finance theory

Classical finance theory suggests that firms take projects with positive NPVs regardless of the amount of cash the firm has available. However, empirical evidence suggests that the amount that firms invest is heavily dependent on their available cash flows. Why might this be?