
crossed: false, meaning the fund's orders were resting in the book and were hit by other traders, rather than crossing the spread. - No fees. Every fill shows a fee of 0.
- Many small orders. The 5,627 fills came from 2,241 distinct order IDs, with no TWAP ID. The buying is spread through the day in small pieces, which keeps its price impact low.
This matters for traders: the fund is a steady, price-insensitive bid on HYPE/USDC, but it is small relative to the market. In the same snapshot, the rolling 24-hour volume was $37.9 million on the HYPE/USDC spot pair and $363 million on the HYPE perp. The fund's $1.41 million on 24 October was about 3.7% of that spot volume and well under 1% of spot and perp combined, so it does not set the price on its own.
How to read the numbers without fooling yourself
- Do not annualize one week blindly. $1.69 million per day would be about $617 million over a year, but only if volume and fees stayed exactly where they were this week. They will not.
- The value in dollars moves with HYPE. The $3.67 billion figure is mark-to-market at $76. The
entryNtlcost basis of about $28.70 per HYPE is what the API reports for the position, so the gap between them is a paper figure, not money anyone can spend: the fund only buys. - Burned does not mean gone from the API. The docs call the fund's HYPE burned, while
tokenDetailsstill includes it in total supply and lists it as non-circulating. Always say which figure you are quoting. - Always date your snapshot. The balance grows every few seconds. A number without a timestamp is not checkable.
Track it yourself with Python
The script below needs only requests. It prints the fund's HYPE balance, its share of supply, and the daily buys for the last seven days. It pages through userFillsByTime, which returns at most 2,000 fills per call; Hyperliquid only serves the 10,000 most recent fills for an address, which at the current pace of 600 to 1,100 fills a day is roughly one and a half weeks. For anything older, store the results yourself, as explained in our historical data guide. It uses three or four light requests, well within the API rate limits.
import time, collections, requests
from datetime import datetime, timezone
API = "https://api.hyperliquid.xyz/info"
AF = "0xfefefefefefefefefefefefefefefefefefefefe" # Assistance Fund system address
HYPE_SPOT = "@107" # HYPE/USDC spot pair
def info(body):
r = requests.post(API, json=body, timeout=10)
r.raise_for_status()
return r.json()
# 1) Current HYPE balance and its value
bal = next(b for b in info({"type": "spotClearinghouseState", "user": AF})["balances"]
if b["coin"] == "HYPE")
hype, cost = float(bal["total"]), float(bal["entryNtl"])
mid = float(info({"type": "allMids"})[HYPE_SPOT])
print(f"HYPE held: {hype:,.0f} value: ${hype * mid:,.0f} avg cost: ${cost / hype:.2f}")
# 2) Share of supply
tok = next(t for t in info({"type": "spotMeta"})["tokens"] if t["name"] == "HYPE")
d = info({"type": "tokenDetails", "tokenId": tok["tokenId"]})
print(f"of total supply: {hype / float(d['totalSupply']):.2%} "
f"vs circulating: {hype / float(d['circulatingSupply']):.2%}")
# 3) Daily buys over the last 7 days (paginate: max 2,000 fills per call)
end = int(time.time() * 1000)
t, fills = end - 7 * 86_400_000, []
while True:
page = info({"type": "userFillsByTime", "user": AF, "startTime": t, "endTime": end})
fills += page
if len(page) < 2000:
break
t = page[-1]["time"] + 1
days = collections.defaultdict(lambda: [0.0, 0.0, 0])
for f in fills:
if f["coin"] == HYPE_SPOT and f["side"] == "B":
k = datetime.fromtimestamp(f["time"] / 1000, tz=timezone.utc).date()
days[k][0] += float(f["sz"])
days[k][1] += float(f["sz"]) * float(f["px"])
days[k][2] += 1
for k in sorted(days):
q, usd, n = days[k]
print(f"{k} {q:>9,.0f} HYPE ${usd:>12,.0f} avg ${usd / q:.2f} fills {n}")
New to the API? Start with our Hyperliquid API Python guide.
