By Keel Research Team
Updated October 15, 2026Walk-forward testing
Walk-forward testing for crypto strategies
Walk-forward testing checks a strategy on data it has not seen: you choose settings on one stretch of history, test them on the next, then roll both windows forward and repeat. Keel is a strategy builder and backtester for Hyperliquid perpetuals, and built-in walk-forward testing is coming soon. Today you can run each window as its own backtest over the dates you choose and compare them.
Coming soon
Walk-forward testing
Built-in walk-forward testing is coming to Keel. It is not in the product yet.
In Keel today
- Backtest over the dates you choose, with fees, slippage and funding
- Change a setting and re-run; every version and backtest is kept
- Paper trade on live Hyperliquid prices with no money at risk
What it is
Choose on one window, test on the next
A normal backtest uses all of your history at once. If you tuned the strategy on that same history, the result flatters it. Walk-forward testing splits the history into steps. In each step you pick settings on an in-sample window and then measure them on the out-of-sample window that follows.
Then you move forward. The next in-sample window starts later, you choose again, and you test on the next unseen stretch. Joining the out-of-sample pieces gives one equity curve made only of results the strategy had not seen when its settings were chosen.
If that joined curve looks like the full backtest, the strategy probably held up. If the edge disappears outside the windows it was chosen on, you find that out before you trade it.
Why it matters in crypto
Crypto regimes change fast
Crypto moves between strong trends, long chop and sharp crashes, often more than once a year. Settings that fit one regime can fail in the next, and a single backtest over the whole period hides when that happened.
Walk-forward testing makes the change visible. You see each out-of-sample step on its own, so a strategy that only worked in one bull run stands out.
Funding matters here as well. Funding on Hyperliquid swings with positioning, so a carry or momentum idea can earn in one window and pay in the next. Each window needs real costs and funding to be a fair test.
A real example
One backtest, several regimes
A walk-forward test splits a run like this into steps. Here is a library strategy over its full window.
Sharpe
1.17
Total return
+151.0%
Worst drawdown
−38.8%
Trades
1,243
Look for the stretches where the curve goes flat or falls: a walk-forward test shows those windows on their own. Momentum Strategy on Hyperliquid, Hyperliquid, Aug 2024 to Sep 2026, after trading fees, slippage and funding. Past results don't predict future ones.
See Momentum Strategy on Hyperliquid in the libraryIn Keel today
Run the steps by hand today
Each walk-forward step is a backtest over chosen dates, and Keel runs those now.
Pick the dates for each backtest
Set a start and an end date on any backtest. Run the in-sample window, choose a setting, then run the next window with that setting.
Keep every window
Every version and every backtest is kept, so the windows stay side by side instead of overwriting each other.
Real costs in every run
Every backtest charges trading fees and slippage and applies Hyperliquid's hourly funding to open positions, so a setting that trades a lot pays for it.
How the backtest worksThen paper trade it
Paper trading runs the strategy on live prices with no money at risk. It is the one out-of-sample window nobody can tune for.
Paper trading on live pricesComing soon
Walk-forward testing
Built-in walk-forward testing is coming to Keel. It is not in the product yet. Sign up free and we'll email you when it ships. Your account works today: build a strategy, backtest it and paper trade it while you wait.
FAQ
Common questions
What is walk-forward testing?
A way to test a strategy on data it has not seen. You choose settings on one window of history, test them on the next window, then roll forward and repeat. The out-of-sample results joined together are the honest estimate.
Is walk-forward testing the same as walk-forward optimization?
They describe the same loop. Walk-forward optimization puts the weight on picking the best settings in each in-sample window. Walk-forward testing puts it on the out-of-sample result, which is the part that tells you whether the strategy works.
Does Keel have walk-forward testing?
Not yet. Built-in walk-forward testing is coming soon, and if you sign up we'll email you when it ships. Today you can run each window as its own backtest over the dates you choose and compare them.
How long should the windows be?
Long enough that each out-of-sample window holds a fair number of trades. With a few years of crypto data, several months in-sample and one or two months out-of-sample is a common start, but the right size depends on how often the strategy trades.
Is walk-forward testing enough to trust a strategy?
It is a strong check, not a guarantee. A strategy can pass and still fail live if the market changes again. Paper trading on live prices is a good next step before real money.
How realistic are Keel's backtests?
Every backtest charges trading fees and slippage on each trade and applies Hyperliquid's real hourly funding to every open position. It is still a simulation: fills are modelled, and past results don't predict future ones. How Keel works explains how the backtest and the live run use the same compiled strategy.
Start with a backtest
Sign up free and we'll email you when it ships. Your account works today: build a strategy, backtest it and paper trade it while you wait.