# What Kalshi Perps Funding Has Actually Cost Since Launch

*By Benny Ricciardi, FSWA Award Winner · Published Author · Ran 4Deep Sports · Led FTN Marketing · Traded Bonds on Wall Street — The 7 Oracles at PredictionMarketsPicks*

We pulled every Kalshi perps funding print since June 3. Bitcoin longs paid about 7.9% a year, most windows paid nothing, and on 10 of 22 perps the shorts paid the longs.

- Source: https://predictionmarketspicks.com/articles/kalshi-perps-funding-rates-since-launch
- Published: 2026-09-25
- Updated: 2026-09-25

---

Everyone talks about Kalshi perps volume. Nobody talks about the bill.

Funding is the rent you pay (or collect) for holding a perpetual future, and on Kalshi every single print since launch is public. So we pulled all of them: **4,640 funding prints across 22 perps, from Bitcoin's first window on June 3 through the gold print at 10 a.m. ET on Sept. 25, 2026.** Here's what holding a Kalshi perp has actually cost.

**Key takeaway.** Three things jump out. **Most windows pay nothing at all.** **Bitcoin is the exception: longs have paid about 7.9% a year.** And **on 10 of the 22 perps the shorts have paid the longs on net**, so if you like being long the smaller coins, the funding has been working for you, not against you.

## How Kalshi funding works, in one paragraph

A perp never expires, so something has to keep its price glued to the real asset. That something is funding. Every window, Kalshi measures how far the perp traded from its index. **Perp above the index: longs pay shorts. Perp below: shorts pay longs.** Crypto settles up three times a day (12 a.m., 8 a.m. and 4 p.m. ET), metals once a day at 10 a.m. ET. Each window is capped at 2% of the position, and **anything smaller than 0.01% rounds to zero**. That last rule turns out to matter more than everything else.

## Most of the time, nobody pays anybody

Across the 18 perps with at least 30 prints, the typical one paid nothing in **86% of its windows**. Chainlink and Dogecoin were zero in 97% of them. Even Bitcoin, the busiest perp on the exchange, rounded to zero in **54%** of its 342 prints.

That's the 0.01% dead zone at work. On an offshore exchange a tiny premium still charges you a tiny amount every time. On Kalshi it's free until the premium gets big enough to count.

## Bitcoin longs are the ones paying

Bitcoin is the only big perp where one side pays almost all the time:

| Perp | Prints | Longs paid | Nobody paid | Shorts paid | Per year, from the long's side | $10,000 long, whole period |
|---|---|---|---|---|---|---|
| Bitcoin | 342 | 45.6% | 53.8% | 0.6% | **7.9% paid** | −$247.07 |
| XRP | 342 | 27.5% | 66.4% | 6.1% | 3.0% paid | −$94.64 |
| Hyperliquid | 327 | 11.6% | 79.2% | 9.2% | 1.5% paid | −$43.61 |
| Solana | 342 | 7.0% | 90.9% | 2.0% | 0.5% paid | −$16.31 |
| Ethereum | 342 | 2.9% | 86.5% | 10.5% | **1.3% collected** | +$42.15 |
| Zcash | 281 | 5.0% | 74.4% | 20.6% | 3.6% collected | +$92.28 |
| Bitcoin Cash | 324 | 3.4% | 84.9% | 11.7% | 2.9% collected | +$86.58 |
| SHIB (1K) | 324 | 0.0% | 86.7% | 13.3% | 2.6% collected | +$75.73 |
| Litecoin | 324 | 0.0% | 84.9% | 15.1% | 2.2% collected | +$64.95 |
| Sui | 324 | 0.3% | 86.1% | 13.6% | 2.1% collected | +$62.42 |

*Read from Kalshi's public funding-rate history, June 3 to Sept. 25, 2026. "Per year" is the average print times the number of windows in a year. Live numbers for every perp are on the [Kalshi perps board](https://predictionmarketspicks.com/kalshi-perps).*

A 7.9% yearly rate sounds small until you remember funding is charged on the **full position, not your margin**. At 5x leverage, 7.9% of the position is about **40% of your margin per year**, or roughly 3% a month, before a single price move. That's the number to have in your head before you park a leveraged Bitcoin long and forget about it.

## Shorts have been paying the longs on the small coins

Flip the table over and it gets interesting. Of the 18 perps with a real history, **10 have paid longs on net**: Ethereum, Zcash, Bitcoin Cash, SHIB, Litecoin, Sui, NEAR, Worldcoin, Cardano and Dogecoin. Litecoin, SHIB, Cardano, NEAR and Dogecoin **never charged a long once**. Every print was zero or paid the long.

Translation: on those markets the perp has tended to trade a hair *below* its index, which means more traders wanted to be short than long, and the longs got paid to take the other side.

## The one time someone hit the cap

Kalshi caps funding at 2% a window, and it has happened exactly once. **VVV listed on Aug. 28, and its very first print at 4 p.m. ET came in at the full 2%.** A $10,000 long paid $200 in eight hours. Every VVV print since has been a fraction of that, but it's a good reminder that a brand-new listing can price wild for a day, and the funding follows.

## What this means if you trade them

- **Before you hold a Bitcoin long for weeks, price the funding in.** The [perps calculator](https://predictionmarketspicks.com/tools/kalshi-perps-calculator) uses each asset's 30-day average and shows funding next to fees and your liquidation price.
- **Funding and fees are separate bills.** Fees hit twice (open and close) on the full position, starting at 0.120% each way for a taker at the entry tier. Your prediction-market volume counts toward a cheaper tier.
- **Metals are too new to call.** Gold and silver have 15 daily prints each since launching Sept. 10. Gold longs have paid on a third of them. Check back in a month.
- **Leverage changes everything.** The same funding rate that's a rounding error at 1x is real money at 5x. And at 5x, your liquidation price is closer than you think. Run your own position through the [Kalshi perps calculator](https://predictionmarketspicks.com/tools/kalshi-perps-calculator).

[Open a Kalshi account →](https://kalshi.com/t/3k35kcs1 "sponsored")

*Perps carry leverage and can lose your entire margin quickly. Trade responsibly.*

---

## Disclosure

PredictionMarketsPicks publishes analysis of CFTC-regulated event contracts. Nothing here is financial advice and every position carries risk.

Links to Kalshi, DraftKings, FanDuel, Fanatics, Polymarket in this article are referral links marked "sponsored" — we may be paid if you open an account. It costs you nothing and never changes what the model says.