Use case

Improve your backtest

Stop backtesting fantasy fills. Price in slippage, impact and funding with point-in-time metrics.

Shown: minute lows of the 465 Binance perpetuals, 20:40–22:00 UTC on 10 Oct 2025, with BTC’s path traced · built with ohlcv 1m

Featured notebook

Backtesting with realistic microstructure cost

A price-only backtest quietly assumes you can trade at the mid-price for free. Real fills pay the half-spread, move the book, and miss when liquidity thins out. This notebook builds a microstructure transaction-cost model — half-spread plus market impact — from Aperiodic's L1 price, impact, and slippage metrics, checks it against measured slippage, and shows how different cost assumptions reshape the same simple momentum backtest.

Loading notebook…

From notebook to production

Pick the metrics

Browse the catalog and pull point-in-time microstructure metrics — no L2 reconstruction, no tick archive to maintain.

Run the notebook

Every notebook runs end to end on preview data with a free account, in the browser or through the CLI.

Plug it into your stack

Swap the preview key for your own and the same code pulls every year your plan includes through the Python package or the API.

Metrics in this use case

More notebooks for this use case

Also featured

Funding-aware perpetual backtests

Improve your backtest

A perpetual-futures position does not only earn the price move — it also pays or receives funding at discrete settlement prints. A price-only backtest ignores those cash flows and can misattribute carry to trading alpha, most of all for a persistently long strategy in a rising, positive-funding market. This notebook isolates realised funding and shows exactly where it binds: it is a first-order cost for a directional (long-biased) position and nets out for a direction-neutral one.

View notebook