Selection Bias
Selection bias occurs when the data collected for analysis is not representative of the entire population, leading to skewed results and potentially flawed trading strategies. In cryptocurrency markets, this often manifests when analysts only consider data from top-tier centralized exchanges, ignoring the fragmented liquidity found on decentralized protocols.
Because the sample is not randomized, the conclusions drawn ⎊ such as volatility estimates or correlation coefficients ⎊ may not hold true when applied to the broader market. This bias is particularly dangerous in quantitative finance, where models rely on the assumption that historical data accurately reflects future market behavior.
If the sample excludes critical segments, the model fails to account for the actual dynamics of price discovery across the ecosystem. Recognizing this bias is essential for building robust risk management frameworks.