Crypto Backtesting: Data Quality, Market Fragmentation, and Continuous Trading
Summary
The document outlines a basic backtesting workflow: define a strategy, gather historical data, calculate returns and supporting statistics, then decide whether to deploy or refine the idea. It notes that crypto strategies may use pairs trades, rebalancing rules, or directional signals, much like equity strategies.
Its central lesson is that crypto data requires extra care. Trading spans centralized and decentralized venues with uneven liquidity and inconsistent records; markets run continuously, so day boundaries and time zones affect analysis. Short histories, uncertain data for smaller assets, whale-driven price moves, regulatory shocks, and blockchain forks further complicate tests. These factors can distort prices and make results sensitive to data choices.
The article recommends broad exchange and protocol data as a way to support research across crypto markets, but this section is promotional and supplies no comparative evidence that a particular data source improves performance. It gives no specific method for handling these issues, so researchers must define data cleaning, venue selection, trading costs, and event treatment themselves. Backtest results remain limited by the quality and history of the underlying data.
Key ideas
- A useful backtest starts by specifying the strategy and the historical data needed to evaluate it.
- Crypto data can vary across venues, especially when liquidity is thin, making price selection consequential.
- Continuous trading means that day definitions and time zones can change measured results.
- Short histories, whale activity, regulatory events, and blockchain forks can undermine the reliability of historical tests.
- Backtest conclusions should account for data quality and the limits of the available sample.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.