Stock market prices are procyclical, while investment good prices are countercyclical. A real business cycle model calibrated to these observations implies that 75% of the cyclical variation in aggregate output is due to an investment-specific technology shock, while the rest is due to an aggregate productivity shock. To test this conclusion, we investigate the model's implications for asset prices and business cycles. The model does not do significantly worse than existing models on these dimensions, and on two dimensions it does notably better. It is consistent with the facts: (i) employment and investment across different sectors comove over the business cycle: and (ii) high interest rates lead low aggregate output. Fact (ii) is often interpreted as reflecting the business cycle effects of monetary policy shocks. Our result suggest that (ii) may, at least to some extent, also reflect the effects of real shocks.