Crypto cross sectional momentum

Checked October 2026

Cross-sectional crypto momentum buys the coins that rose the most over three weeks and sells the ones that fell the most. Liu, Tsyvinski and Wu published it in the Journal of Finance in 2022. Backtoast ran it. The equal-weight version failed our test, losing 22.6% a year over 14,770 trades.

Backtoast

You pay $99 for two full tests of your own strategy, and a pass is an offer of 30% of the profit Backtoast makes trading it with its own money.

Crypto cross sectional momentum

Each week the rule ranks coins by their return over the past three weeks, buys the top fifth, shorts the bottom fifth and holds a week, betting the coins already running keep running against the ones already falling. The bitcoin-only weekly momentum rule is a different test.

The paper weights each coin by its market value. The run gave each of 60 longs and 60 shorts an equal share and used every Binance USDT pair with no floor on size, so small coins counted as much as large ones.

Crypto weekly contrarian bets the other way, buying last week's losers.

When it works

Cross-sectional momentum does best in long crypto trends, when the same coins lead for weeks, and loses in sharp reversals, when the leaders fall hardest. Traders keep using it because the paper found momentum is one of three factors, with the market and coin size, that explain why some coins return more than others. It showed up mainly in the larger coins, so a floor on size is the usual fix, and a draft can test that before you spend a full test.

Origins

In their 2019 working paper, Liu, Tsyvinski and Wu reported 4.1% a week for the top fifth minus the bottom fifth from 2014 to 2018. Among the larger half of coins it was 4.2% a week, and among the smaller half 0.6%, which wasn't significant. The run is from late 2021 to September 2026.

The run is Backtoast's, from 29 October 2026, after costs.