This paper presents a model of alternative sources of credit – bank vs. bond finance - to examine the credit substitution hypothesis. Our framework produces testable hypotheses about the behaviour of price- and quantity-based information variables. Examining data from ten Euro-area countries, we find that a credit spread outperforms a finance mix as a predictor of economic activity in both time series and pooled data regressions. There are clear signs of asymmetric and maturity effects in the data. Positive changes in the credit spread predict decreases in economic activity while negative changes bear no informative content. The asymmetric effect is exceptionally strong in pooled data and is present in short-term, long-term, and total credit spreads. In country-specific time-series regressions the asymmetric signalling property is strongest for the long-term credit spread. By contrast, we find no substantive evidence that changes in a quantity-based finance mix have robust predictive power.