Abstract
Narrative economics holds that stories reshape collective motives and, through them, asset prices. Yet the field lacks a framework for decomposing narrative content into interpretable psychological channels with testable implications for risk. We introduce the vector of needs (VoN), a five-dimensional coordinate system comprising security, status, control, empathy, and transcendence, constructed as a minimal spanning vector across canonical motivational theories. Using propositional anchor sentences and a sentence encoder applied to 4.5 million US news headlines between 2007 and 2022, we measure daily narrative alignment within each dimension and test whether these motive-specific signals carry information about market returns, volatility, tail risk, factor premia, and cross-sectional covariance structure. We find a sharp bifurcation: empathy and security operate as crisis-anticipating regimes, forecasting elevated volatility and tail severity through distinct channels. Empathy compresses the return covariance matrix toward a single factor, while security increases cross-sectional dispersion along a safety gradient. Transcendence is the dominant predictor across all market moments, simultaneously forecasting higher returns, lower volatility, and a dramatic decoupling of industry correlations. The framework explains about 18.7% of the variation in future (out-of-sample) realized volatility, indicating meaningful predictive power beyond the data used to estimate the model.