Macroeconomic and financial crises exhibit structural asymmetries, skewing risk distributions rather than strictly shifting the conditional mean. This paper introduces Quantile Bayesian Predictive Synthesis (QBPS), a dynamic framework optimizing forecast combinations directly at specific quantiles. To solve the computational "zero-variance trap" of standard normal-exponential mixtures at extreme tails, I develop a novel scaled parameterization for the Asymmetric Laplace predictive likelihood. I further introduce an observation-driven asymmetric discounting mechanism anchored in the Generalized Autoregressive Score (GAS) framework to trigger rapid parameter adaptation during sudden structural breaks. Applied to forecasting the extreme left tail of US real GDP growth (Growth-at-Risk), QBPS yields competitive out-of-sample predictive accuracy relative to standard machine learning, symmetric Bayesian, and two-step parametric benchmarks. Finally, the framework's dynamic weights enable real-time predictive tail risk attribution: while financial subcomponents share predictive value during economic expansions, the forecast decomposition isolates intermediary leverage as the primary leading signal for financially driven downside risk, revealing distinct predictive signatures for endogenous market stress versus exogenous real shocks.

Moving beyond conventional mean-targeted forecasts, I predict the entire distribution of bond risk premia to capture extreme, crisis-driven tail movements. Augmenting standard predictors with an extracted upper-tail macroeconomic factor—representing a latent "macro-panic"—achieves full predictive dominance over baseline models across the quantile spectrum. Methodologically, a Quantile Random Forest (QRF) efficiently captures non-linear crisis thresholds using mean-selected predictors, offering a computationally tractable, high-performing alternative to Quantile LASSO. The resulting forecasts exhibit pronounced regime-dependent cyclicality, becoming heavily countercyclical during recessions. Crucially, this approach reveals a term-structure duality: short-term tail risks are driven by acute macro-liquidity panics ("dash for cash"), whereas long-term risks are anchored by structural safe-haven demand ("flight to safety"). Finally, I document a decay in macroeconomic predictability along the yield curve: while the QRF dominates at shorter maturities, naive empirical quantiles become highly competitive at the long end.