Track it without code
Because the fund is a normal address, any read-only wallet tracker can follow it. In Hyperfolio, paste 0xfefefefefefefefefefefefefefefefefefefefe to see its spot balances and recent fills next to your own wallets. The same approach works for whales and vaults: see how to track multiple Hyperliquid wallets. Hyperfolio never asks for a private key or a signature.
FAQ
What is the Hyperliquid Assistance Fund address?
0xfefefefefefefefefefefefefefefefefefefefe, as stated in Hyperliquid's fee documentation.
How much HYPE has the Assistance Fund bought?
On 25 October 2026 at about 08:05 UTC it held 48,251,014 HYPE, about 4.83% of total supply. Check the live figure with the script above.
How much does Hyperliquid buy back per day?
From 19 to 24 October 2026 the average was 21,484 HYPE, about $1.69 million, per day. It changes daily with trading fees.
Is the HYPE in the Assistance Fund burned?
Hyperliquid's docs say yes: HYPE in the fund is burned and removed from circulating and total supply. In the API, tokenDetails lists it as non-circulating.
Does the Assistance Fund ever sell HYPE?
Not in the data we checked: all 5,627 fills in the seven-day window were buys.
Does trading on Hyperliquid feed the buyback?
Yes. Trading fees fund it, so more volume means more buying. Staking HYPE and referral codes lower the fees you personally pay; see the staking guide.
Do I need an API key to check this?
No. All the endpoints used here are public read-only info requests.
Trade for less. If you are new to Hyperliquid, use referral code HYPERFOLIO for 4% off trading fees (how the referral works), then follow your PnL, your fills and wallets like the Assistance Fund for free in Hyperfolio.
Try Hyperfolio for free
Track your Hyperliquid portfolio in real time with PnL, Smart Money, Markets, Perp Calculator, multi-venue portfolio and push alerts.
HYPERLIQUID
-4% feesTrade on Hyperliquid · 4% off fees
Hyperfolio is an independent app, free to use with no subscription. We're funded by Hyperliquid referrals, ad space and donations; we truly appreciate you trading through our link.
Free for you · keeps Hyperfolio running
Related articles
Hyperliquid Liquidation Map: Real Liquidation Levels (2026)
Build a Hyperliquid liquidation map from real on-chain positions instead of estimated heatmaps. Dated example from the 300 largest accounts on 2 Nov 2026, how to read clusters and concentration, and a Python script.
GuideBest Time to Trade on Hyperliquid: Hours & Days (2026)
When Hyperliquid volume and volatility peak, by UTC hour and weekday, measured from 672 hourly candles (4 Oct to 1 Nov 2026) for BTC, ETH, HYPE and SOL, with what it means for slippage, stops and funding, plus a Python script.
GuideHyperliquid Long/Short Ratio: Read Trader Positioning (2026)
Why the market-wide long/short ratio on Hyperliquid is always 1, how to measure it for a cohort of wallets from public data, a dated example of the top 100 leaderboard traders, and a Python script.
GuideHyperliquid Sharpe Ratio: Risk-Adjusted Returns Guide (2026)
Calculate Sharpe and Sortino ratios for any Hyperliquid wallet or vault from the public portfolio endpoint: the deposit and withdrawal trap, dated leaderboard and HLP examples, and a Python script.
GuideHyperliquid Correlation & Beta: Hedging Crypto Perps (2026)
Measure correlation and beta on Hyperliquid from the public candle API: dated 30- and 90-day matrices for BTC, ETH, HYPE, SOL, XRP and DOGE, three portfolio and hedge examples, and a Python script.
GuideHyperliquid Volatility & ATR: Stops, Size, Leverage (2026)
Measure Hyperliquid volatility yourself from the public candle API: realized volatility and ATR for BTC, ETH, HYPE, SOL, XRP and DOGE on a dated snapshot, plus stop, position size and leverage examples and a Python script.