Reducing Crypto Strategy Turnover with a No-Trade Buffer

Summary

This article demonstrates a practical way to reduce trading costs in a crypto statistical-arbitrage portfolio: keep existing positions until they drift sufficiently far from their target weights. The example uses perpetual futures, excludes stablecoins, and selects the thirty highest-volume assets using trailing volume. It combines carry, momentum, and breakout signals into long and short weights, then evaluates the portfolio with a backtest that includes prices, funding rates, and commissions.

The no-trade buffer controls how much a position must deviate before a rebalance occurs. The author reports that a buffer of 0.04 produced the best historical after-cost Sharpe in the explored settings, and that performance remained relatively stable over a range of values. The illustration also shows lower turnover and fewer daily trades. These findings are historical and depend on the sample, cost assumptions, and model construction; they do not establish future performance. The article emphasizes careful alignment of signals, prices, and funding data, and suggests choosing a somewhat larger buffer to account for possible out-of-sample deterioration.

Key ideas

• A no-trade buffer delays rebalancing until current holdings drift far enough from target weights.
• The example ranks crypto perpetual futures by trailing volume and combines carry, momentum, and breakout signals.
• A realistic backtest needs to account for commissions, funding, and correctly aligned prices and target weights.
• The selected buffer reduced turnover, while its reported Sharpe advantage is limited to the historical simulation.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.