# Crypto backtesting: test rules on Hyperliquid perps

> Backtest crypto rules on Hyperliquid perps like BTC and ETH, daily or hourly, with funding and costs counted. Free. Backtest and paper or testnet only.

Written by [Tactfolio research team](https://tactfolio.com/about/editorial-policy.md), published 2026-10-16, updated 2026-10-16, data as of 2026-10-16.

Crypto backtesting shows how a set of trading rules would have done on past
crypto prices. Tactfolio backtests rules on Hyperliquid perps, such as BTC and
ETH, on daily or hourly prices. Funding and a cost on every trade are counted.
It is free, and crypto on Tactfolio is backtest and paper or testnet only.

- Test crypto perps such as BTC, ETH, SOL, and HYPE, and perps that track
  stocks, indexes, and gold, such as XYZ:TSLA and XYZ:GOLD.
- Pick daily or hourly bars. Rules can run every hour, every 4 hours, or once a
  day, week, or month.
- Prices are adjusted for funding, so every result is net of it. Every trade
  pays a 0.05% cost.
- A crypto strategy holds only perps, never stocks or ETFs as well.
- It holds each perp long, at 1x. It never borrows or sells short.
- A backtest is hypothetical. It can't tell you what happens next.

## What is crypto backtesting?

Backtesting means testing a plan on past prices, as if you had followed it at
the time. The SEC's investor site warns that back-tested results
[are hypothetical and do not reflect actual performance](https://www.investor.gov/introduction-investing/general-resources/news-alerts/alerts-bulletins/investor-bulletins-47).

Crypto adds three things to watch:

- **The market never closes.** Crypto trades every hour of every day, weekends
  included. A test needs prices for all of those hours.
- **History is short.** Bitcoin has traded for years, but most coins and most
  perps have not. A short test can miss a bad year.
- **Perps charge funding.** Holding a perp long usually costs a small fee every
  hour. A test that leaves it out looks better than real life.

## What you can backtest on Tactfolio

Tactfolio tests perps listed on [Hyperliquid](https://hyperliquid.gitbook.io/hyperliquid-docs),
an exchange for perpetual futures. A **perp** is a contract that tracks a
price, such as the price of bitcoin, with no end date.

| What you can test | Examples | Notes |
| --- | --- | --- |
| Crypto perps | BTC, ETH, SOL, HYPE | Hyperliquid's main market |
| Stock perps | XYZ:TSLA, XYZ:NVDA, XYZ:AAPL | Listed on trade.xyz, a market on Hyperliquid |
| Index perps | XYZ:XYZ100, XYZ:SP500 | XYZ100 tracks the Nasdaq 100 |
| Commodity perps | XYZ:GOLD, XYZ:SILVER | Can mix with crypto in one strategy |
| Bars | Daily or hourly | Daily bars close at 00:00 UTC |
| Costs | 0.05% on every trade | Close to Hyperliquid's base taker fee of [0.045%](https://hyperliquid.gitbook.io/hyperliquid-docs/trading/fees) |

A strategy picks its market when you create it, and the market never changes.
A crypto strategy can mix crypto with stock and gold perps, such as BTC and
XYZ:GOLD. It can't hold SPY or any other US-listed fund.

## How funding is counted

Hyperliquid [pays funding every hour](https://hyperliquid.gitbook.io/hyperliquid-docs/trading/funding).
When the rate is positive, people who hold a perp long pay people who hold it
short. Tactfolio adjusts past prices for funding, the way stock prices are
adjusted for dividends. So a backtest's growth is already net of the funding a
long position paid.

Here is how much longs paid over the 20 days to October 16, 2026, as a yearly
rate:

| Perp | Funding paid by longs, per year |
| --- | --- |
| BTC | 7.7% |
| ETH | 8.3% |
| XYZ:TSLA | 8.2% |
| XYZ:XYZ100 | 5.4% |
| XYZ:GOLD | 6.6% |

Rates change all the time, and they can turn negative. The point is the size: a
test that skipped funding would look about that much better each year.

## How far back the test goes

| Prices | Where they start |
| --- | --- |
| BTC, daily | August 19, 2020, when it listed on Hyperliquid |
| Stock perps on trade.xyz, daily | Late 2025 for most of them |
| Hourly bars | About 208 days back when a perp is first used, and longer from then on, since Tactfolio keeps every new hour |

A backtest starts once every perp in it has prices. For a perp that tracks a
stock or gold, you can turn on [simulated history](https://tactfolio.com/docs/simulated-history.md) on
daily bars. It fills the years before the perp listed with the prices of the
asset it tracks, such as GLD for the gold perp, and charges a funding cost on
top. Simulated history is off until you turn it on, and the results say which
part is modelled.

For a crypto perp, simulated history fills the years before it listed with
Hyperliquid's own reference prices for the coin, which reach back to 2020 for
BTC and ETH, and charges the perp's average funding on top. With it off, a
crypto backtest covers only the years the perp has traded on Hyperliquid.

