Pairing Uncertainty and Attention Indices to Explain Cryptocurrency Returns: Quantile and Causality Evidence

Does the pricing of uncertainty in cryptocurrency markets depend on where in the return distribution one looks, and does investor attention carry information beyond uncertainty itself? We address these questions for ten cryptocurrencies spanning dominant and less-dominant coins over September 2018 to December 2021, using quantile regression, joint (multivariate) quantile regression, and Granger causality tests, together with a COVID-19 sub-sample. Grounding the analysis in the uncertainty- and attention-based asset-pricing literature, we test three hypotheses: that uncertainty is priced in the tails of the return distribution rather than at its centre; that cryptocurrency-specific uncertainty matters more than broad macroeconomic uncertainty; and that uncertainty and attention are complementary, so that pairing an uncertainty index with an attention index explains tail returns better than either index alone. The evidence supports all three. Single indices are largely irrelevant at the median but become influential in bear-market tails and over longer horizons; the cryptocurrency-specific UCRY indices dominate the broader macro proxies; and price-or-policy-plus-attention pairs show stronger and broader tail effects than either index alone, though part of this reflects the additional regressor in paired specifications. Causality runs mainly from indices to returns at longer horizons. The study shows that uncertainty is a tail phenomenon in cryptocurrency markets and that attention operates as a distinct, complementary channel.