In this paper, we focus on the transmission mechanism of quantitative-easing (QE) policy and address which matters most: the size of the bond purchases in each period (flow effects) or the total amount of bonds taken away from the private sectors (stock effects). To this end, we estimate a dynamic stochastic general equilibrium (DSGE) model in which short-and long-term bonds are imperfect substitutes due to market segmentation and preferred habitats, using Japan's data from the 1980s to 2017. We find that (i) Japan's data support imperfect substitutability between short-and long-term bonds, which implies that government bond purchases by the Bank of Japan (BOJ) reduce the term premium, exerting an expansionary effect on economic activity and prices; (ii) government bond purchases by BOJ have compressed the 10-year yields by 50-100 basis points as of the end of 2017; and (iii) the compression of the term premium has been mainly driven by stock effects, which underscores the importance of the size of BOJ's government bond holdings in determining the term premium.