Anticipated psychological spreads: Cryptocurrencies’ hidden short-term monitors and implications for price forecasting

Abstract

Behavioral finance applications to cryptocurrency markets often neglect investor psychology surrounding support and resistance levels. This study introduces the anticipated psychological spread model (APSM), which formalizes chartist investors’ reactions to psychological price thresholds through loss aversion. Two behavioral indicators are defined: buyers’ anticipated psychological spread (BAPS), representing the perceived profit margin near resistance levels, and sellers’ anticipated psychological spread (SAPS), representing the anticipated profit margin near support levels. To examine the short-term price impact of these indicators, the study applies panel quantile regression to 32 cryptocurrencies from January 1, 2020, to January 31, 2024. An autoregressive integrated moving average with exogenous variables (ARIMAX)-based generalized autoregressive conditional heteroskedasticity (GARCH) framework is further employed to test robustness and evaluate the forecasting accuracy of the APSM. The results show that BAPS exerts a negative influence on prices, particularly during bear markets, while SAPS has a positive effect, especially in bull markets. Behavioral asymmetry analysis reveals buyer dominance over sellers throughout the study period. The APSM substantially improves short-term forecasting accuracy compared with classical ARIMAX–GARCH models. These findings indicate that BAPS and SAPS are valuable components for algorithmic trading strategies based on autoregressive models.