## How to backtest a crypto strategy on Tactfolio

1. **Open the builder.** Go to the [builder](https://tactfolio.com/builder) and pick Hyperliquid as
   the market on the start page.
2. **Start from a pattern or a blank strategy.** Two starters are ready:
   **Crypto trend or cash** holds BTC and ETH while bitcoin is above its
   200-day average and steps out to cash when it is not. **Bitcoin and gold**
   balances BTC with the gold perp.
3. **Add your rules.** An **If** block can switch holdings when a condition is
   true. A **Rank** block can hold the strongest few perps. An empty **Weight**
   block means cash.
4. **Pick how often it runs.** Hourly or every-4-hours runs use hourly bars.
   Daily, weekly, or monthly runs use daily bars.
5. **Read the results.** The backtest shows growth, the deepest fall, what it
   held and when, and BTC held alone over the same dates. Save it with a free
   account, and every save is kept.

When a test looks good, run it forward on a
[Hyperliquid paper account](https://tactfolio.com/paper-trading.md) or your own Hyperliquid testnet
account. Neither uses real money.

## Crypto backtesting mistakes to avoid

- **Leaving out funding.** A long perp often pays several percent a year.
  Tactfolio counts it, but a test elsewhere may not.
- **Trusting a short test.** A strategy built on one bull run can fail in the
  next bear market. Check how it did in 2022, when crypto fell hard.
- **Testing too many versions.** Hourly bars give you many more numbers to
  tune. Try enough rules and one will look great by luck. This is called
  **overfitting**.
- **Trading too often.** Hourly rules can trade many times a day. Each trade
  pays the 0.05% cost, and it adds up fast.
- **Forgetting the drops.** Bitcoin has fallen more than 70% from a high more
  than once. Look at the deepest fall before you look at the growth.

[Why past results mislead](https://tactfolio.com/docs/concepts/why-past-results-mislead.md) covers
each trap. [How to read a backtest](https://tactfolio.com/docs/concepts/reading-a-backtest.md)
explains every number in the results.

## What Tactfolio does and does not do

| Tactfolio does | Tactfolio does not |
| --- | --- |
| Backtest Hyperliquid crypto, stock, index, and gold perps | Backtest coins you buy and hold, or other crypto exchanges |
| Test on daily or hourly bars, with funding counted | Test minute by minute |
| Run a tested strategy on paper or your own Hyperliquid testnet account | Trade crypto with real money |
| Hold each perp long at 1x | Use leverage or sell short |
| Keep crypto and stock strategies apart | Mix perps with US-listed stocks or ETFs in one strategy |
| Label simulated history on stock and gold perps | Simulate history for crypto perps |

## Try it on Tactfolio

Open the [builder](https://tactfolio.com/builder), pick Hyperliquid on the start page, and open
**Crypto trend or cash**. Change the average, the perps, or how often it runs,
and the backtest runs again. For stocks and ETFs instead, see
[backtesting on Tactfolio](https://tactfolio.com/backtesting.md).

## Keep learning

- [Trend following](https://tactfolio.com/learn/trend-following.md): the idea behind the crypto trend starter.
- [Moving average](https://tactfolio.com/learn/moving-average.md): how a 200-day average is worked out.
- [Max drawdown](https://tactfolio.com/learn/max-drawdown.md) and [Sharpe ratio](https://tactfolio.com/learn/sharpe-ratio.md): two numbers to read on every backtest.
- [Leverage explained](https://tactfolio.com/learn/leverage-explained.md): why Tactfolio keeps perps at 1x.
- [Backtesting explained](https://tactfolio.com/learn/backtesting-explained.md) and [quant trading on Tactfolio](https://tactfolio.com/quant-trading.md): more on testing rules.
- [Best no-code backtesting tools](https://tactfolio.com/comparisons/best-no-code-backtesting-tools.md): how Tactfolio compares.

## Common questions

### Is crypto backtesting free on Tactfolio?

Yes. Everything is free while we build. You need a free account to run and
save your own backtests.

### Can I trade crypto with real money on Tactfolio?

No. Crypto on Tactfolio is backtest and paper or testnet only. Real-money
trading on Hyperliquid is switched off.

### Which crypto can I backtest?

Perps listed on Hyperliquid, such as BTC, ETH, SOL, and HYPE. You can also test
perps that track stocks, indexes, and gold on trade.xyz, such as XYZ:TSLA,
XYZ:SP500, and XYZ:GOLD.

### Can I backtest bitcoin on hourly bars?

Yes. Pick hourly or every-4-hours runs, and the strategy uses hourly bars.
Hourly history is shorter than daily history, so the test covers fewer years.

### Is a perp the same as owning bitcoin?

No. A perp tracks bitcoin's price, but you never own a coin. Holding it long
usually costs funding every hour. Tactfolio tests perps, not coins.

### Can I mix crypto and stocks in one strategy?

You can mix crypto perps with perps that track stocks or gold, such as BTC and
XYZ:GOLD. You can't mix perps with US-listed stocks or ETFs like SPY.

### Is a good crypto backtest proof that a strategy works?

No. A backtest shows how rules would have done on prices that already
happened. Crypto history is short, so this matters even more. Paper trade a
strategy before you trust it.

## Sources

- [Funding](https://hyperliquid.gitbook.io/hyperliquid-docs/trading/funding), Hyperliquid Docs, accessed 2026-10-16
- [Fees](https://hyperliquid.gitbook.io/hyperliquid-docs/trading/fees), Hyperliquid Docs, accessed 2026-10-16
- [Investor Bulletin: Performance Claims](https://www.investor.gov/introduction-investing/general-resources/news-alerts/alerts-bulletins/investor-bulletins-47), U.S. Securities and Exchange Commission, Investor.gov, accessed 2026-10-16