Economics Working Paper WP16113
Abstract: We examine the first QE program through the lens of an open-market operation under- taken by the Federal Reserve in 1932, at the height of the Great Depression. This program entailed large purchases of medium- and long-term securities over a four-month period. There were no prior announcements about the size or composition of the operation, how long it would be put in place, and the program ended abruptly. We use the narrative record to conduct an event study analysis of the operation. To do this, we construct a dataset of weekly-level Treasury holdings of the Federal Reserve in 1932, and the daily term structure of yields obtained from newspaper quotes. The event study indicates that the 1932 pro- gram dramatically lowered medium- and long-term Treasury yields; the declines in Treasury Notes and Bonds around the start of the operation were as large as 128 and 42 basis points respectively. A significant proportion of this decline in yields is attributed to the portfolio composition effect. We then use a segmented markets model to analyze the channel through which the open-market purchases affected the economy, namely portfolio rebalancing and signaling effects. Quarterly data from 1920-32 is used to estimate the model with Bayesian methods. We find that the significant degree of financial market segmentation in this period made the historical open market purchase operation more effective than QE in stimulating output growth. Additionally, if the Federal Reserve had continued its operations in 1932, and used the announcement strategy of the QE operation, the upturn in economic activity during the Great Depression could have been achieved sooner.